MBA Mondays: Revenue Models - Subscriptions

When I got into the venture capital business in the mid 80s, software was sold. You would pay a large sum to acquire a license to the software and then a much smaller annual fee for maintenance and support. Today, most software is sold in a subscription model. You pay a monthly fee for the right to use the software. If you stop paying the monthly fee, your right to use the software goes away. Maintenance and support is bundled in.

The emergence of the subscription model has made the software business better. In the old upfront license fee model, software companies would trade at 2-4x revenues. Now they trade at 6-8x revenues. That reflects the recurring, almost annuity nature of the subscription model.

Software is not the only technology oriented business that utilizes a subscription revenue model. Content has also moved from an upfront fee (buy an song or a movie on iTunes) to a subscription model (a monthly fee for Spotify, Rdio, Netflix, or Hulu).

And infrastructure is now also sold on a subscription model. Amazon Web Services (AWS) is a great example of this. Need a server? You can provision it for yourself in the cloud and pay a monthly subscription for it. Same with storage and a host of other infrastructure services.

The emergence of the subscription model for software, digital content, and infrastructure has led entrepreneurs to offer all of these in limited form for free with a paid upgrade to a monthly or annual subscription. The term for that form of subscription is freemium, a term that was invented by Jarid Lukin in the comments on this blog in March 2006.

Subcriptions can be paid monthly or annually in advance. Many companies opt for the latter model and that works really well because it creates a favorable cash flow dynamic in the business. Cash comes in before most of the revenue is recognized, leading to a healthy and predictable business model. In addition, the subscription business model allows a company to book most of the revenue for the year in advance. Companies with subscription based revenue models often have great visibility into the next twelve months of revenues, a feature which makes for a great public company.

The big gotcha in subscription revenue models is churn. If you churn more than 10% of your customers every year, subscriptions can be a challenging model. You need to grow new customers at 10% just to stay even. I encourage our portfolio companies that utilize a subscription model to be very active at managing customer satisfaction and to actively monitor the customer's usage of the software or service to identify customers with high churn potential. This kind of data can be automated and leveraged to proactively manage problem accounts and reduce churn.

The first several years of a subscription based business will typically require a fair bit of funding because the revenues come in over time instead of up front. But once a subscription business reaches scale, it has very favorable cash flow dynamics, as mentioned above. For these reasons, subscription based businesses are good businessed to raise capital for and investors generally find them attractive to invest in.

In many ways, the subscription business model is the most attractive of the "big three" (advertising, commerce, and subscriptions), all of which we have now covered in this MBA Mondays series on business models. Subscriptions are more predictable and reliable and as a result create more investor confidence leading to higher multiples and more valuable businesses. Of course that is not universally true of all businesses, but I have found it to be generally true over long periods of time and many different businesses.

#MBA Mondays

Comments (Archived):

  1. jason wright

    well, as it is the new year, are people being seduced by the great gym membership..ripoff?i hate gyms πŸ™‚

    1. awaldstein

      I’m completely seduced by fitness as part of health and my gym time is a daily reward :))

      1. fredwilson

        i am with you arnold. i love my gym.

        1. Duncan Logan

          Modern Gym Membership is to pay if you don’t go. Co revenue increases later in the year but people are less likely to cancel their subscription.

          1. pointsnfigures

   30 min a week. eat right and don’t worry about the rest. lost 25 lbs in 2011. kept it off.

        2. LE

          I luckily have enough room (that’s suburbia) to have whatever equipment I need and don’t have the portal to portal time of a gym (or the germs).My latest addition (and I HIGHLY RECOMMEND THIS) is a water rower:…It’s small enough and nice enough (natural wood) to have in any room in the house. Very classy and makes a great whoosh noise when you row very relaxing. You could even keep it at the office.


          I made my *last stand* at 42 just a few years ago. I benched 305 (wanted to do 315) and squated 495 (wanted to do 500). A hip injury while deadlifting put the end to my power desires..I’m’ moving to a *fitness* approach now. It’s gonna’ suck. Well maybe some dumbells here and there. But, nothing over 100lb’ers.

      2. jason wright

        my bicycle is my gym – maybe i’m a little claustrophobic πŸ˜‰

        1. awaldstein

          Big urban bike rider myself.Honestly though, depending on your goals this is not balanced enough exercise for me. From a pure exercise perspective anything that doesn’t bring in your core and combine aerobics and focused muscle is not inefficient.I’m an extremist though…

    2. Aaron Klein

      It’s true. They will all cancel and I’ll have my parking spot back in a few weeks.

      1. awaldstein

        What’s a parking spot?

        1. Aaron Klein

          You adorable New Yorkers. πŸ™‚

    3. JimHirshfield

      I’m with @awaldstein:disqus on this one. Morning swim is part of my daily ritual and worth the abundance of cash the gym vacuums out of my bank account on a monthly subscription basis. But yeah, I hear ya that lots of “breakage” in this model where the customer’s intention doesn’t match their behavior.

      1. Anne Libby

        Where do you swim?

        1. JimHirshfield


    4. ShanaC

      i love gyms. Try lifting

      1. jason wright

        umm, how much do you weigh?

        1. ShanaC

          i’ve been missing the past few weeks. (sick, stress, holidays). Normally though I am high rep low weight. I regularly do 20 reps 2x of 40lb deadlift-highpull though. Which is high for a girl from what I hear

  2. Driftaway Coffee

    there’s still a big difference in enterprise vs. consumer on this. enterprise still sells licenses vs. consumer app. sure would be nice to see a photoshop license renew every month!

    1. kidmercury

      have you tried adobe creative cloud? that is a subscription service to the whole adobe creative suite, photoshop too. I LOVE IT. all caps, that is how much i love it. frankly i should go bold too. it is great for everyone — adobe will end up making more money on this in the long run, and i love having the latest version of all adobe software at my fingertips at all times on the computers i want them to be on.

      1. falicon

        Wow. that’s a great find…I’m sold. Can’t believe I didn’t know about this already…very awesome. Thanks!

        1. kidmercury

          they’ve done a terrible job of promoting it…..i know so many web developers that haven’t heard about it.


            Then it will probably canceled in a month. It seems marketing is the only way to make things profitable today. I see so many great tools come and go because of poor marketing!.Did I mention HealthRMS or SafetyRMS? lol

      2. awaldstein

        I need this now! Thanks.

      3. Aaron Klein

        +1. Totally agree. We have two seats already.

      4. Richard

        Thoughts on their behance aquisition

        1. kidmercury

          not sure, i think there are potential synergies and behance seems like a great business from my perspective as a casual observer, but lets see if adobe can integrate it properly though or if they will pull a yahoo…..

      5. Driftaway Coffee

        thanks. i knew it existed but haven’t given it a try. thought it was in beta / dummed down. will give it a try now.

      6. takingpitches

        $50 bucks/month. Sweet – thanks for the headsup!

      7. ShanaC

        and a lot of people also hate it. They think early versions of the software work just as fine, and can be had for a lower cost over time. Adobe is having some serious problems converting over their hardest core users as a result.

    2. JimHirshfield

      But enterprise is transitioning more and more to subscription, don’t you think?

      1. Driftaway Coffee

        they should. pace of change in enterprise is much slower, even though employee’s individual experiences as consumers are affecting enterprise more and more.

  3. kidmercury

    there are countless reasons i love targeted niche/vertical businesses, and subscription opportunities is one of my favorite reasons. i think it works much better on a niche basis and enables greater share of customer wallet (aka lifetime value of customer). because it requires a high level of customer satisfaction, i think it works better on a niche/vertical basis as does have their subscribe n’ save feature, which is an example of subscription in non-niche. i guess those sams club type things is an example of how this might scale in a low margin business.

    1. takingpitches

      Amazon used to offer 15% off on subscribe and save products, but in the last year or so, reduced it to about 5%.Anything is better than nothing, but the problem — as a Prime subscriber — is that it takes longer for the items to arrive if I go the S&S route rather than the Price route, so I often find myself in the cognitively dissonant position of buying the product at the higher price.They should fix this.

      1. ShanaC

        yes. It is often cheaper to buy locally and stock up because of this weirdness

      2. kidmercury

        yeah i don’t really use subscribe and save anymore — the discount is too small to warrant the commitment and inconvenience. amazon does use dynamic pricing and so perhaps they are just screwing you and me. πŸ™‚ i seriously have begun shopping a bit less from them, switching a bit more to or sometimes even daring to go real world when i see a better deal, although i still buy the vast majority of everything from amzn.

  4. karen_e

    A good offshoot of this could be Fun Friday Subscriptions: What Do You Subscribe to Monthly? (New York Times and Spotify.) Annually? (Memberships to three local cultural organizations.)

    1. William Mougayar

      I like that one. It would be a real market test for what people pay. We should also add the subscription services that companies pay for too, e.g. UserVoice, Relic, SendGrid, etc…

    2. fredwilson

      that’s a good idea

      1. Techman

        I look forward to Friday then.

    3. JimHirshfield

      Ha….”What’s in your wallet?” theme.

    4. Duncan Logan

      +1. What have you subscribed to for more than 2 yrs?

    5. jason wright

      Spotify hasn’t got it right. the music ads i MUST listen to on the free service never chime with my listening list and musical taste as expressed through my usage of the service. i MUST to stuff that does not move me. i don’t get their process.

      1. Dan

        Are they intentionally obnoxious to drive you to convert to get rid of them?

        1. jason wright

          they’ve driven me to uninstall their software. it’s cow pat.

          1. Anne Libby

            Cow pat. Hah!


      I only have one monthly subscription based discrentionary spending item. I’ve been moving to on-demand-only. The reason is becasue on-demand gives me control. I *grab* what I want when I want it. Also, no more subscription based software. I want to be able to keep a prior version around just in case!

  5. Mancel

    I feel like Subscription based businesses are also good because the customers form some sort of bond and hence loyalty with the business which makes it tough for them to leave unless you truly screw them over.

  6. William Mougayar

    Do you think that subscription models can co-exist with a “large network of engaged users” strategy where the revenue will come later from an atomic unit?Another way to ask the question is- If you see a very early subscription model (e.g. RebelMouse, GetLittleBird) at the seed stage, is that a warning signal that the busines will be stuck with that model, or do you think these companies are testing things out and might change to another revenue stream later?

    1. Aaron Klein

      We shipped Riskalyze Pro with the premium part first. Now we’re about to ship the free piece and there is some fascinating atomic unit revenue potential there.Shipping it “backward” has been really great for us though. We’ve brought just the right customers aboard to figure out the best parts of our roadmap to ship first.

      1. awaldstein

        I love this…Feedback in a scattered market is honestly very tough to get. Starting with transaction connected paying customers first is clever.

        1. Aaron Klein

          I have to admit that we stumbled into that benefit but it’s been awesome to get that particular slice of market feedback.

    2. JimHirshfield

      “Do you think that subscription models can co-exist with a “large network of engaged users” strategy where the revenue will come later from an atomic unit?”Happening right here, right now. πŸ˜‰

      1. William Mougayar

        But are they both expected to grow equally big or was one put forward while the other one was getting ready?

        1. JimHirshfield

          Will the revenue and the large network of engaged users grow equally? Yes, I think they can, but that doesn’t mean everyone in the network is paying (see “freemium model” or 1/9/90 rule).And no, they don’t have to happen coincidently.

    3. Richard

      “large network of engaged users” the quote of the year!

    1. Ptaco

      It’s ironic that companies such as apptegic and servicesource came into being because of the B2B subscription model. You can also argue they came into being because companies fail to serve their customers. Subscription model encourages and reward companies that actually care about the value their customers are realizing.

      1. Karl Wirth, CEO Evergage

        There are many things that a suscription company should be doing to serve its customers. This includes building a great product, making it easy to use, continually adding value, providing excellent support, and listening to customer feedback.But, in an online business, where you can’t talk to every customer personally, how do you know the value that each particular customer is realizing with your service? And, how do you communicate effectively to each customer to help them get more value from the service?I think that you do it by understanding what each user is doing within your service and by communicating to them within that experience the information they need to be successful. We are seeing companies, who are already very good at serving their customers, serve them even better by using our solution to do this.

    2. fredwilson


  7. awaldstein

    On a tactical marketing thought, personally disdain all one month free then auto over to paid subscriptions.From company side logical as u need accounts for free but just always feels posed.

    1. William Mougayar

      Just clarifying…are you saying you don’t like “try before you buy”?

      1. awaldstein

        No love freemium but don’t like free trials with automatic paid roll in. Different.

        1. William Mougayar

          Automatic with warning 1-2 days prior is ok?

          1. awaldstein

            This is an impossible marketing message to give.’we will take your credit card and warn you 2 days before we start charging”Putting that in a phrase to encourage me to think about the value of the service just ain’t happening.

          2. Kirsten Lambertsen

            Totally agree with this. Always scares me away.

          3. PhilipSugar

            I love Credit Karma’s advertisements on this….if its free why do you need my card. I also agree with LE there is an incentive to make it very hard to cancel. That always seemed really sleazy. It must work, but it is not something I want to do.

          4. awaldstein

            I’ll use all the methodologies at my disposal to keep customers, maximize revenue and on and on.But…when you are starting from a dead stop and need to build community and users and have them be your voice to a larger market, you need people that love what you do or offer.Solve that problem then be smart with everything else and you have a better chance. That’s how I want to play the game.

          5. ShanaC

            true. But if you have a mewing community that doesn’t pay for your service, then you have other problems

          6. awaldstein

            ‘mewing community’….please clarify.

          7. Richard

            Im very reluctant to subscribe to anything with a free to autopay model. Believe in your product!

        2. JimHirshfield

          Agreed, but it’s all in the presentation. Sneaky is not a marketing strategy.

          1. Matt A. Myers

            Sure it is, but it’s a bad one.

          2. JimHirshfield

            true enough

          3. Anne Libby

            “Sneaky is not a marketing strategy.” Upvote.

          4. JimHirshfield


        3. fredwilson

          i agree

          1. Richard

            TIme to put this to the A/B test?

        4. James Ferguson @kWIQly

          Agree totally – I want to commit to pay when I know I am happy to pay and not before – Especially on-line where a shopfront is wholly anonymous and means very little in terms of transparency

        5. Abdallah Al-Hakim

          especially those that automatically enroll you in a annual subscription

      2. Aaron Klein

        I think it’s the idea that half your new paid accounts are people who forgot you were going to charge them and had no intention to buy.They make sucky long term customers and your brand suffers as a result, imho.

        1. awaldstein

          Earned customers are the ones you want w/o doubt.

        2. PhilipSugar

          That’s exactly right. Sucky long term customers. But if you are living for the moment trying to make a number, then it must work.But again that is not what I want. It is a strategy and it does work, which is why I never want to build stuff to “flip”

          1. Aaron Klein

            It’s abundantly clear when you are “burning the future” to make short term numbers and it always bites you in the end.Slap me if I ever do this.

    2. PhilipSugar

      I can say from a BtoB perspective its actually better to charge an installation fee upfront, I’ve tested this.The problem with even saying free, is that you get no commitment. You might feel better short term, because you “sign up” more accounts but the biggest problem is this: Salesperson says try it what do you have to lose? And you’ve now turned your implementation people into the salespeople with no hammer.The churn rate (I don’t even know if you should call it that) during the first three months is insane.

      1. awaldstein

        Heard some saying that b2b abd b2c models are becoming the same.I don’t buy it. Different behaviors. Different approaches. Different ways of budgeting and thinking about investments.

        1. PhilipSugar

          Agree completely. For individual use and reimbursement the are becoming the same. I give examples of cell phones, applications you might use like Dropbox or Basecamp. Eventually I believe BYOD will include your laptop.But core systems? No. There is too much to lose. I got called out for being anti-big business on Brad Feld’s Blog but I’m not. There is a reason why systems and incentives are setup to insure screw ups don’t happen…that’s because a screw up can cost a ton of money.

          1. ShanaC

            Do you think this means it is going to become less clear about who owns what device because of local vs cloud storage?

          2. PhilipSugar

            No, it makes it simpler, you own the device, you store nothing locally on the device. (with the exception of encrypted applications) I know people in NYC and SF find it quaint to own a car, but lots of us do. We use it do stuff for the company (like go the airport) and we get reimbursed.The company doesn’t want to own the car because there are all sorts of issues around that. Just like the company doesn’t want to own your phone. Now we have to keep track of how often you call Aunt Millie, or what apps your kids have on the thing.No, we’d rather figure out how to lock it down. (Password) what policies we have towards customer info on it and then deal from there.

      2. ShanaC

        ouch. why bother then?

      3. K_Berger

        I totally agree with this. Your customers have to have some skin in the game.Another problem with free is that most people equate free with worthless (or not worth much). When we were just starting out, we had a 6 month free trial. Across the board there wasn’t much usage. When the trial was over, some stayed and some left, but the ones that left were a waste of time from the beginning, we just didn’t know it at the time.

    3. Dave McLaughlin

      What do you folks see as the relative merits of Freemium vs Free Trial (assuming you don’t auto-enroll the free trial participant for the reasons mentioned here)? What situations call for one vs the other?

      1. awaldstein

        I’ve just learned my lesson around this (again!) on a personal project.The more you tear down the walls and give value without asking for a signup, the more the value that drives a signup becomes crisper and the better the systems works.People will pay. Don’t annoy them just let them experience value first.

        1. Matt A. Myers

          “People will pay. Don’t annoy them just let them experience value first.”Extra upvotes for this.

    4. LE

      Only thing worse than that is the same thing where there is a significant amount of effort and friction required to cancel the plan and/or get charges reverse.Note to anyone out there that expiration date of credit card does not block a company from continuing to charge your card. Under a certain dollar floor you can easily recharge a credit card and use any expiration date and the charge will go through for the majority of cards. One of the disadvantages of asking a customer for a new card is that it gives them an opportunity to say “um just cancel that”. So the ability to continue to charge an expired card is important.

      1. ShanaC

        ouch. Why is crediting a canceled card allowed?

        1. LE

          We’ve been doing this since the upper 90’s. Can’t remember a single person who ever said “why did you charge my card it had expired?”.You give someone a credit card and you (supposedly) have a contract with them. The contract doesn’t state “this contract is only good while your credit card is valid and not expired”. The fact that the card expired does not cancel the contract just like the fact that a check bounces doesn’t cancel a contract unless of course for some odd reason that’s in the contract.I think you mean “debiting” by the way. And the card isn’t cancelled it’s expired. Of course when cards are replaced with new card numbers if you are setup with automatic renewals the credit card processor will cheerfully give the merchant the new card number.

          1. Anne Libby

            This wasn’t my experience this past summer when I accidentally tossed a card. Maybe a lost card is treated differently, as it’s assumed to be possibly stolen?

          2. LE

            It depends on who the merchant is and specifically if the charge is marked a certain way as recurring.In general most people with recurring charges (from reputable companies) who lose a card don’t even know how to contact and update the card with all the people who are charging them which could be many. So this is actually a big benefit. For example I pay $100 per month for storage. In no way do I want that charge to not be paid so they auction off my belongings (say email is bad etc.)

    5. Matt A. Myers

      Agreed. Freemium is so much better anyway. If a product isn’t useful enough for someone then they won’t use it anyway. If they can use it freely as long as they want, maybe don’t even actually need it immediately but doing research in advance, they then can get a feel for if they like it and essentially teased about additional features (or just use up what capacity that the free version has, and then required to upgrade).

    6. ShanaC

      well, what better methods do you think work for subscriptions?

      1. awaldstein

        Freemium, no wall and open for all.Earn the account and the registration. New rev of theLocalSip for example will have the wall completely down and open, with reg needed for certain actions like a transaction or RSVP and to get alerts and newsletters.

      2. Aaron Klein

        I’ll let the cat out of the bag for Riskalyze Pro.We’re about to ship a free version of our platform for advisors. You’ll get less insight into your clients, and you get a generic set of ETFs to use in your model portfolios.We’re lucky that those are tradeoffs most advisors will not want to live with, but they can still get a ton of value out of the free product. So they bump into those limitations all the time in a way that makes them say “gosh, this is working so well. I feel like I could totally land that new client if I could personalize this to them.”Click to upgrade. Enter credit card number. Boom, it’s activated.That’s an intentional sale, as opposed to “use Riskalyze Pro FREE for 30 days” that requires your credit card and automatically signs you up afterwards.I refuse to do those trials any more.

        1. Anne Libby

          Good call.

    7. Anne Libby

      I’ve gotten burned on this by my own inattention so often that I’m done with them: the product would have to be a date with George Clooney for me to buy under this model.

  8. Brandon Burns

    The “gotcha” is a really big help, I don’t naturally think to look out for those because I don’t know what the are β€” please include all the “gotchas” for the upcoming posts!And putting all of this trough the light of how investors think about something vs. another sheds a light that most of us here do not see. Again, please make that a regular occurrence in this series, and wherever applicable on AVC. :o)

    1. Brandon Burns

      Unrelated, but several people here have mentioned that I should blog about UX. I started, and have actually managed to keep it up. A new leaf has been turned.http://lifeexperiencedesign

      1. fredwilson

        that’s great. a tumblog to boot!

        1. Brandon Burns

          disqus commenting is in there, too. gotta support the community. πŸ˜‰

      2. takingpitches

        Thanks for blogging man! Look forward to reading.

        1. Brandon Burns

          no, thank *you* for urging me to do it. hope to keep it up everyday. it really helps me think through things.

          1. takingpitches

            nothing better than writing it to think through things. Congrats!

      3. Kirsten Lambertsen


        1. Brandon Burns

          thanks for the vote of confidence!

      4. ShanaC


        1. Brandon Burns

          thanks for the love! follow me :o)

          1. ShanaC

            i did, and sent it to people

  9. Aaron Klein

    Subscription revenue is also an amazing competitive weapon. In my last job leading product for a division of a publicly traded multinational, I strenuously tried to take our core business to a subscription model.And failed. The addiction to hit-driven pops in revenue was strong and no one was willing to go through the 2-3 years of extra work and pain to pull off that transition.

    1. Ptaco

      I agree – the subscription model forces sales organizations to change their motivation. Best performers typically shun subscriptions unless their company can upfront the annualized “big hit”.

      1. FlavioGomes

        The smart sales org would prefer the sub model and consistent income. But be very careful. Over time the annuities create fat cats that still get commish even if they don’t sell anything new. Manage renewal periods and differing commish structures so that attracting new business has strong incentive.

        1. JimHirshfield

          Don’t pay commish forever. Set the period upfront, even if client relationship continues on.

          1. FlavioGomes

            Thats a tough one to implement as the sales person always views it from the perspective of attracting the overall deal. Would you base it on the first contract term? Nothing for Renewals?

          2. JimHirshfield

            I go with first year’s revenue; paying a percentage; spread out over the year. Salespeople bring in the deal, the product and the company as a whole keep the client. So beyond first year, I don’t pay commish. The balance can be tweaked by what percentage commish you pay on the first year’s revenue…dial it up/down as you see fit.Renewals that are net increase in revenue would trigger add’l commish, in my view.

          3. FlavioGomes


          4. JimHirshfield

            You’re welcome

          5. PhilipSugar

            Yes, the issue is if you have a pay as you grow model. In that case you have to pay on increase like new sales, otherwise you get salespeople swinging for the fences which short term can hurt your sales because it increases the cost of entry and long term can hurt your sales because when you pay ahead of value in a subscription model it really hurts your churn.

          6. JimHirshfield

            I think I follow you, just not sure what you meant by “increases the cost of entry…”

          7. PhilipSugar

            You, k_berger, and me good have a good chat because it seems we all have lived this.What I meant was an example of where the cost of the subscription increases. This happens quite a bit, more seats, more usage, etc. This by the way is the Holy Grail of churn because if you can get existing customers to pay more your churn rate can actually be positive not negative.So if you only pay the salesperson the new sales rate commission for the first year of sales, they rightfully will set a high minimum/also feel screwed as they watch the account grow but don’t get paid. You have to pay the new sales commission rate on the increase.This is the real long term view of the world because what you are doing is getting into the account at a low cost of entry, proving value and then increasing the cost as you provide more value. So if you only pay commission during the first year, the salesperson instead will want to set the contract for the highest initial rate possible. That makes the sale harder and also increases churn because when the client isn’t using all the seats/however you charge they are more likely to churn

          8. K_Berger

            I agree with this. I think it also depends on your overall churn rate. We had a churn rate of less than 3%. Since clients tended to stick around, no need to incent the salesperson based on renewals.Having said that, we were toying with the idea of a small commission if a client was signed to a 3 year term versus a 1 year. While the odds are they will stick around, there is some value in making that more of a certainty.

          9. JimHirshfield

            good point

          10. PhilipSugar

            I think we probably could have a great talk.We have a very low churn rate as well, but I felt that if the salesperson wasn’t going to get the benefit of a big one time check when they signed the deal, they should get the upside of a small commission on the renewal.I didn’t like incentivizing on the longer term because I found then salespeople played games on the terms.I actually have customers push for the longer terms for price certainty.

      2. Aaron Klein

        Another approach is to advance commission up to a cap (like increased base salary) for a short term period during transition. Salespeople are smart…if you can make sustainable better for them over quick hit revenue, they’ll go there.

    2. PhilipSugar

      Changing the model is brutal.

      1. Aaron Klein

        So true. But death is worse.

    3. JimHirshfield

      That’s unfortunate and indicative of hit-and-run style business. I can’t imagine that clients were as happy as they could have been in a subscription model. I think saleforces work harder to keep clients happy in subscription models because they want to keep the subscription revenue coming in. In the old model of upfront sales, it’s wham-bam-thank-you-sam (I just made that up to be PC)…then they wonder why years later the client upgrades to competitor’s product.

      1. Aaron Klein

        Totally agree. Incentives get aligned across the company.

    4. FlavioGomes

      It is an extremely difficult transition to make from one time hits to a recurring stream. Offer both models and slowly wean from one to the other as the opportunities present themselves.Easier to do with B2B on a case by case basis.

      1. Aaron Klein

        Argued and lost. Execs senior to me simply wouldn’t go for it.The company has suffered as a result. I have a lot of friends who lost their jobs there (long after I left).

    5. fredwilson

      it’s a hard transition to go from up front sales to subscription

      1. awaldstein

        Really tough comp model and challenging building a blended model sales org that has both hunters and farmers.Be great to have a sales exec/comp pro (Phil Sugar) do a post on that. Taking the abstract model to a comp based org discussion would be thought spurring.

        1. PhilipSugar

          Its not so tough if you start that way. You can give a high commission percentage on new recurring revenue for the first twelve months (both the initial and increases) and a small slice of ongoing for lets say the next five. The salesperson is actually very aligned with the business versus only caring about the initial sale and the hit of revenue, literally not giving a shit the day after. Their yearly raise is baked into the plan via the renewals.No you are not going to get the Oracle, IBM, SAP reps, but that isn’t who you want. Frankly even for those people they realize those days are over, but I don’t want to be the one that resets expectations.Changing the model is just brutal, both internally and externally. Internally you are going to take a 70% revenue hit, and externally your customer is going to say I own this, I bought this, now you want me to rent this? I’ve looked at this, and I don’t have an answer, other than to literally start a new company to compete with the old. I’ve seen people try a hybrid approach and it is a huge mess.

          1. awaldstein


      2. Aaron Klein

        Indeed. Which is why it’s such an amazing competitive weapon for disrupters. πŸ™‚

      3. ShanaC

        what are some good ways to manage the sale

  10. Richard

    Any sense for the number of paying but non-using subscribers a typical b2C software company might have? Β 

    1. William Mougayar

      Yup, I’d love to learn about these numbers for HootSuite and DropBox for eg., although DropBox straddles b2c and b2b.

      1. Richard

        I assume that the b2c > b2b (btw, maybe b2b/b2c are outdated, seems like the the b is superfluous?)

        1. PhilipSugar

          No B to B still applies. That is when there is a centralized payment. I can tell you in that case the number of non using payers goes almost to zero after one budget cycle.

          1. Richard

            Wow, that’s quick.

          2. PhilipSugar

            If you are getting real money from the enterprise (which you have to do because it is so expensive to sell to them) it will be a line item in a yearly budget. If there is no use then it goes away because its a big number.

          3. K_Berger

            On the other hand, if you are a small-ticket item, then you can fly under the radar for a while.The problem then becomes this: you have some customers paying but not using. Now you want to help them start using it again but are afraid to draw attention to the fact it isn’t being used. It is pretty clear that the right move is not to ignore them, but there will be a revenue hit in doing so.

          4. PhilipSugar

            Yes, if you are small ticket you can fly under the radar for a long time. And you bring up a conundrum. If its big ticket you know if its not used they are going to cancel so you work to get it used. If you are small ticket that’s an interesting trade-off.

        2. ShanaC

          no. i think as small businesses grow more common, people will feel more of a need for downtime

    2. JimHirshfield

      I find it hard to believe that there’s a one size fits all number for this. So many different parameters: type of company, product, price point, use cases, culture, etc.

      1. Richard

        My guess is that consumer characteristics factors are the principal components (particularly with auto-payment)

        1. JimHirshfield

          Sure as @jasonpwright:disqus point out below, prevalent with gym memberships. But I’ve been thinking of this blog post as it relates more towards software, and because of my background, more so enterprise.

        2. Anne Libby

          Auto-payment: most annoying thing ever. (And sometimes dancing on/over the border of ethical.)(Pro-tip: if you’re having a hard time unsubscribing: lose your credit card. Easy unsubscribe, so much so that I’m thinking about “losing” a card more often.)

    3. fredwilson

      meaning likely to churn off?

      1. Richard

        no, perpetual (3months-24 months) payer/nonuser?

    4. Chris Phenner

      At a previous, subscription-based music firm at which I worked, we called this ‘The Sludge.’ And I agree with Jim H. & Fred that these kinds of users will vary widely from service to service, and that these users are not attractive because they are likely to churn once they ‘wake up’ and realize they’re still paying.

      1. ShanaC

        How fast does this occur?

    5. muratcannoyan

      Very good question. My company helps dentists build their referral base on a social platform, and simplifies ecommerce. So they need to invite/engage their patients to get value from the platform. Getting them to pay for the monthly sub is one thing, getting them to use the platform is another. It’s been a interesting lesson.

  11. JimHirshfield

    Love this model. As compared to advertising or commerce, I don’t need to resell product to my existing customer base next month. I know how much revenue, a min, I’ll have next month.As for churn, keeping customers happy and relationships genuine (even if at times they’re not happy) is key to maintaining their trust and business.

    1. karen_e

      You must be great at what you do, Jim. I mean it.

      1. JimHirshfield

        Flattery will get you everywhere πŸ˜‰

  12. Ivan Yuen

    Of course there’s subscription commerce as well. Consumers pay for convenience (monthly supply of razors) or the element of surprise (cosmetic samples) This model is particularly interesting because there are network effects that create value as the number of subscribers grow and manufacturers would compete to get access to these consumers.

    1. JimHirshfield

      Hybrid FTW

      1. ShanaC

        i don’t see them as hybrid. I see them as variations.

  13. PhilipSugar

    I love subscription models. I wrote an article in 1999 describing the benefits. I’ll find it and post a link. The impetus was Y2k. As 1999 drew to a close new license revenue for software companies literally dried up because nobody wanted to install any new software before the year 2000; that really sucked.You are right that revenues grow more slowly. If you think about it you could sell a license for lets say $100k, that would translate into a monthly sale of roughly $2-$3k. Because sales happen on average in the middle of the year and you only recognize monthly that translates into only $15k to $20k of revenue in the first year.It is in the following years the leverage really applies, because you can concentrate on making your software good and still get paid, instead of working on new features to try and sell upgrades which might just schlock up your software….for example see Microsoft. Once you hit breakeven the gravy train starts….but don’t kid yourself it is brutal ride over the mountains of little revenue and tons of work to get there.There are certainly benefits for end users. If you don’t use you don’t pay, and you are right it is extremely important for the company to focus on satisfaction because shelfware disappears.New customers are not the ones dictating features because under the one time license the huge money hit is like a crack pipe, you can’t give it up. A new customer gets their pet feature put in because “you can’t walk away from the sale”. This is a good thing because it craps up your product.And actually that’s why I don’t like to try and get a 3 year subscription upfront. I’d rather get it monthly or quarterly because just like Cortez burning his ships, everybody has to focus like hell on customer satisfaction. You also don’t get blasted so badly during the negotiation because both sides realize you aren’t arguing over a lot of money short term.Finally even though you should be able to account for this in either scenario nobody does: When somebody actually buys and uses your software there is an ongoing liability to that. They will want stuff to get their jobs done. When you take all of that money as a one time hit there is a temptation to spend that money when you get it and that makes for lean times down the road.

    1. awaldstein

      Smart…like the Cortex metaphor. Comp driving behavior is the key to sales performance and subscription ain’t always natural for sales.

    2. acoustik

      “New customers are not the ones dictating features because under the one time license the huge money hit is like a crack pipe, you can’t give it up. A new customer gets their pet feature put in because “you can’t walk away from the sale”. This is a good thing because it craps up your product.”That is so true. I know a few legacy businesses that can’t ditch feature X or feature Y because that one big customers (sometimes the government) asked for it in their sales contract.And that just leads to feature bloat, inability to innovate, and overall makes you more of a services business than anything else. Which is okay if that’s what you want, but usually it isn’t what you thought you signed up for.

      1. Dave Pinsen

        Good points. With monthly subscription, you can use the 37 Signals / “Getting Real” approach to feature requests.

        1. PhilipSugar

          I think it is great to ignore feature requests, I do think you should keep track of them however. It is not hard with software. I’ll plug Joel’s company here.

          1. LE

            Joel says regarding feature requests:”How do you manage them? You don’t. Just read them and then throw them away.”This is one of the things I hate about internet advice. Advice like that might be great at scale where there is enough data to divine statistically what it is your customers want. Or great in certain circumstances. But not all. Or maybe for you (Philip) or from your perspective. But for a blank slate (young entrepreneur) taking that info in as gospel I don’t agree with at all. Everybody is looking for the play book as if they can follow that play book and leap to the top of the heap.What (as Jobs would say) customers don’t know what they want? Or what if one customer is more equal (and I don’t mean by amount of business) then the other animals? Maybe they are smarter in what they suggest? Maybe they know something others don’t? What if they are the guy that was featured last night on 60 minutes that knew Jobs well enough to be called by him at 3am? What if they don’t identify who they are when they make the suggestion so there is no inference of their importance?The fact that many people ask for something is certainly significant.But so is something that is only suggested by one person. If someone doesn’t want to put the effort (Joel) into digging deep into suggestions and to keep them then that’s fine. But it doesn’t mean that because only one person or 3 had an idea that it’s not a good idea or not worth pursuing or you should throw it away. So if you are throwing these away after you decide “nah don’t think that’s a good one” and it never comes up again you aren’t taking advantage of information that might play a purpose in a future context. Or at the very least the resources of a user who could be valuable as well as a chance to build loyalty.

          2. PhilipSugar

            You completely missed the point of my comment.I was promoting Joel Spolsky’s software that totally tracks every single one of these requests:…Prioritizes, keeps lists, stores forever.We track absolutely every single one. I am saying we don’t do every single one, because frankly some are stupid (one where somebody wanted to have a unique id as phone number, but allow multiple people to have the same phone number comes to mind), but we absolutely track every single one.

    3. ShanaC

      doesn’t the lack of new customers make you ignore innovation in the field because subscriptions aren’t churning (Until they do)

      1. PhilipSugar

        No, its not that you don’t want new customers, you want them like crazy, but you don’t want their requests to come before your user base. There is a saying in software that is as old as the hills: “I know I told you what I want, but now that I am using it I know what I need”Let’s take a real life example from one of my past companies, we don’t have enough of these. Everybody likes to state opinions but nobody cites real events. Union Carbide wanted to buy a one time license. The price was around $2mm. They wanted a custom splash screen that every time an employee logged in the were reminded that they worked at Union Carbide. BASF needed a TitleV report that the state of Texas in the ship channel mandated, but they had already paid us a couple of quarters ago.Union Carbide said they would not buy unless we would put in the feature. They were about equivalent in time to implement. I don’t care who you are: Microsoft, Oracle, IBM, whomever. You only have so many resources and that is a good thing, so we could do one or the other.I’ll pose a question in your style. Which do you think got done?As JLM so aptly says every generation thinks they invented sex.

    4. Techman

      Oh wow. I wonder if any UNIX based systems were affected by that bug, as I don’t really run Windows.

  14. Tom Labus

    Subscription needs cash flow+++ to build the necessary infrastructure for growth. The markets will turn on CRM if and when they perceive slower customer growth. They need a constant flow of new products to entice customers.A huge percent of their revenue is from new offerings, They have to run fast all the time to stay on that treadmill.

    1. PhilipSugar

      The issue is that their expenses keep growing to fit the size of the aquarium.If you look at somebody like 37signals, the revenues grow but the expenses don’t

      1. Tom Labus

        I’ll check them out, thanks

  15. andyswan

    I’ve been in the subscription game since the day I dropped out of law school…there’s no better (legal) game in town.Seems like a decent opportunity to spam…err…mention our new startup..Voomly is a platform for individuals to start their own subscription content business.Win.

  16. JimHirshfield

    I forget where I read it, but subscription revenue based businesses are valued using future revenue, whereas traditionally companies are valued based on historical results. For instance, why should we evaluate on last quarter’s or last year’s revenue when we know exactly how much money they’ll make in the coming months?#lookforward #dontlookback

  17. Charles Kenny

    I would agree with you Fred that subscriptions are a good business model, but would counter that healthy competition is a necessity for honest success.One need only look at the mobile phone, cable and telecom industries to see how a subscription model can be abused when there are few comparable alternatives available.

  18. andyswan

    I’ve been in the subscription game since I told BU law to funk off (or was it the other way around? LOL)There’s no better (legal to operate) game in town.It’s pretty much the foundation of our new startup, A platform for individuals to start their own subscription content business. All kinds of win.If ya don’t know, now ya know.

    1. Aaron Klein

      Very bullish on this super trend you are riding. I can’t believe how many applications I’ve thought of for my wife and friends to effortlessly start a Voomly business since you told me about it.It’s like Etsy for everyone who can’t knit.

      1. andyswan

        New tagline: “Etsy for people who can’t knit” Where do I send the check?Incidentally, that’s a pretty common data-point we got from beta: “I can think of a voomly for everyone I know, but not myself.” Ahhhh onboarding…

        1. awaldstein

          Lot’s more people have something of value to sell that have no idea how to find a market for it.Took a quick look at Voomly and that’s what jumped at me.Doing some paid for ‘office hours’ now locally and considering this as a product on your platform.

          1. andyswan

            We’d love to have you. Voice connect, webinar and text coming soon. We’ll grow with you.Just let me know how I can help….you’d do very well on voomly and would love your feedback as user.

          2. awaldstein

            On my calendar to get done in the next spurt of things outside essentials.

          3. andyswan

            #word. andy at andyswan dot com I will do whatever it takes to make this a win for you and would love to promote once live.

          4. Drew Meyers

            Have you thought about white labeling your solution so people in various niches could integrate it easily as part of a subscription? I’d be interesting in chatting if so

          5. Drew Meyers

            and you would be correct…sorry about the confusion

          6. awaldstein

            No biggee. You can connect directly to Andy.

          7. andyswan

            @drewmeyers:disqus shhhh you just nailed our business model! is the self-serve sandbox….but we are actively in talks with several niche communities to white-label our solution and turn their network into a marketplace of win πŸ™‚ andy at andyswan dot com

          8. Drew Meyers

            Just sent you a quick email

          9. Drew Meyers

            Awesome chatting with you today. See you in person next week.

        2. Donna Brewington White

          I can see it now, a whole new field developing around marketing OTHER people’s Voomly’s. Discovering new Voomly talent.

          1. andyswan


    2. Max Yoder

      Dig. Good luck, Andy.

    3. LE

      That’s a great idea. [1]Seriously I really like it. (We need to get you a new name though that’s the one thing I don’t like).Suggest tweaking this concept though:”You’ve built a reputation for yourself as an expert. You’ve got a following.”I think many people might underestimate the degree to which they can be thought of as an “expert”. (Esteem thing could scare away some people). Might be an idea to add some cute wording to take into account people who would be “voomlers” (domain available btw) but don’t think of themselves that way. People think “I’m really good at this but I’m not an expert”. And the people who think they are not experts are people who often know so much. Depending on one’s self esteem and upbringing you might see yourself as notches down from the value you can provide in the eyes of others. And that’s what counts not what reality is. (It’s really the inverse of what I’ve said before “it’s not what others think of you it’s what you think others think of you”)[1] If this doesn’t work it’s because you can’t execute.

      1. andyswan

        Awesome +10000

      2. Aaron Klein

        I think it’s definitely true that it’s easier to figure out what someone else should be doing on Voomly than yourself. I’m guessing that’s more of an opportunity than a problem…

        1. LE

          Suggest: “suggest an expert” would be a good start. Complete with all the trimmings (rewards, points, cash, happy ending) for the referree who would be someone who hasn’t heard of voomly perhaps.Who wouldn’t take an email where someone else suggested they were an expert at something? (As long as it doesn’t turn into the type of spam that linkedin is doing now of course..)

          1. Aaron Klein

            Exactly. The key is to emphasize quality over quantity.It’s almost a challenge:LE, Aaron thinks you are an expert at Guerilla Internet Marketing. Did you know there’s a way to turn your experience and knowledge into revenue? Voomly can help.Don’t disappoint, LE. Aaron could be your first customer.CC @andyswan:disqus

          2. andyswan

            I’d pay LE too

          3. LE

            Thanks. I feel like Sally Field. (I’ve always noticed a correlation between people who accept my help and people who are really smart.)I always thought a similar approach would be to put together an “a” team of people that could help startups or existing legacy businesses. As a package. People with different areas of expertise bundled together as a loose group. This group (which is fluid and it’s members depend on the particular project) would then contract and be blessed by the investors to help their startups and increase the chance of them succeeding. The group revenue would be negotiated and once done, team members could decide if they are in or not and the revenue could be whacked up ascap style. In some cases some team members would cajole others to join “make time for this one” etc.If you read HN you will see how people jump on the chance to offer advice to anyone who asks for free. (Actually that’s an idea for you to get some of those people involved even as individuals Aaron can write a blog post about it..) No reason those people can’t somehow get money for whatever they are doing for fun right now. Then the entrepreneur can sit back (like the president does) take it all in and know everything they need to know to make a decision.)

          4. andyswan

            Yes I like bundling quite a bit.

          5. andyswan

            Absolutely…. we’re just working on making this clean.Learning….

          6. Donna Brewington White

            That’s a good idea. Some game mechanics could come into play as well. Rewards for being a talent identifier. I’ve been thinking about something like this for an online recruiting product. That’s all I’ll say for now. πŸ˜‰

    4. FlavioGomes

      I’ve been enjoying your service on voomly thus far Andy. I’m not a startup…but many of the principles you deliver apply to mature businesses as well.

    5. Charlie Crystle

      Looks pretty good. Signed up. Have no idea what I’m offering yet but what the hell–$200/month, 100,000 subscribers…it’ll do. πŸ™‚ Looks really nice, Andy.

      1. andyswan

        We can work with those numbers. Thanks Charile. Happy to help get you rolling and promote you. I’m sure I’m not the only one on AVC that could tell you EXACTLY what your Voomly should be πŸ™‚

        1. Charlie Crystle

          Hmm… not sure how to take that :)Let’s bump it to $5k/month and see if there are any takers… πŸ™‚

          1. LE

            After personally helping lotsa people for free, and understanding the enjoyment that that brings, somehow I can’t see how charging $29 for 4 emails could possibly be economically beneficial to you! Note the example of the lawyer who will do work pro bono but if asked if he would accept $65 per hour says no. Free yes. Low pay, no.Most importantly time wise this will do the following. At 1 or 3 people this is essentially something that is fun and that you can stuff into your day (like commenting on AVC). It’s background radiation in the airplane. I think it’s great.But now let’s say you actually get 50 people paying you $29 per month. That’s only $1450 per month in total revenue. That’s not much. Even in Lancaster PA. And my guess is that you would be totally distracted from your daily activities with all the emails and back and forth. And would have an enormous opportunity cost to you.I really like the page that you put up and think there is something there. I would definitely refer people to you at that price it’s a no brainer.The problem that I have found (in helping startups and the type of people that aren’t corporations with open pocketbooks) is that they have no money at all to spend. So it’s not as if you could even charge them enough so the only thing you can do is to keep your client numbers low.(Perhaps @andyswan has to get this idea targeted towards corporate types with their nice juicy budgets budgets of OPM)(Side note: I have several people on retainer for many many years that I pay $250 per month and $500 per month and their only job is to answer my emails when I have a technical question that I need answered quickly that I don’t know the answer to. The first person started by doing this for free and at my suggestion about 10 years ago agreed to let me pay him for what he was able to provide. In return I get long detailed emails. I’m sure though that if he did this (he has a full time well paying job) for others he wouldn’t be able to give me value even at $250 he’d be to busy. N=1 it works well though.

          2. Charlie Crystle

            of course it’s too low–we’re just testing it out. I’m going to A/B test a few services, like $5000 for a Rent-an-Opinionated-CEO, or $20,000 for half-time enlightenment services πŸ™‚ Maybe 5% equity to get your startup its first 5 customers πŸ™‚ Or $9 for a newsletter.Fun to try πŸ™‚

          3. JimHirshfield

            Do I get lunch with that package?

          4. Charlie Crystle

            I’m thinking dinner.

          5. Donna Brewington White

            What, no freemiums?

        2. Donna Brewington White

          +1Dude’s a goldmine.

    6. takingpitches

      I am pro-tools to help people fully utilize and monetize their talents. This seems like a great idea in that vein!

    7. ShanaC

      I’m tempted. Most people can’t tie back in growth marketing and analytics. But could I get you to subscribe.

      1. andyswan

        Probably… Let’s find out

        1. ShanaC

          I’ll do it, and if you like it, you have to forward it to 5 people and say they must try. πŸ™‚

          1. andyswan

            consider it done.

  19. Pat Clark

    Subscriptions are great on the enterprise model but do you think we’ll see a more refined approach to how subscriptions are priced in the near term, away from ‘seats’ and more towards the unit of consumption for the business (# documents created, # jobs posted, # files stored)? Certainly, the business model should dictate the strategy (what, actually, are you offering?) but atomic-unit pricing (to steal Fred’s phrase) could lead to more usage and upgrades downstream as their usage expands when they enter at the basic pricing level.

  20. Ed Lea

    In a few cases, I think targeting impulse buys can be better than subscriptions. You mentioned Netflix, at Β£5.99 per month it’s a good price. But apple charge me Β£9.99 to watch a film and I’m very likely to buy more than 6 films per year – in which case apple come out on top.Equally, in app purchases are another indicator of how great impulse buys are. I’m tempted to switch from subscriptions to impulse buys in future products (even tweak the product towards that revenue model).

  21. Max Markowitz

    Great post on the subscription model! At OrderGroove we’ve been bringing this business model to retailers and brands like Sally Beauty, Grainger, Rogaine, L’Oreal, etc. – allowing customer to subscribe to frequently purchased products, while increasing brand loyalty – with fantastic customer receptivity. What do you subscribe to?

  22. Barry Garnet

    With a subscription model, it’s also important to consider when revenue is recognized for tax purposes. Generally revenue can be deferred. To over simplify if there are equal number of subscribers in each month then half the revenue is earned and half can be deferred. This would be a good topic for a Monday MBA post.

    1. PhilipSugar

      I can tell you having done this for more than a dozen years, there is no way to talk about “revenue recognized for tax purposes” generically. It in no way resembles reality. Are you cashed based or accrual based? Do you structure installation over the life of the contract or one time? Are you trying to maximize revenue now because you feel taxes are going to go up? Do you have enough expenses to offset the revenue? Are you taking a big Section128 deduction this year? Those are just some of the factors.

      1. LE

        True and many games played in this area. Back in the day would add “needing to get bank loan in first quarter and have to show good end of year income so …”Although some of this behavior is borderline it is done.Also (along the lines of both our examples) nothing preventing you from reporting one thing for one purpose and then restating at a later date.As an entrepreneur it’s important to know what is possible, what the accountant will allow, what the IRS will allow, and do what you have to do to get things done. Depending on what you are doing the accountant may or may not bless the behavior which is why you get advice from them and then decide the risk you are willing to take after hearing that advice or doing your own research.

  23. andyidsinga

    as an end user subscriptions tend to bum me out as they seem to often be more expensive next year …and bundling new ‘features’ to justify an increase.i’d like to have freemium with pre-paid, small features.

  24. JaredHoughton

    I have found myself victim to the “pay but don’t play” subscription game lately – specifically with Mail Chimp. I haven’t sent an email in over 3 months but I enjoy the security of knowing my mailing lists and templates are safely kept in my mail chimp dashboard.I wonder how many other paying but non-using customers mail chimp has?

    1. PhilipSugar

      But you are using it. To store your lists and templates.

      1. LE

        “But you are using it. To store your lists and templates.”Good point and Mailchimp marketing research should/would hopefully uncover this and make it part of the message. Really no different then what Volvo did with safety. You didn’t say “I’ve never been in an accident but so my Volvo is wasted”.

        1. Anne Libby

          It would be smart for them to look at usage, and send a note for offering lower tier price for people who are using it to store lists and templates…and reactivate at the higher price by sending an email.

          1. LE

            My twist would be to monitor cancellations.When somebody cancels find out the reason why. At that point offer to lower the price to keep the sale. This is what the cable companies do. You say you want to cancel they try to keep you with deals and lowering the price. Entire separate CSR queue to handle this. From a buyers perspective best thing to hear is “anything I can do to keep your business?” If the person replies to that (meaning there is no hard reason they want to stay you just have to give them a reason.(I could argue either side of the approach and in some cases what you are saying could be better or vice versa.)Had a case today where somebody wanted to cancel something that they pay $30 for. They said “everybody else is charging $10 for the same thing”. Our reply was “well here is why we charge more but no problem we can give you a price of $15 but not $10 that’s below cost”. They went with the pitch (was about two paragraphs) and are now paying the $15. And most importantly they are happy to stay and not switch. So there was no need to roll over and acquiesce to their low price demands but rather point out the value we provide and throw a bone of a discount (throw a bone is key people have their pride and need to feel validated and come away with something).

      2. JaredHoughton

        True, but how long until that perceived value runs out? As @MsPseudolus:disqus mentioned below, she has been paying netflix (not using) for 3 years. I have no intentions of doing that, but on the other hand im still paying. It has some interesting psychology behind & Id love to see some data on the mail chimp side (e.g. avg. time of non-use before canceling subscription)

    2. Kirsten Lambertsen

      I think I have you beat. I’ve had the same NetFlix DVD’s sitting on my dresser for about 3 years.

      1. ShanaC

        4 years for me. I should cancel

        1. Anne Libby

          You and @MsPseudolus:disqus don’t use Netflix for streaming?

          1. ShanaC

            I do, but not the disc service

    3. ShanaC

      i don’t know. It is a good question, because it goes back to churn question

  25. Not anyone

    I’m just about to launch with a subscription model (not going to be shameless here) and I believe Fred’s post captures quite a bit. Here are my concerns about subscriptions that I know I’m going to have to experience first hand. I was wondering if anyone else had gone through these.1. What if you have to change the price? For example, what if you establish a competitive price but find that you should increase the fee?2. Freemium is a wonderful model that’s gotten beat up a great deal; however, I think freemium can work against segments of customers (for example, single employee businesses v. businesses with more than one employee); however, I’m still concerned about giving it away at all. In the back of my mind, if someone is convinced of the value proposition, a fair deal benefits everyone.3. I am giving a two-month discount for subscribing annually. I would prefer giving away one month but following my personal purchasing habits, two months seems to be the right number. Any thoughts here?

    1. Max Yoder

      FWIW, here’s what I think.1. Increase your price for any new subscribers. Honor your old pricing model for the length of your current clients’ contracts; when you renew these clients, use the new pricing model.2. Try this: only offer your freemium model to folks who ask for it directly via email. That way, you don’t need to put anything about your product being free on your site, but you can still appease the people who like to play in the sandbox before they buy it.3. There isn’t one right answer here. All you can really do is put your discounted offers out there and see how they perform. Just be sure you definite what success looks like before you do any testingβ€”that way, the facts will do the talking, not your rationalizations.Good luck, John!

    2. LE

      “What if you have to change the price?”Change the price but inform the customer. Once you have people hooked you can send them a notice that the price will be increasing on “x” date. You hook people at a low price and then jack up the price. Anyone who complains simply give them an opportunity to continue at the old price. Of course you run the risk anytime you ask for a confirmation of any auto charge that someone will say “just cancel” even if the price is the same.”I am giving a two-month discount for subscribing annually. I would prefer”You can try different approaches. You can also say “first two months free don’t pay until the third month”. Different variations simply do testing (assumes enough scale to do this of course).

    3. Kirsten Lambertsen

      You might find this useful, if you haven’t seen it already…My opinion is, it’s all direct marketing. Test everything. A/B test one vs two months vs no months free.


      “(not going to be shameless here) “.ABS !!!

  26. Dave McLaughlin

    Anyone have thoughts on what the customer expectation is / should be if they’ve purchased the annual subscription and then decide they want to cancel halfway through the year? Wondering from the perspective of prudent cash management. Thanks.

  27. JLM

    .The traditional subscription model is very much like the high rise office building model.You have a classic “make or buy” decision point — can we afford to build our own building or are we resigned to renting space?Renting, like a software subscription, also allows a tenant to reassess its changing needs over fixed periods of time. They can expand or contract or change their location.The marketing of office space in soft markets in which landlords provide a period of “free rent” is just like the “freemium” concept in which basic services or initial periods of subscription are discounted to zero in order to establish a relationship over which the revenue starts flowing.Proving there is really nothing new under the sun..

  28. Max Yoder

    I’m building a subscription-based business right now. I want to share a mantra that my mentor passed down to me:Protect margin not by volume, but by value.

  29. JamesHRH

    Noted another comment on the 2006 post:Maybe something wil come of this: From an economic perspective, free flows like water. Money introduces a value decision and a transaction, both of which slow the flow and change its path.Guess who?

  30. bruno boutot

    Great wrap-up of the subscription model, Fred, thanks. But I can’t wait to read what you have to say about “marketplaces which to me are really peer to peer businesses”. I think that marketplaces are the most promising model, even more than the “big three” combined, especially when they are an extension of a community model.

  31. JT

    As always, great insights in here. Another fantastic reasons to love the subscription model is how the baseline subscription can act as marketing material for higher margin product upsales. We all know how Freemium works, but the fact is that some of the best professional services firms (banks, advertising agencies, talent agencies) use a variation of this model and publish weekly or monthly insights into market trends and share these thoughts with a close group of paying customers. (Think: equity services to hedge funds, future market trend forecasting for consumer product companies)While these baseline subscriptions are revenue events that support overhead for the professionals services firm, the underlying content of these materials is meant to demonstrate intelligence, market breadth and ability to comprehend and simplify any market trends.Therefore, the clients that purchase the base subscription are purchasing the call option on information flow. And if and when a topic interests the client enough, a more substantial revenue event occurs back to the professional services firm. The Yankelovich Report (now distributed by a WPP company) was famous for exceeding at this structure.

    1. ShanaC

      which wpp company?

  32. Alex Murphy

    Very thoughtful and complete.I like to think of Subscriptions as “Pay for Performance” for content and software. Like CPC, it aligns interests of all parties and maximizes the potential.

  33. matthughes

    Of the “big three”, a subscription must be most intimate connection between a company and its customers.If you are willing to pay a subscription, you really want the product.

  34. Ryan Connors

    Great piece Fred. I think churn rate is one of the most important leading indicators of a subscription-based company’s success. I believe that if people are empowered and informed they’ll perform better and churn less.I work at a business where our primary purpose is to enable customer satisfaction and reduce churn. It’s much more cost effective to keep customer happy than it is to replace ones who are lost.

    1. jusben1369

      For the churn lover academics this is a great post from someone right in the heart of it:

      1. Ryan Connors

        Justin, that’s a tremendous resource. It makes a lot of sense to tease out the variables that can distort the reporting for churn. I’ll be sure to share it around.

    2. ShanaC

      what are the top three largest sources of churn?

      1. Ryan Connors

        Shana, I could speculate and say things that come to mind such as confusion around utilizing the tools and subsequent poor results, poor delivery on promises and a better offer from competitors. But what’d I’d like to do though is dig a little deeper and craft a blog post around that topic. It’s a simple, valuable question to answer. Have you encountered any sources of churn that stand out from doing exit research?

        1. ShanaC

          yes: Poorly planned out navigation on the top. It needs to be consistent but also get people if they don’t show up one “the home screen”. Also if you serve a lot of products that have a lot of content marketing backing them, to make sure there are consistent links leading back to that specific product.

  35. Abdallah Al-Hakim

    There is an excellent series on SaaS put together by a Canadian VC Mark MacLeod and one of the points he always emphasized is about ‘churn’ – here is the link…. The trick is pricing a plan that will appeal to your target segment. On of my favourite examples is Dropbox

  36. jusben1369

    I think the best part about subscription models are that they keep you very close to your customer. That’s beneficial for multiple obvious reasons.I think the valuations argument and the cash flow/financials piece above are a little overly simplified. (3x revs in license revenues typically = 3x annual recurring levels so a 3x revs on a licensed software company = 9x on SaaS etc) In general I think the +’s and the -‘s between each model cancel each out.Two examples: Back in the “On Prem” days many a startup self funded by closing that first 7 figure deal. Getting cash up front was super powerful and afforded a range of options not possible via the recurring model. Secondly, undelivered elements can hang up revenue recognition for a very long time too – something not as applicable in a licensed software world. Point being, the model isn’t great all the time.Finally, particularly in B2B, I suspect there is a backlash coming as it relates to the ability to customize offerings. Perhaps the private/public cloud will solve that problem though.

  37. Abdallah Al-Hakim

    It is also cool to learn that freemium as term was invented in the blog comments. I followed Fred’s linked and here is the first mention”Free + Premium = Freemium?Posted by: Jarid | March 23, 2006 at 10:56 AM”

  38. Matt A. Myers

    Porn’s the industry that first took advantage of the subscription model on a large scale — and then really took advantage of it in a bad way with things like pre-checked cross-sales (hidden too at times), etc.. They got away with the deceptive tactics for at least a decade.


      Yes! Porn. The best websites to check system performance. They are always on the cutting edge of video etc. If you’re system doesn’t choke at a porn site it’s ready for some serious gaming.

      1. Matt A. Myers

        That’s the only reason why I visit porn sites..


          I don’t play video games anymore so I buy best bang for the buck systems now. But, you still need to worry some about video speed for playing movies.


    “The first several years of a subscription based business will typically require a fair bit of funding because the revenues come in over time instead of up front.”.That would be a higher barrier to entry. Anyone working on ways around that? Better yet, has anyone found a way around that? Of course being able to write software yourself helps.


      Oops… Disqus!!!.Since I’m usually years ahead of the curve. I would have to say my actions indicate a prime time to move in to subscription based services. While looking to make a change in 3-5 years based on my *current* move to on-demand.

  40. takingpitches

    Some subscriptions die very slowly. At the end of 2011, AOL still had 3.3 million subscribers paying for dial-up access, although only about 11% were actually using it (if I understand their 10-K right).

    1. Anne Libby

      There was a very sad story in the news some years back about some of the former RBOCs who had been charging some Grandmas and Grandpas a monthly rental fee for their phones. (Remember that?)Some had paid thousands in rental fees over the many years since deregulation allowed you to go to Target and buy a $10 (crappy) phone.

  41. Llodavinci1

    From a consumer perspective, let me say that I absolutely despise software as a subscription model. So much so, that I will not buy it under any conditions. I consider such a model to be parasitic at worst, vampiric at best. If I can not find a non-subscription alternative, I do without.

  42. David Clarke

    Of course, like sex the subscription model wasn’t invented by this generation. From the ’60s to ’80s, and beyond, mainframe software was essentially leased. So in a sense the perpetual license phase is kindof an anomaly related to the client-server interregnum. I didn’t live through the mainframe era and obviously the dynamics of what software was and who procured it were very different then, but I’d love to hear some insights from anyone who leased mainframe s/w. Can we learn from that experience?

    1. PhilipSugar

      It was a little different, because as Fred says “software is a commodity”. I’ll disagree a bit with this, but I will say the huge margins/bloat you could achieve back in those times was different. However, now if you are willing to be hyper efficient you can still achieve the margins just not at the scale.

  43. $27180517

    I found this post to be a bit ironic this afternoon Fred having just finished reading an article by David Tecee on business models for a Strategy class in which you were quoted. Fitting for MBA Mondays.

  44. David Petersen

    Churn. The worst thing in the history of subscription businesses. I consider a 10% annual churn rate a pipe dream for the stuff I deal with. 20% is what I aim for and I have yet to achieve that magical number. I have a feeling a site like has annual churn rates of 60-75%.I saw a genius give a talk at SXSW a few years ago about overcoming churn, and it was probably the most valuable 20 minutes of my life. Unfortunately he didn’t respond to my emails (unlike Fred who always humors me with at least 5-10 words, thanks Fred you are the best!) or Twitter requests for more information and now I can’t even remember his name or the company he worked for.I learned enough, though.To overcome churn, you need to look at your user behavior, hard. Pay very close attention the way people use your site, and identify the patterns of those who cancel. Figure out what they are not getting, and give it to them. Reach out to them in advance, if they fit the statistical profile of someone who will cancel. Tell them that you are aware they are not getting the most out of your service, and help them do it. There is always a pattern to people who cancel, and it can be found and fixed. Or at least slightly improved ha.

    1. fredwilson


  45. takingpitches

    At least in regular everyday domestic life, monthly subscription bills seems more associated with bills we don’t like to pay and businesses that are seen as lacking top-notch service — utilities such as phone, cable, gas, electricity etc.Just goes to show that where subscription businesses are associated with necessities and with services that may not have many competitive options, subscription price itself does not necessarily incent better service or innovation.

  46. takingpitches

    I’ve done a lot over the course of my career with churn in telecom/media industries — cable, sat tv, long distance, sat radio, wireless. industries where the customer base churns out every 4 years or so. Those battles become ugly, a lot of running to stand still with advertising, promotions, save offers, contract stuff.ugly, but in the context of this blog, perhaps a royal problem to endeavour to.

  47. William Mougayar

    How about voluntary subscription as a revenue model like what Andrew Sullivan just announced for the Daily Beast so he can publish ad-free content. It’s been also called “leaky meter” model. He uses to collect $19.95 & has raised already $440K 4 days after announcing it.Β I thought about that as a possible revenue model for AVC where whoever wants to, contributes $20 (or more if they wish). If 5,000 readers contributed $20 per year, that’s $100K. You could donate half of that to your charity of choice & use the other half to organize an annual gathering of AVC’ers in NYC for example. Β 

    1. fredwilson

      its a fine model. but i don’t want to do that here.

  48. George

    Subscription business model is the most attractive for goods whose marginal cost is close zero (non-subtractable goods). That’s why it works so well with software and content. However, for e-commerce startups selling private goods (shoes, razors, cosmetics) through subscription, it remains to be seen if subscription is really a scalable business model or just a marketing tactic/pricing scheme.

    1. fredwilson

      such a great comment. i am skeptical too.

  49. bernardlunn

    Inspired by your talk about Ankle Biters, I wrote a post about how enterprise software revenue models is moving from licensing to transactions. Monthly subs are better than perpetual but they are still licensing and I think the next turn of the wheel is towards transactional revenue models:http://bernardlunn.wordpres

    1. fredwilson

      yeah, our portfolio company workmarket offers both models. transactional turns into a feeder for subscription. it’s a really nice model.

  50. andy_mcf

    Appreciate comments relative to churn risk and associated ways to mitigate through deliberate (and sustained) focus on customer experience/satisfaction. The subscription model can reduce “contract time” asymptotically, which in turn reduces barriers to exit. All this means that companies (large, small, high tech, low tech, etc) must invest more to solidify the value proposition _each moment_, not just when the renewal event occurs.

  51. tylernol

    where do you get the 10% number from?

    1. fredwilson

      made it up

      1. tylernol

        ha! I was hoping that there was a spreadsheet of company data someplace that yielded this bit

  52. Malcolm CasSelle

    Fred, much to my amazement, nearly every kind of content monetizes significantly higher just by offering a paywall. MediaPass (mediapass dot com), the leading solution out there, shows results 5x to 20x over advertising on a CPM basis for everything from cat photos to recipies. I believe this due to a flight to quality on the web. App stores and NYTimes $91m in paywall revenue the last 12 months is the weatherbell.

  53. Modernist

    how do governments adopt a subscription model?

  54. Volnado

    Little late jumping in but Fred let me ask you a subscription question. If you could pay to have a subscription to your favorite artists that would provide you a stream of their new multimedia content right when it’s made and behind the scenes peek into process of it being made as well as concerts and tours and you get the content before everyone else (the free people) get it… would you pay? What about if it included an app store that was open to devs building on top the artists content? I asked you at Billboard Futuresound in SF who your favorite artists were and you said Action Bronson. So would you pay so when AB goes in the studio (which he probably does several times a week) he leaks the new tracks directly to you and other paying subscribers? And you get to see what that video guy following him around shoots on a regular basis. And you have a stready stream of AB themed apps and games coming out from his developer fans around the world? Or even better question… would your kids want a subscription?

    1. fredwilson

      Not sure I would pay for that

  55. John

    I’d love to get some real world examples of churn rates. How did you get to the 10% number?

  56. patrick

    the huge strides in web analytics make it so much easier to readily identify churn causation. i can’t imagine running our site 15 years ago (granted, i would have been in elementary school). the fact i can connect with a customer directly, ask why they left the service and a/b test solutions once a general theme is presents itself is a heaven-send.