MBA Mondays: Revenue Models - Advertising

My friend Darren Herman helped me think about this post. He sent me this deck along with some thoughts. This slide from Darren's deck is a good place to start this discussion:

Monetizing with ads
It is true that the vast majority of consumer web apps have been and continue to be monetized with advertising. On mobile that is less true, but becoming more true every day.

Mobile app revenue

There are all sorts of ways to generate advertising revenue online. Here are the entries under the advertising category in our revenue model hackpad:

Various ad models
This list is most certainly not exhaustive but it does cover the most common advertising approaches and you can see how many there are on the Internet. There has been a lot of innovation in this sector in the past 18 years since the first banner ads were created and sold.

The famous Luma Partners slide shows just how complex the online ad market has become over time.

Luma partners slide

And this market map is by no means exhaustive either. Online advertising is a big and complicated business.

I would break up advertising into two big buckets; ads that are sold and ads that are bought. The first is a relationship business, requires a direct salesforce or a salesforce that you can tap into, and will bring a higher revenue per impression in most cases. The latter is a data business, automated by machines and software, is a volume game and will bring a lower revenue per impression in most cases. Much of the online advertising market is moving inexorably toward the latter category, for good and bad.

The reaction to this move away from high value "brand" advertising to commoditized and programmatic advertising is native ad formats and advertising models. I have written about native advertising at AVC before and am a big fan of this approach. Examples of native advertising are promoted tweets on twitter and radar and spotlight on tumblr. In both examples, the ad unit is the same atomic unit of content as the users create in the service. I think we will see more and more of this as the value of the impression is driven lower and lower in the programmatic model.

When you think about an advertising revenue model, you need to think about one of two things; scale or niche. Scale means hundreds of millions of impressions a month or more. Niche means a valuable audience that advertisers will pay a premium for. But even if you are going for the niche approach, you will still need to have a lot of impressions. Here is why:

Advertising is sold many ways, including:

CPM: Cost per thousand impressions

CPE:  Cost per engagement

CPA:  Cost per acquisition

CPC:  Cost per click

Sponsorship:  Fixed cost for a fixed program

[thanks to Darren for that list. I took it directly from his email to me]

With the possible exception of Sponsorship, all of these methods will converge to the same number. For example if you sell a click for $1/click, and one out of every hundred page views turns into a click then you are selling a page view for $1/100 (1 cent), and that turns into a $10 CPM (10/1000).

CPMs have been in decline for years on the Internet. That's because the Internet keeps on creating more and more inventory. There is no scarcity. And as a result the supply/demand clearing price just keeps going lower and lower. Ten years ago, a $10 CPM was acheivable. Today, you will be lucky to get a $1 CPM. A $1 CPM means that 10 million impressions will generate $10,000. That's enough revenue to sustain a one or possibly two person business but not much more. You will need at least 100 million impressions and ideally more than 1bn impressions per month to have an interesting advertising supported business at scale. 1bn impressions is a lot of users using your service a lot.

Niche will work at slightly less scale. If you have a unique and valuable audience, you might be able to get a $5 to $10 CPM. So you will need 100 million impressions per month instead of 1bn impressions. That's still a lot of super valuable users engaging a lot.

If you are going with a scale model and you have a service that has that level of inventory to sell, then you have the choice of building a sales force inside your company or using a third party to sell your inventory. You don't need just one third party. You can use many of them. That's where the Luma slide (above) comes into effect. There is an entire industry built to take the inventory you give to a third party and put it through endless machines and algorithms before it is shown to an end user. I will not get into this in more detail here but Darren's slide deck, which I linked to above, has some good information on that. When you use a third party to sell your advertising you can give away anywhere from 50% of ad revenue to 20% of ad revenue. Most commonly it is somewhere in between.

If you are going with the niche or native approach, you will need your own sales force and you will need to hire a leader for that sales force (a VP Sales or Chief Revenue Officer) who can build and lead that team. The sales leader is a critical hire. There are people who do this for a living, who really understand how to put a team together and generate advertising revenue predictably and reliably, and they are highly compensated and are worth every penny. Do not skimp on this if you are building your own sales force. You may choose to build your own sales force if you are going with a scale model, but you don't need to do that right away.

In the interest of keeping this post a reasonable length, I will end here. I highly recommend diving into the comments where we will discuss and debate this post. I will conclude by saying that an advertising model is a viable revenue model option if you are building a service that has a lot of scale. But if you don't have millions of users a month, you should think hard before going in this direction. There is a limited amount of ad dollars out there (except CPA budgets which are in theory infinite) and more and more services trying to tap into them every day which is why advertising rates on the Internet seem to be in permanent and systemic decline.

#MBA Mondays

Comments (Archived):

  1. William Mougayar

    What do you think of native AND mobile together? Is that an emerging segment that’s still small now? Is it Foursquare’s approach mostly? 

    1. fredwilson

      I like it a lot of you have scale. Foursquare has roughly 10 million monthly actives and most of them are highly engaged. I think that is about the minimum amount of scale needed to make the model work

      1. awaldstein

        I agree.10 million engaged is a huge accomplishment.When you look at investments Fred, making the mental leap from a core behavioral connection between a handful of early users to that number is a giant leap of faith I would think.

        1. fredwilson

          Its about funding maybes

          1. takingpitches

            I loved the “maybes” from yesterday.At the end of the day, a great VC (it seems to me) makes those leaps of fate on the future. It’s about stomach and instinct.Separating out the possibility that your idea is just legitimately bad, you know you are talking to a wannabe VC when that person has no history of making those bets and either needs the validation from a real VC or wants you to completely de-risk everything for him or her.

      2. ShanaC

        wow. that grew

    2. PJSweeney

      Yelp is doing this very well. You search for “breakfast” it gives the list by rating, and many of the places have an offer (20% off, etc) all happening on a mobile app that is getting hit 100 million times per month. I believe merchants drive the offer so it is very scalable in that no sales person is required.

      1. fredwilson

        Yup. Yelp and Foursquare are two great examples of models that work well on mobile

        1. pointsnfigures

          I get no value from Foursquare. Check in for what? Yelp gives me value by having crowd sourced reviews.

          1. fredwilson

            I am the other way. yelp is useless to me. Different strokes for different folks. They have about the same number of monthly actives on mobile

          2. PJSweeney

            Fred – not really fair to compare Yelp and 4SQ. Yelp did $36m in Revenue last quarter with 84 million unique monthly visits. 4SQ did under a million in revenue last quarter with 8 million unique visitors. One is a market leader and the other is a follower. I know you are invested in 4SQ but it does’t validate the model yet at very small scale.

          3. fredwilson

            I would prefer to have a business of the future than a business of the past. Foursquare has 1/3 more MAUs on mobile than Yelp does, it is deeply social and monetized natively. You can have Yelp. I will take Foursquare.

          4. pointsnfigures

            My problem with Foursquare is I never get anything from checking in. Maybe a free coke occasionally. I never check in on Yelp, or rarely, but at least I get a list of reviews. Facebook check ins have more value for me if I want to meet people. The technology that 4sq has is interesting, but I am unsure of the value proposition for the user at this point. See huge value for a business wanting to target customers.

          5. fredwilson

            Have you tried explore on foursquare? It is way better for finding a place to go than Yelp

        2. William Mougayar

          According to this report, the global online mobile Ad market was $5 billion in 2011.

          1. fredwilson

            Yup. Not sure that’s accurate

          2. John Revay

            seems low

          3. AbbeyPost

            I believe that the IAB is working on a new report for 2012, and we’ll all be blown away by the figure for online mobile ads this year. It’s absolutely exploded by orders of magnitude (for better or for worse).

        3. Abdallah Al-Hakim

          native and mobile seem like a match made in heaven!

        4. Sriram Krishnan

          Do you see search traffic on mobile moving from Google to inside specific apps? For example, the ‘breakfast’ example above.It would be great to have a model where all apps could get search traffic (and not just for the words they stuff into their descriptions and keywords).

          1. fredwilson


  2. PJSweeney

    Fred – tough topic to cover in a short blog, but nice effort! It’s worth a few more posts. I have found that getting “Endorsed Impressions” by making the customer the voice of the brand is the Holy Grail of Marketing and the best use for any word-of-mouth network. At the 2012 Olympics Cadbury used the idea of making every guest at the Cadbury House venue a brand champion by automating their social sharing – in exchange for the guest getting great photos and memories without lifting a finger, Cadbury put a logo in the corner of each shot. (See the Cadbury video here:…This endorsed impression is much more valuable than a CPM impression, because it is user-driven content. I believe this subtle idea of product placement in social media will become more powerful than product placement in the movies. It’s product placement in the movie of your life – and that’s everyone’s favorite movie!Where would you put social media product placement?

    1. AbbeyPost

      That’s affiliate marketing, strictly speaking. We did a hugely successful lead gen campaign for VISA leading up to the Olympics, in just this way.

  3. Rich Edwards

    The thing that struck me as I read through you’re assumptions on CPM and impressions is that anything with an solely ad support revenue model that might scale cannot almost by definition, be a niche play. It must have a fairly broad appeal in order to get to a volume that makes sense with those revenue metrics (or be a significant outlier in the revenue rates it generates).

    1. fredwilson

      Unless you are building a small lifestyle business

  4. jason wright

    the Luma slide – makes me want to retreat from the modern world and live a simpler life.I start to hear Marvin Gaye in my head, singing songs from ‘What’s Going On’

    1. William Mougayar

      Yup. I had seen it in someone’s office & it gave me a headache looking at it.

      1. AbbeyPost

        My “day job” boss actually emailed it to the entire sales team with the expectation that we understand it.I wanted to run and hide.There’s a reason I’ve had very little to say about this post, even though I actually work in the digital marketing platform space.

    2. awaldstein

      Any day I hear Marvin Gaye in my head is a good one…

    3. fredwilson

      It sure does

  5. Kasi Viswanathan Agilandam

    What do you think of advertisement of internet companies on TV channels?Did that happen in U.S. decades back to create awareness of e-commerce? It happens a lot in India right now … lots of e-commerce (especially e-commerce and not tech) are spending lots of money (a lot) on TV ads …Do you see that as awareness creation or they suck in getting the ball rolling? I don’t like seeing TV ads of internet companies … it sort of feels like going backwards in time.

    1. JimHirshfield

      That’s not actually advertising as a revenue model. That’s _using_ advertising to market your start-up. Different altogether.

      1. Kasi Viswanathan Agilandam

        Yes. I understand it is not a revenue model.My question is did that happen few decades ago (or may be still happening in developed countries)?Is that a good sign of the product/service? (internet product/service) OR is it sign of failure of the product/service?

        1. JimHirshfield

          I think it depends on the product being advertised and the creative they’re using to do their ads. If it’s blanket advertising from start-ups, then that’s crazy times.

          1. Kasi Viswanathan Agilandam

            ‘crazy times’ it is 🙂

        2. takingpitches

          definitely used to happen in Web 1.0, with startups using a lot of their money for Super Bowl ads (generally the most expensive tv ads in the US for the year).it’s in most cases probably doesn’t say something good about management, with exceptions (the Apple Mac launched with a famous Super Bowl ads years ago…)

          1. Kasi Viswanathan Agilandam

            Yes. If it is about launch of a new product then a TV makes sense. Like Samsung making a funny video on some competitive product….OR … Micheal Jordan endorsing a new Tesla electric car….etc.,But advertising ‘what we do’ or ‘how it makes your life easier’ seems really bad … sounds to me like ‘too much money and no business’.

    2. fredwilson

      I am not a fan

  6. pointsnfigures

    “If you have a unique and valuable audience, you might be able to get a $5 to $10 CPM. “-seems to be the money line to me. Everyone thinks their audience is unique and valuable, but they can’t all be. This is where advertising is not relevant, and your app is. Without a sticky and useful app, your unique and valuable audience won’t show up and your company isn’t valuable to anyone.

    1. awaldstein

      Is advertising ever relevant? I think not.It you want millions of uniques you need to focus on the most core behavior. There is simply no other way to do this that I can see.Sponsorships, lead gen are value based models built less on behavior of the user and more on the intent of the user to buy. Less is more in this model at times.

      1. brian trautschold

        i’ve said the same thing before – but there are many times that an advertiser actually impresses me or moves me with a great commercial or cleverly sponsored tweet – and I say, “wow, that was a great ad”

        1. awaldstein

          I’m a marketer by dna and yup, agree.But it happens very infrequently. i may be a corner case but simply don’t think that advertising adds value to life. It’s some weird part of the food chain that just ‘is’.

          1. Kasi Viswanathan Agilandam

            Ad is for suckers and sucking product … not me said that on this blog… i am just repeating what someone said… guess who

          2. awaldstein

            No mystery there.I kinda agree but the discussion here is that ad and F1000 marketing spend are the lifeblood of our economy to some degree.As a business person though I don’t build companies based exclusively on them for some moral factor, but because they are almost always ‘go big or go home’ and really hard tactically to target.

    2. Dave Pinsen

      The flip side to having an audience that can generate high ad revenue is that such an audience is expensive to build via advertising.

      1. awaldstein

        hmmm….if you have have to pay to acquire the volume to make advertising work, does this work at all?I can’t imagine buying 10M subscribers. If you have to buy them I question whether you can keep them.

        1. Dave Pinsen

          The conclusion of Norby (the entrepreneur whose post Fred blogged about last week) was that it doesn’t (at least not on mobile).

    3. Brandon Burns

      I’m not sure the audience needs to be so “unique” or “valuable;” there just needs to be one. Third party ad sellers don’t care, and they buyers aren’t looking at anything but reach and basic demographic data.The $1 CPM vs $5 to $10 is really more about type of ad. Full sized banners get the higher numbers, small banners and text ads get the lower ones.Either way, like you said, it means that having a sticky app / experience / content is what matters. As long as its good and people show up, you can charge for their eyeballs — whoever they are.

  7. awaldstein

    You need to inherit a real slice of the worldwide online population for advertising to be a legitimate model. And in order to do that you need to own the most core behavioral activity of the market. Facebook and Twitter have done just that.I love the marketplace model cause it’s transactionally based, selling value and consumer/merchant connections.I love the niche model as well cause even in an advertising–sponsorship or % of transaction, it’s value based on the connection.This distinction is not moot. To get millions of monthly uniques you need the simplest behavior to be captured to get the broadest possible population. To get thousands that are valuable, one at a time, you need to be brokering value.I’ll take value any day as a consumer and as a marketer building a community.

    1. Aaron Klein

      You and I have talked about this and I totally agree. Businesses that help to arrange transactions are distinctly more valuable than businesses that try to interrupt user attention with ads.That’s why I think we will continue to see Facebook and Twitter head closer toward transaction businesses. E-commerce inside the tweet is going to be a very interesting business when they launch it.

      1. awaldstein

        Agree…Facebook is the interesting case. A transactionless platform for the most part.I hear about products, Kickstarter projects, events all the time on Facebook for a few verticals that I engage in there.Once I hear of them I leave to transact. That’s the Facebook and social web dilemma in a nutshell.

        1. takingpitches

          Great point — Facebook Credits, Facebook Offers, Facebook Gifts…a continuing series of attempts — none of which seem to have developed enough traction given expectations — to make it a transaction platform.

      2. Darren Herman

        Aaron, I like this comment a lot. Closest to the transaction is key.

        1. Aaron Klein

          Thanks. It’s been the power behind Google’s exploding revenues and that’s why you see Facebook and Twitter building out the tracking tools to be able to show advertisers how their ad products drive transactions as well.It’s all about the transaction… 😉

    2. Brandon Burns

      “You need to inherit a real slice of the worldwide online population for advertising to be a legitimate model.”Not necessarily true. Depends how you look at it. Not everyone needs to be Facebook. Independent bloggers make a good living on 50k uniques per day. Now that doesn’t make you VC fundable, but that does put and keep you in business.

      1. awaldstein

        50K uniques per day is rockstar blogger status. Certainly for them you are correct.The gap between them and Facebook and them and most startup businesses or the avg blogger are equally huge.My point is that targeting a media based business is just a bad idea unless you are willing to take a ‘go big or go home’.And 50K a month uniques avg is not a startup norm. If you have that number you have potential market proof and have options beside media if you desire.

        1. Brandon Burns

          i’ll agree with all of that. :o)

    3. fredwilson

      I agree Arnold. Did I make that clear enough in my post?

      1. awaldstein

        You did Fred, I’m just ‘redundantly excited’ it seems to see this brought out.Getting enough uniques to make this work is just a bear. Makes focusing on core value and transactions, while not easy, a tangible process to attack tactically.

    4. Matt A. Myers

      It will be interesting to see in 10-15 years from now what major platforms exist.

      1. awaldstein

        5 years out is unknowable really.

        1. Matt A. Myers

          Agree. 🙂

  8. JimHirshfield

    Do you think CPMs are also down due to better analytics? IOW, are advertisers seeing more data and realizing they don’t need to pay as much because they have a better idea which 50% of their dollars they have been wasting?

    1. fredwilson

      No. I think it is because they have so many ways to reach me now.

  9. takingpitches

    Thanks Fred. Excited about this series.I remember reading Chris Dixon’s purchase intent/non-purchase intent distinction a couple of years ago.If the customers you are delivering ads to are non-purchase intent, you are faced with the CPM decline issue as inventory on the Internet forever increases.If purchase intent users, where you are delivering ads to folks who are interested in purchasing something like Google, you could have a gold mine. Some of the paid advertising keywords go for hundreds of dollars a click.I can imagine foursquare having a model that is more purchase intent than non-purchase intent if it can deliver relevant offers through Explore (looking for brunch, a drink, a hardware store, etc.).

    1. fredwilson

      Yup. I think the purchase intent world is broadening and that Google is losing its monopoly on it as that happens

      1. Robert Dewey

        I’ve been looking into how to target based on purchase intent, at scale. It’s not that difficult of a problem (many ad startups popup daily claiming to do this using various data), but unless you start your own ad network (which takes a lot of relationships and a lot of time), you generally have to go through the bigger guys like DoubleClick with their RTB/Real-time bidding platform. Real-time bidding basically allows you to target the individual in real time vs. simply targeting the content they are viewing (or the generalized demographics of the site itself). Right now, this appears to be limited to banner ads… but I’d love to see other formats done this way, like text ads or video ads.Simply put, this allows advertisers to reach people who are actually interested in their product (or their competitors’ product), based on their search terms, sites they have visited, etc.. Imagine tweeting about your favorite hockey team, and then visiting CNN to see an advertisement from a vendor selling tickets to the game. That’s the future of advertising IMO.EDIT: Actually, what I’m talking about above is advertising based on your interests, or an interest graph which is accumulated as you interact on the web. Actual intent to purchase something might be harder, as not only do you have the interests, but now you need to decide if that person is in the market to buy.

        1. ShanaC

          you can do rtb by content. But, as I like to say about the web – it exists as a zen paradigm, it doesn’t exist until someone sees it.

      2. Techman

        I wonder if what you say will come true. Currently Google basically has a monopoly in the advertising market, and their biggest source of income is though advertising.

    2. Brandon Burns

      One key thing everyone forgets about the nature of advertising:Delivering a message has worth beyond immediate action or “clicks.”The digital advertising industry sold brands on the ability to track and drive behavior, something TV, print, outdoor, etc. couldn’t do as well. Now we place the value of an ad online almost solely on whether or not a user clicks through to make a purchase.But seeing, say, Coke’s branding everywhere, all the time, leaves an impression in your head that, if you act upon it, will usually result in purchase that is asynchronous with the time you actually looked at an ad. This is why TV, print, outdoor, etc. advertising are all still effective and profitable.Advertising isn’t always about immediate action, but its always about mental conditioning; a much more complex process that takes time and, usually, indirect methods to achieve success.

      1. takingpitches

        You have an instinct for this Brandon. What is your take as to why the online display ads don’t seem to leave an impression in your head the way TV, print, outdoor does?One of the most interesting thing in Meeker’s presentation from Saturday was (going by memory) something like that print took 7% of user’s time but 25% of ad dollars, suggesting either or a combination of both legacy utility for the advertiser or print ads give better impact

        1. kidmercury

 many billboards do you see in a day? how many banner ads do you see in a day?

          1. takingpitches

            yep, back to the Big Bang-like expanding pool of ads

        2. Brandon Burns

          kid said it first. i 100% agree.though i’m not sure its so much over exposure to a particular brand – because that’s the ideal situation for a brand — but over exposure to the supply of banner ads in general. there are just soooo many.but that’s why retargeting works so well — it guarantees super high exposure to a brand you’ve shown interest in. drop by a company’s website that uses this model (relayrides and taskrabbit come to mind), and you won’t be able to avoid their ads on the web for the next month. they’ll stick out, and you’ll remember them.btw, my instinct comes from 8 years in advertising, in both traditional and digital media agencies, working on a vast array of brands.

          1. takingpitches

            doing that experiment with taskrabbit!

        3. Chris Halligan

          Also, humans read on the web with an F shaped reading pattern (… which explains the rise of ads embedded in text below the fold, whereas ads in other media are integrated in a manner that makes them harder (but not impossible) to skip.In some media, the ads are a even a welcome and necessary part of the content (women’s magazines, as an example).

        4. Michael Brill

          Online display ads are competing with the content that we’re trying to focus on, so we’ve become much better at subconsciously recognizing and blocking out online ads because the payoff is low. TV ads and full/half-page print ads don’t have any immediate content competition, so they get a lot more of our attention. Same with offline ads that break up the background we’ve seen thousands of times before. I can still remember Accenture ads in airports from 15 years ago. I have no recollection of any ad I’ve seen online today.

          1. Jim Peterson

            Great point here. Eye studies show people don’t even look at many of he ads.

          2. ShanaC

            umm, other tv shows? those recording thingys that allow you to skip the commercials?

          3. Michael Brill

            Then you’re not being exposed to the ad. Point is that if you’re lying on the couch in front of 55″ of colorful and noisy hamburger-swimsuit-explosion close-ups, there’s a higher probability that you’re going to remember that than a banner ad off to the side of your cat video.

          4. ShanaC

            that’s an assumption. testing might be better for high creative display. which isn’t being done.

          5. Anuj Agarwal

            Absolutely. Those youtube ads before the video takes off, infact grabs my attention.

      2. ShanaC

        Yes. And also that there isn’t a way to do infinite ads. Every web model claims infinite ads are possible: and many startups fail on this model.*headthunk* why do people insist on saying that ads are only about immediate intent and that there will always be the space you need…I don’t know

        1. Brandon Burns

          I don’t know either. Its really the digital advertising industry that started and continues to perpetuate this false notion. They think it helps them sell against non-digital media placements, but really they’re only hurting themselves.

          1. ShanaC

            yes. they can and should demand more money, especially in light of both the creative possibilities and the data offerings.Vive la revolution (or if someone would let me say that..)

      3. awaldstein

        While certainly you are correct that exposure to messages over time impact our attention and word and image memory.And of course what you say is the the exact pitch that every agency and platform makes to every person who advertises on them and asks for some quotient of why I’m doing this.We all heard it a million times. Been going on since we lived in caves and scratched on the walls.Not saying it’s not true but I could simply say’ spoken like a true ad salesperson’ 😉

        1. Brandon Burns

          Ha! Really, after so much time in the industry, it rolls off the tongue.People ask me all the time why I don’t build an ad tech product to help make sense of this mess.The reason: I cringe every time someone tell me something to the effect of “spoken like a true ad salesperson.” :o)The thought of building the digital version of the ad sales guy makes me ill. Even thinking about it makes me want to take another shower to wash it all off!

          1. awaldstein


      4. Kirsten Lambertsen

        I was wondering about this point, reading all the comments. It’s been 4 years since I sold ads. I can’t help but think that agencies would be pushing to demonstrate value of ads beyond just clicks. Are they using overall lift now, to demonstrate effectiveness?I think everyone here should keep advertorial in mind, too. A lot of advertising is the just the entry fee for getting friendly editorial, access to user data, etc.

      5. Luke Chamberlin

        To your question of “value”, look at what we’re comparing. A brilliant or funny TV campaign can sustain a brand for years (“where’s the beef?”).On the internet what do we have? Banner ads? Pop up windows? Those are annoying and don’t respect the user and probably hurt brands as often as they help. On the internet, ads in their current forms have no brand value, so they can *only* be measured by click throughs.Can you imagine pitching a client on the brand value of a banner ad?(this is one reason ad agencies are flocking to social campaigns: it’s finally something on the internet that they can sell and use “brand value” to inflate their rates).

        1. ShanaC

          yes, if only because I am starting to see great ones for fashion brands. A number have started to embrace direct buys and make pretty, interesting, semi-native banner ads (j crew, gap, and calvin klein come to mind)

      6. Anuj Agarwal

        I think this works well for fortune 1000 companies. For rest of the advertisers, clicks matter the most.

        1. ShanaC

          no basis in the data for that. Lots of people don’t click that do respond

          1. thinkdisruptive

            Yes, but unless you can quantify the “lots”, the ROI math gets very fuzzy. Big brands can measure enough proxies with big data to be able to tell what’s working and what’s a good investment. The little guy can’t, and shouldn’t waste their money on anything they can’t immediately measure.

      7. JamesHRH

        Brand advertising reinforces a market position (on customers, or for very large companies, employees) which is why buying impressions matter.Its the classic balance issue that pops up in so many business areas: how much to spend positioning, how much to spend selling & how much to undermine the position (occasionally) with a price drop.

    3. Luke Chamberlin

      When I search brunch on Foursquare I see brunch specials (sometimes). I think this is where they are heading.The next step will be convincing the small business owners. Foursquare is in a unique place where they will struggle with “closing the loop” since their transaction begins online but ends in the physical world.

      1. takingpitches

        i don’t know where they are going specifically, but i assume they could have some sort of self-serve platform like google adwords, twitter, etc., no?

        1. Luke Chamberlin

          I think that’s where they want to be. Feels like if you owned a restaurant or coffee shop you’d want to spend money on Foursquare before buying keywords on Google. Especially if Foursquare completed the transaction loop with something like a check-in.Imagine if you could set a budget of $100 and each check-in cost you $5 and you only paid when someone checked in.

          1. takingpitches

            An adwords focused on local — would be very cool

          2. AbbeyPost

            Actually, check out Privy; that’s a big part of what they are doing. They’re a Boston based startup currently taking part in the fall season of 500 Startups. VERY sharp team.

          3. takingpitches

            Thanks Cynthia — When writing my comment, I was thinking within 4Square since it already has the Explore behavior going. Thanks for the heads-up re: Privy — will check out what they are doing.

    4. Mark Essel

      Student loans and debt consolidation are sadly high value keys.

      1. takingpitches

        Yes and very relatedly, online higher education. Four of the top eight spenders on search engine advertising if you see the list in my comment below.

        1. AbbeyPost

          What’s interesting is that credit card issuers don’t show up on your list, but they are absolutely massive spenders. It’s just that they pay for performance instead of eyeballs. Amex alone did something like $200M spend in their affiliate program this year.

          1. takingpitches

            That’s true. Various credit card offers start appearing heavily after the top ten. #11 is Citi Cards. #12 is compare, etc.

  10. Dave Pinsen

    Informative post. I wonder if Darren or anyone else here has first hand experience with Apple’s iAd for Developers (I’m actually curious how effective it is for ad buyers).

    1. fredwilson

      I think Flurry’s ad product for devs is way better. The data is fairly conclusive. But full disclosure that i am biased by our investment.

      1. Dave Pinsen

        I’ll check it out.

  11. ShanaC

    why are people not pulling inventory? I mean, wouldn’t cpm go up if you have less ads per page.

    1. takingpitches

      a collective action issue, right? The whole industry would need to do it, and the industry is too fragmented.

    2. fredwilson

      Nope. There is no scarcity online. You can’t crate it.

      1. ShanaC

        you know what is going on in the back of my mind already on the subject.

    3. Kirsten Lambertsen

      In my experience, if you are direct-selling your ad inventory, then you can command a higher cpm for an ad that is alone on the page. However, it’s usually part of a big package that you’ve sold to the advertiser, and usually it’s the advertiser requesting it, not the other way around.

  12. Tony Yi

    i would argue that you don’t need “millions” if you have a very targeted, high value audience (eg HNW/affluent luxury shopper) who is highly engaged. leveraging your behavioral data and advanced modeling tools, you can create “off-site” networked audiences in the millions that index to the same relative actions that the advertiser seeks. in fact, even many large publishers are leveraging this same audience extension dynamic.on the buy side, totally agree that CPMs are continuing to be compressed but trends like programmatic buying are delivering efficiencies that make the economic model very interesting. again, tons of data, geo/demo targeting and less human interaction create better margins for buyers even on lower CPMs.

    1. awaldstein

      Example please of a luxury segment that is focused enough and large enough and not based on transactions that can charge high enough ad rates to make it a business in itself.I believce iche based markets work best and more naturally when they are connected to a transaction in the value chain.

      1. Tony Yi

        several come to mind. robb report is a a classic. thrillist nyc is a more recent incarnation of highly localized, niche targeting. now they are in several cities in the usa focused on the young, affluent male.transaction and lower funnel targeting is great for DR campaigns and display is growing. but brand spend online could be argued as furthest from the transaction yet more is spent online on these types of campaigns. – deloitte report (there are several emarketer, forrester rpts on the same topic).granted a lot of this spend is sponsorship where synergy is factored in and/or bundled…but as upper tier inventory becomes more addressable/biddable i believe we will see much, much more brand dollars as biddable.exciting times for our industry.

        1. awaldstein

          Thanks…Didn’t know that Robb report was still around. A dinosaur that I knew a lifetime ago. I know thrillist but I’ve never considered these venues advertising, much more sponsorship as you say and email rentals.I’m not certain I see the same future as you do.The web is terrible at targeting specific demographics and anything specific to location. If I wanted to get to this demographics I certainly wouldn’t use advertising as the media. As a monster brand, sure. As a upstart cracking in, not feasible.

          1. Tony Yi

            sorry we’re not seeing eye to eye but this isn’t the future i’m describing. we are doing a lot of this today with great success but it is not widely known yet.agree – vanilla demo targeting is not good. more sophisticated targeting involves offline data, crm data, email lists (but not sold/rented as you cite above), etc. that are appended to online cookies (which are after scrubbed of PII before being targeted). that’s offline or email data match. this seed pool is then modeled for similar cookies that exhibit high correlations for like search, site, social and other data traits. no guessing on loose demo data. rather hard data on what people are doing/reading/consuming with high recency. you can then filter this data with demo/location data further.about the web targeting location – data companies are excellent at zip+4 location targeting in many areas using similar matching techniques and assigning location data to cookies.last but not least, robb report is still around. just an example. there are hundreds of niche publishers with value – luxury, moms, small business owners, etc. some are doing audience extension now but many are not…yet

          2. awaldstein

            Thanks for the details.I don’t know your background but I would be open to having a chat re: targeting groups within specific locations/[email protected] if interested in pursuing.

          3. Tony Yi

            also – branding ads are bought via rtb. it’s not all sponsorships. i didn’t mean to totally discount it like you did. it’s not only kellogg – unilever, coca cola, P&G, Ford, amex…the list goes on and on for programmatic buying of branding display advertising .

          4. ShanaC

            that hasn’t been mainstreamed yet, and the measurements for roi for programmatic brand buying haven’t gotten good. Nor has creative optimization….

          5. Tony Yi

            the top 100 advertisers are doing it or planning on testing it. the roas is there which is why it was invented. the creative optimization is no diff than ad nets with a/b and mvt.

          6. Tony Yi

            sorry you said roi for brand advertising. agreed but roi for branding has always been fuzzy in all mediums – panels, surveys, grp’s, etc.

          7. sprugman

            Not to hijack this discussion with an ad :), but my company just announced a new approach to measuring ROI that we’re pretty excited about. I’d be curious to hear your reactions.

          8. sprugman

            Hmm… I guess that marketing page doesn’t say anything meaningful about our methodology. I’m not sure how much of that we’re explaining, so it might be difficult to have a fruitful discussion about it. 🙁

  13. Mario Bucolo

    How can be a real average salary + success fee for a CRO in a startup? Thanks for any number

    1. Aaron Klein

      Success fee = equity.

    2. fredwilson

      250 base + 150 to 250 incentive comp at plan

      1. AbbeyPost

        250 base, up to 500 at plan? Fred, I need to work for one of your startups.

        1. fredwilson

          anyone who can manage a $50mm to $100mm per year revenue generation engine is worth all that and more

  14. kidmercury

    affiliate marketing is mentioned in the hackpad list embedded, but not discussed in the post. i believe aff marketing really breaks all these rules. you can build a business that has extraordinarily high CPM — hundreds of dollars/CPM — with the right affiliate model. it’s tricky, but entirely possible. a number of microentrepreneurs with a strong background in search marketing employ this type of business model.

    1. awaldstein

      Super interested in this.Examples or links please. My gut tells me you are right but I can’t put my finger on the proof.

      1. kidmercury

        i don’t want to cite any examples because search marketers are protective of their approach, but i can elaborate in a way that may illustrate and give folks ideas.first it should be noted that internet audiences in a way can be divided into two categories: those with commercial intent and those without. the vast majority of internet sessions are those where one does not have a commercial intent. we see this criticism of social media sites all the time; i.e. facebook and twitter are not designed for commercial experiences, and so the ad quality is not the best (so the story goes) — and so you get low CPMs and thus need a big audience.if, however, you are focused on reaching an audience with a high commercial intent, and if you are selling a product with high profit margins and are using affiliate marketing where you get a significant percentage of the sale, voila! you’ll end up with high CPM.put another way, google makes the bulk of its search advertising money on the highly lucrative, highly commercial search queries that occur on its search engine. queries that reflect a high commercial intent for a high margin product. if you build a business focused just on that realm, and you execute it properly (it is understandably a very competitive realm, but a different kind of competition as you are not necessarily competing for a large audience), you’ll end up with extremely high CPMs/CPCs/CPAs.

        1. awaldstein


        2. William Mougayar

          It seems that most of the ad dollars spent where there is non-commercial intent is a waste, isn’t it?

          1. awaldstein

            So every ad dollar on Facebook and Twitter is useless?

          2. kidmercury

            not necessarily…..i do believe branding ads could work well for certain types of businesses. also, certain types of branding ads can help a company position itself well for commercial queries when they do occur, and branding ads can also help create commercial demand in the first place. to use an offline example, cigarette companies have invested in creating an association between sex and a post-sex cigarette. so they used branding ads to create demand.

          3. awaldstein

            At it’s most prime level, if people have heard about you before they come face to face with your product it helps.Human nature at play and the core ‘why’ for branding.Golden ring of course is when that brand become the reach for the product you make or the higher price you’ll pay within a category.

          4. kidmercury

            this is very true. advertising can play an important role in setting expectations, and expectations are huge. i just read “predictably irrational” by dan ariely, which talks about behavioral economics, advertising, and expectations a good bit, so the subject is on my mind. i think it’s a great book for those interested in advertising/branding and one that is particularly relevant to our discussion today in fredland.

          5. awaldstein

            Just bought the book.

          6. takingpitches

            it’s carpet bombing (display ads) versus drone strikes (search ads or other ads to purchase intent audiences.

          7. Abdallah Al-Hakim

            but non-commercial intent could change to a commercial intent depending on the flow of the conversations. For instance, I could be reading a blog about advertising but then a discussion comes up about Tablets (discussion are very serendipitous in nature!!). This might trigger a commercial intent because I was planning on getting a tablet. This will require real-time automation for native advertising which I don’t know if it is technologically feasible yet??

          8. William Mougayar

            I think that case describes re-targeting services.

        3. takingpitches

          Kid, to add some context on your great point about high commercial intent search queries on high margin products, this is spyfu’s estimate of who the top 10 spenders on search engine advertising (and their daily budget) are. 7 of the top 10 are online education or insurance. That gives you an idea of the industries looking to spend that you can target.University of Phoenix (277K)Geico (163K)Statefarm (127K) (109K) (96K)Amazon (95K) (80K)Devry (70K)Google (66K)Allstate (66K)

          1. kidmercury

            yes. also, i remember back in the day of housing bubble, mortgage-related terms were huge search engine spenders. every SEO i knew had a mortgage refinancing blog lol

          2. Mark Essel

            Aha, found it. This puts online code schools in a new light, it’s big ad $$. Next up online insurance ed startups

    2. Brandon Burns

      really? i find affiliate marketing the hardest thing to crack. stumbleupon and pinterest promised breakthroughs in this area, yet i’ve seen none.what intel do you have?

      1. kidmercury

        commercial intent + high payouts + your ability to close sale (perhaps hardest part) = affiliate $$$i’m not hugely successful as an affiliate, and have had lots of ups and downs and stress along the way, but i feel like i’m getting there.

      2. AbbeyPost

        Pinterest specifically stripped affiliate links; they had a breakthrough, they just shut it down.

    3. fredwilson

      Affiliate fees are so small that I have never seen it produce enough revenue to justify a venture investment. Could be great for a lifestyle business

      1. kidmercury

        yes probably not a venture investment, save a collection of them. the best aff marketers i’m aware of are pulling in a few million per year in revenue (at 90% margins). but IMO a general rule of thumb is that high margins do not sustainably scale.

  15. brian trautschold

    advertising is so interesting because people are conditioned to actively block out (specifically bad advertising)… its quite incredible..whether X’ing out of the full ad on USAToday, ignoring the right hand ads on Facebook, or the myriad of ads on ESPN, – I’ve already subconsciously “ignored” many thousands (or millions?) of ad dollars in the 14 minutes I’ve been at my desk… However, a great ad – whether it be a great worded sponsored tweet or post, or more likely a great ad or movie trailer, I will actively seek out… If an ad is quality content, i.e. Google’s Dear Sophie or the recent Nike Fuel campaign, great ads do work.dear sophie ( fuel (

    1. awaldstein

      Are you the norm? Or I who almost never seeks out an ad or clicks on one?If you, as in 50% of the market that there’s sense in the model. If I, then you got to wonder.Anyone have any data here?

      1. brian trautschold

        Maybe a hybrid… I never click on ads – especially banner/ facebook… but when I find something great…I will indulge. Maybe just an appreciation for great marketing. As I mentioned, the companies in question are 2 of the best in that regard.

        1. awaldstein

          There are times when I have to stop looking at myself as the market and accept that I”m the corner case.Ads boggle me. They fund the world as we know it yet (most) disdain and (most) claim to never click on them.

          1. FlavioGomes

            I think that technology folks who’ve spent the vast majority of their careers online are the fringe “non-ad” clicking group. They witnessed its progress closely and know the “tricks” to get you to click.I firmly believe its the general casual web user who provide the bulk of ad revenue. I have a friend that manages TV shopping network channels and the revenue rates for such asinine products were unreal. I think the flowbee did north of 30 million in revenue. While not a direct ad related correlation…it should give some insight to the profile of an ad clicker.

          2. awaldstein

            I just don’t know. You may be right.I’ve sold products on the Home Shopping Network and done quite well.I don’t think though that it’s an analog between those that buy on TV from people on live video selling stuff and those that click on innocuous and annoying ads on the web. I get where you are coming from but i don’t see these as behaviorally analogous.

          3. Robert Dewey

            I don’t either. The people watching the shopping network are wanting to be marketed / pitched to, whereas most people surfing the web do not. Like my previous comment on here, I could see someone making a successful “search” engine that displayed only promoted content. There is a niche group of people that would use that — the ones that actively want to be marketed to. Those are the people usually looking for a product. Almost sounds like a joke.

          4. Ivan Mojsilovic

            Like my uncle once said after long debate of how useless the adds are: oh crap, they never actually targeted me, so why im bothered at all 🙂

  16. William Mougayar

    But let’s look at where the puck is going versus where it has been. Really, when was the last time you clicked on an online ad?When I look at that list from Hackpad, I’m interested in discussing where the growth is going to be, what are the emerging models, and which are the dying ones?For e.g. online video Ads segment apparently has the highest growth at 55%, followed by sponsorships at 27%. Search is still the big Kahuna at 49% market share, and we could spend a whole post just diving into “advertising models for search”. And that same report says the whole online Ad market is growing at 23% in the US. http://www.marketingcharts….So, where is the innovation taking place in Online Advertising: Native, Mobile, Sponsored…?

    1. Brandon Burns

      “Really, when was the last time you clicked on an online ad?”People love to say this. But if no one clicked, there wouldn’t be a multi-trillion dollar industry build around ads.

      1. Carl Rahn Griffith

        But who does, lol? That’s the ultimate irony in all of this.In my many (many) online years I can never remember doing so – and when I discuss it with others this is not uncommon.Maybe, like the fiat system, it’s all one great big con (OK, let’s be kind and call it an ‘illusion’) – or ‘Hollywood Accounting’, as it were…

        1. Brandon Burns

          I rarely click on ads, but it happens from time to time.Considering that a .2% click through rate is good these days, if everyone clicked on 1 out of every 500 banner ads they saw, the ad owner would get their money’s worth.I find the ads placed by retargeting engines work best. Recent example: I stopped by, didn’t rent a car, but now I see their ads everywhere I go on the web (they’re clearly doing retargeting). Since then I’ve visited their site several times and have used it twice.And I’d say that’s possibly a one out of 500 experience for me. Rare, but it happens… and ad revenue / pricing is tied to that frequency. So, en masse, it works for a big advertiser, which are the only ones that can afford banner ads anyway.So I guess the system works?

          1. Carl Rahn Griffith

            I believe ‘the system’ – like many of our now de facto systems – are built on nothing more than a house of cards, I’m afraid to say.Thing is, in this context, even if a nominal click-through rate is an accepted metric, how many of those actually manifest as real transactions?Trouble is, advertising has become an industry itself and almost too important to fail – or even be questioned.Just like the banks…

          2. Brandon Burns

            I spent my whole career in that industry. It makes my head spin.At the end of the day, brands need to pay for the ability to get the word out about their products. They want eyeballs, and someone to find the right eyeballs to put their message in front of.There are almost infinite ways to skin that cat, and all those ways will keep the debate on how to do it and measure it going, the market too fractured and chaotic to reign in, and the whole system trucking along… and making money.Like you said, too big to fail. Or, in this case, maybe too unwieldy to manage.

        2. Ivan Mojsilovic

          I’m one of those “clickers”. I love to check the adds on search and even in my paid google apps gmail. So many times I have found super interesting stuff that i would never found otherwise.

          1. Robert Dewey

            While it sounds like a joke, I actually think someone could make a “search engine” that only shows sponsored results (filtered of course). People would use it, I’d probably use it when looking for products.

          2. kidmercury

            google has been increasing the advertising presence on its search results page. some serps are just google ads and links to google pages (i.e. google+, maps, images, etc)

          3. Ivan Mojsilovic

            I think that’s a good idea. I’ve clicked on sponsored results so many times.

        3. Robert Dewey

          I clicked on a re-targeted banner ad for an online shopping cart platform, a few days after visiting their page. I was already planning on going back to their site, and the ad just gave me that extra push.As far as search ads go, I click those all the time. Sometimes they are actually more relevant than Google’s results, especially if I’m looking up a product. For example, if I want to find a company that sells cheap servers, the search ads usually produce the best results.It’s the same as comparison shopping. Those are all just one big advertisement with referral links, but they are pretty useful and get a lot of traffic.

      2. kidmercury

        people click on ads on search results pages. outside of that, there is A LOT of click fraud, click bots, and accidental clicks (many of which are designed to be conducive to accidental clicking, and thus could be regarded as a form of click fraud). impression-based ads and all the other forms of ads noted in fred’s original post are also major drivers of the big online ad business, so it doesn’t all boil down to clicks.

        1. Brandon Burns

          yep, fraud is a big issue.but online fraud is rampant everywhere — ads are just one piece. twitter bots, fake facebook likes, etc.not sure what that means, but the system seems to keep running anyway.

    2. fredwilson

      Native mobile

  17. Tom Labus

    Revenue via advertising has gotten so convoluted. Someone is going to find a better, streamlined route.

  18. Brandon Burns

    I think advertising, specifically banner advertising, can be an entrepreneurs best friend.On the low traffic end, with 20k unique visitors per day:$6.4 CPM** x 3 ads per page x 2.5 page views per visitor x 20k unique visitors per day x 22 weekdays per month x 70% ad space sold = $14,784 per monthYou can bootstrap a team of 3 on that, and scale from there.The key is getting 20k uniques per day; a modest number, but still an uphill battle and takes a considerable amount of knowhow. But then again, which do you have more faith in your ability to get — the 20k uniques, or investor funding?__________________________**$6.4 CPM = $8 (average going through Federated Media) minus 20% commission

    1. awaldstein

      600K uniques a month?If you can target and get that you prob can get the funding.I think its a more fundamental choice not if you can get that number but what the core value of your product/community/niche is.If you can build a plan to get that repeat attention, I would think there’s a better model than media there.

    2. kidmercury

      as a UX guy, don’t you think banner ads are a UX disaster? i understand revenue trumps all, doubly so for the bootstrapper, but i think if you cripple UX you’re often crippling long-term growth/revenue as well.

      1. Brandon Burns

        I think if you’re going to do ads, you have to design with ads in mind.One key is to have an experience that feels more content discovery than utility. People are used to banner ads on blogs, news sites, video channels, etc. — so if you make your experience feel like one of those, the ads will feel natural instead of distracting. But if you don’t have a platform that can fit into one of those buckets, it probably won’t feel right… but you also probably won’t have the high page-view activity to make it worth it anyway.Behance did an excellent job of making banner ads feel right on their platform, and having that be there main source of revenue that supported them for 7 years before they took first funding from USV.

        1. takingpitches

          Wow — re: behance. I know Fred once blogged about them bootstrapping. Was ads their main source of revenue during the bootstrap phase? It’s amazing to think that they used ads for 7 years to bootstrap a platform that is ultimately transactional (matching talent with opportunity).

          1. Brandon Burns

            Well, it wasn’t always transactional… you have to build a crop of talent before you can pimp them out.Behance started as a place for creatives to post work and look at what others were doing. The experience was and still very “content discovery” feeling, and that comes with high page views per user and makes advertising a feasible source of solid revenue.They also made and sold physical products (their notebooks are amazing for disorganized creatives like me), books, etc. Scott Belsky is the definition of a hustler.

          2. takingpitches

            very cool. a lot of stored value on the network kept people coming back and back even apart someone with opportunity looking for talent to execute that opportunity.(i just get served an ebags ad)

          3. Matt A. Myers

            Advertising rates were much higher back then..

    3. fredwilson

      I use federated. Their CPMs have collapsed this year. Try $1.50

      1. Brandon Burns

        That’s worse than a collapse. That’s armageddon.Why did this happen at Federated?

        1. fredwilson

          More supply than demand

      2. Brandon Burns

        Looks like maybe AVC rates being down is an outlier case?https://advertisers.federat…$10 CPM looks like a median price point for properties served by Federated.

        1. fredwilson

          I think they got out of that business last month

      3. Techman

        That sucks. This is the downfall of advertising as something practical by itself: ad rates keep falling.Do you have an idea on how to fix this Fred?

        1. fredwilson


      4. ShanaC

        (used to use). The cpms for this site shouldn’t be collapsing…..

  19. Vineeth Kariappa

    the number of impressions u get /month does not matter. The actual stat is the number of adverts/ month that can be shown for a site’s traffic. Thought u never liked ad supported models.

  20. Carl Rahn Griffith

    For all our supposed demographic targeting, market intelligence and analytics, it still boils down to this…”Advertising is the rattling of a stick inside a swill bucket.” ~ George Orwell.

  21. Iggy Fanlo

    Having been President of as well as CEO of AdBrite, I can tell you that most publishers don’t have a revenue problem. They “think” that they have a revenue problem because their CPMs are always falling. The simple mathematical truth is that internet gives you measurement. Measurement means that you will see what sales your advertising is driving. You will quickly move to CPA or CPC at the most. You need intent for that, search is awesome. Other than that, finding intent (which is what the entire LUMA slide and the ecosystem is trying to do), is like panning for gold; fun and interesting but very little there.If and when brand dollars come to the internet, which I do not think will happen in a material way outside of video, since branding is about emotion and you just don’t get much in digital, we shouldn’t expect CPMs to rise.What we should and need to do is get our costs in line with current to falling revenue per impression. NO more Cassandra.The great news is that total revenue is growing rapidly. But impressions are just growing faster. It just means that table stakes are now a smaller slice but of an enormous pie. You need that billion impression level. Tat use D&O be huge; it just isn’t any more.

    1. kidmercury

      the bull market in video advertising is just getting started. it will be epic.

      1. Robert Dewey

        They need to fix it, and fast. Seeing the same ad every commercial break, or at the beginning of every YouTube video is annoying. What’s even more annoying is when it’s for some pop star’s music that I’m not even interested in, or a product that I’m not even in the right demographic for. I can tolerate banner ads being that way, because I ignore them. But video ads. Loud and obnoxious, and cannot be ignored.

        1. Mr. Majani

          YouTube basically needs to increase their inventory. They know that their ads are few and repetitive, that’s why they allow you to skip the ad, but I have a strong feeling that once the inventory gets up to a point where there is considerable variation on the ads, then they’ll probably remove the skip option.

    2. Robert Dewey

      There’s a lot going on in the ad market when it comes to intent/interest based advertising. Re-targeting is a step in the right direction, and makes sense… but it seems limited thus far. If someone has been visiting online retailers and searching for key items, it makes sense to target those people specifically. Right now that capability is limited to ad networks with real-time bidding on display ads, as far as I can tell. Would be interesting to be able to pull that data into other places as well, like video ads. If I’ve been looking up high end computers, show me a blade server ad from Dell… not an ad for feminine hygiene products.

    3. Dan McReavy

      I agree that marketing is about connecting, emotionally. I disagree that there isn’t much emotion in digital. When you affirmed online do you feel warm fuzzies? When someone disagrees with you, how do you feel?Context is everything in the physical and digital worlds. adSense and adCenter are price discovery mechanisms with a scalable revenue model sitting between advertisers and producers…and end consumers.Judging by the corporate failure rate of ideas as quantified by start ups, angel, VC and then mature self-sustaining businesses, the price discovery of <$1/CPM might be rich! lolUnfortunately, the garbage companies compete for scarce advertising opportunities and drive up the advertising expense of the good and viable startups. In the end, anything that produces something of value such that income exceeds expenses, will survive.

    4. ShanaC

      if cpm is gone, then they aren’t panning for gold, they’re panning for a way to drive emotions. And I’m starting to see that happen, oddly enough.

  22. Abdallah Al-Hakim

    Where does one place a revenue model that is charges you a premium for getting an Ad-free version of their service. For example, Sparrow mail (before it was bought by Google)

    1. fredwilson


  23. jonathanmendez

    It’s not about choosing a model. As you said all the models break back down into the same performance measurements. Online advertising is about creating marketplaces – and in that regard is not all that different from building anything else that has value online.

  24. Dorian Benkoil

    Fred, I actually see the market splitting into two: the commoditized inventory, for which prices are continually pressured downward as you note; and the “premium” spots — those with some unique value, such as content that can’t be had elsewhere. Consider, for example, a 20- 30- or 60-minute video that people will sit and watch even if interrupted by (a reasonable number of relevant) ads, so compelling is the material. True, the advertiser can reach those people other ways, but within that programming only via ads placed in that specific video stream. I can envision an analogy for particular Twitterers, Tumblrers (word?), etc. There’s only one Shaq, one Steve Martin, one Fred Wilson, etc.

    1. Matt A. Myers

      Differentiation of quality of content, yes; Doesn’t mean it won’t be potentially annoying and intrusive ads.

  25. JLM

    .Typical great discussion of an arcane subject.Where do I send the tuition check?Well played!.

    1. Carl Rahn Griffith

      The best advertising in the world!Word of Mouth.

      1. Matt A. Myers

        Do we need to differentiate between authentic word of mouth and incentivized?

        1. Carl Rahn Griffith


    2. Matt A. Myers

      I’m collecting tuition checks for Fred

      1. Big Red Car

        .Mattie, why do I think you are stretching the truth a bit? Yoga?.

        1. Matt A. Myers

          Humour. :-)Surely your name isn’t Big Red Car? 😉

          1. Big Red Car

            Ooops wrong reply.

          2. Matt A. Myers

            Don’t follow..

          3. Matt A. Myers

            Oh, I figured it out now …

        2. Big Red Car

          Actually it is JLM but I have started a new blog and I can’t figure out my Disqus identity. [Don’t tell anyone.]

          1. Techman

            Did you fix the Disqus problem?

          2. fredwilson

            Yup. I just commented there

          3. fredwilson


          4. takingpitches

            Diana Ross!! I am going for the tee-shirt!

          5. ShanaC

            i’m so confused – why did you change your identity?

    3. fredwilson

      We pay our tuition with participation here. I think I owe you money my friend.

  26. Matt A. Myers

    Ads are intrusive, relative to there being no ads; Decide how you’re going to monetize long-term from that point.The vast amount of ad inventory that exists today, that is diluting the value of ad space, may start shrinking. It can’t be known if as platforms solidify, as needs are filled in by them, that users will need to visit less alternate places, to access information or do what they want – and therefore the ad inventory space that does exist will become more valuable. The neverending barrage of creativity us humans pump out, assuming aligned with ad space, will be a constant fluctuation of available impressions available; Much of this isn’t monetizable due to various factors. And not to forget more people are using AdBlock.This all just makes me realize and see costs of producing the next unit or good is reaching $0 very rapidly – good for people. The only thing delaying it are old dinosaurs that will fight but will slowly break apart, and government and laws that don’t reflect what natural behaviour dictates to happen. The sooner all of these come into alignment, the sooner we can all live as fulfilled lives as we want – with as little sickness possible – and where all competitiveness comes from people trying to out-create (creatives) others, improve themselves, attract good-suited partners – friends, spouses, sexual, business, teachers, mentors, other – through their creativity, their interests and passions, their knowledge and wisdom, their kindness, and whatever else the human spirit has to offer./Start tangent/ There’s enough pressure alone with trying to improve yourself, finding and then attracting good people that we don’t need the added pressure of survival. There’s a lot of work still to bring the world up to par, to manage the natural fluctuations that exist with nature, and then to help each individual discover possible paths and facilitate all of this learning – to ultimately guide people to become their version of best, no matter what life has thrown at them, no matter what circumstances they are born into.I’m in this for the long-haul, at least that’s my vision so far. And I have my backup plans now that I’d be happy with – though I’d rather have a bigger impact on the world than a smaller one. There are a few billion dollar businesses I’d like to try to build, and I feel I am on my way to doing so – assuming I don’t fuck up somehow – though trying my best so far has kept me moving forward, and so I don’t see why this won’t continue. /End tangent/

    1. Carl Rahn Griffith

      Interesting stuff, Matt. Love it.The world is a very f*cked-up place, nowadays, and there are many with a vested interest to ensure things don’t change too much. If the mess accelerates (as it will), so be it. They can live with an increasingly polarised society – as long as they can have the cream.

      1. Matt A. Myers

        I’m hopeful because of what the internet allows – access to information, curation by experts, and then rapid consumption by the masses. And then you need leaders who understand the information people have learned, and lead them to the change that they believe is needed and that they believe will help vs. the status quo. Everything I’m focusing on is towards the greater good – the wellness of others, of everyone. It feels good now that things are progressing. Even getting free help from the local university’s business law clinic to help setup a non-for-profit charity, so I can start the process of fundraising for testing systems that I want to see in place – to get the proof of concept for them (societal systems, not businesses-ones) – and with the not-for-profit then I can apply for grants, etc.. The parts for the my business plans are coming together too, aligning, knowing who initial core group of team will be to move it all forward. I let myself feel little glimpses of excitement, while refocusing on the work that still needs to be done.

    2. takingpitches

      Last sentence of third paragraph to end = EPIC mission statement

  27. Albert R

    So you consider Amazon/iTunes affiliate sales in the category of an advertising model (as opposed to commerce or transaction processing)?Different, unrelated question: do you think that having a freemium (game-type, though not necessarily for an actual game) model is worth the effort when getting a site off the ground? If you had to choose one to start off with for a small site, which would you choose?

    1. fredwilson

      At a minimum you need a free model

    2. AbbeyPost

      Affiliate IS advertising, or more accurately, marketing. When you look at it from the brand perspective, it’s simply CPA. They pay their affiliates to create a need, place the content, and facilitate a transaction; all the brand themselves do is fulfill.

  28. Chris Sparks

    How many start-ups out there which are not pure content-plays are capable of producing 1B/page views/month (or 100MM to a engaged niche audience)? I am racking my brain and can still only count them on less than two hands (compared to the dozens coming to mind which started ad-supported until they were forced to pivot to a different model).Not to mention: 1) the trend of declining CPM’s will likely continue into the future as inventory/efficiency increase – leading to a higher bar of user traffic needing to be sustained and more intense competition for those eyeballs and 2) that you are constantly at the mercy of the major players (Google, Yahoo, Microsoft, AOL, etc) which are becoming increasingly vertically integrated across the chasm separating advertiser–>publisher.

  29. Sean Saulsbury

    Another important aspect to note when choosing the advertising model route is to look at the type of content you are producing. CPMs vary widely from industry to industry. If you’re in the movie/entertainment space (as I have been), CPMs are relatively low ($1-3/cpm if you’re lucky). If you’re in the medical, high end camera, cars, or tech space, they can be much higher.Looking at the cost of keywords for your industry is a first step. If the CPCs are high (such as in the areas I mentioned), your CPMs can be very high as well. If they’re low, you might look at a different revenue model or be confident you can garner a massive audience.At Box Office Mojo we were pushing 50 million pages/month at our peak before we sold to Amazon/IMDb, but we were around four people total–and supplemented our business model with data licensing and paid content.

  30. Max Yoder

    “The sales leader is a critical hire.”Would love to see a guest post from Rick Webb that dives deeper into advertising in general and Tumblr’s native advertising in particular.

      1. Max Yoder

        Ah, news to me. Good to know. Thanks, Fred.

  31. gomobile

    Mobile has the means to be the game changer like no other internet adv model before. The advent of Html5 rich media that crosses the line from weak minded banner ads to real inspired marketing messaging. Comparing what has transpired before is short sighted. As with all things, demanded price is based on value. Celtra’s research shows 10%+ engagement rates for mobile ads. Since forever Internet ads are bland, unappealing and insulting to the brands and ideas they are trying to push. The industry has blamed everything else, from ad serving to banner blindness to time of day. No matter what, are you going to click on something that weak? Right less than 99% do, but so far HTML 5 is driving 4x to 10x more. Why’s that?

  32. ericfranchi

    Great points regarding oversupply of standard units. However there are some really interesting things going on in the space to address it: a movement towards viewable impressions as currency (many of the ads that are available across exchanges, etc are never within view) and IAB Rising Stars (larger more engaging units).

  33. Joe Tareen

    Any plans for writing about Cost Per Transaction model? How do you see business models that take a cut per transaction matching up in the future against other models discussed above?

  34. gregor

    I’m young and naive, but I think that advertising should be an artform. The internet turned advertising into a science problem that was (and is) solved through optimization and targeting. I honestly don’t understand why display advertising is still a 15bn industry and is expected to grow so much when it is so ineffective.I think that native, niche ‘advertising’ needs to become more automated and be an available solution for low traffic sites with high value audiences. We can look at twitter and tumblr, but also look at Buzzfeed, Thrillist, or Pandodaily (to name a few, please reply with more) to show how vertically integrated business models can be so effective at providing value both to the audience and the buyers/sponsors. Buzzfeed handles the mechanics of virality, so brands can actually create art and copy. Publishers should be experimenting more with new revenue products and pull the CPM IV out of their arm. If the one-way broadcast model of communication is dead, and Twitter is bringing us into the new age of the Agora, anyone with a voice that can be heard has an opportunity that has not yet been realized.There’s a lot of room for innovators to leverage the current ecosystem when building crazy products. With countless ad tech companies out there offering layers upon layers upon layers, newcomers can recognize the eventual commoditization of current technologies and focus on new ways that inventory buyers can interact with an audience by tapping the leaders.Anyone else see opportunities?

  35. thinkdisruptive

    Far, far too many applications are ad-supported. I don’t mind well-targeted ads, although I find the ones that follow you around based on previous clicks and searches creepy (and regularly delete all my cookie crumbs as a result). As a consumer, there are many apps I’d rather just pay for, and be left alone. As an ad-buyer, unless you have a pretty good idea of intention (location-based, search-based), it’s hard to justify the spending, especially when you know that if you don’t specify placement tightly, you’re going to offend a lot of people.Ad-supported business models are the dumb-ass default of too many entrepreneurs, and the reason many fail. If they thought more about the reasons why people choose their app (or choose not to use it — ads are often a contrary reason), and the value it provides, they’d realize that ads are counter-productive for monetization and long-term value, even if they juice numbers in the short term.One thing to consider if you’re planning to take the lazy, ad-supported monetization path is to remember that someone who comes along with a true freemium app that is better focused on the job to be done and perceived value will usually disrupt you. It’s only when advertising is part of the added value (i.e. I was looking for something to buy, or something to do) that it helps your overall relationship and lock-in with the customer.

  36. Chris Kenst

    Fred, have you considered using AVC in your examples? The advertising model could be explained in terms of how you might go about monetizing AVC (not that you would), the cost / benefits of each type of advertisement (display, text, etc).Regardless thanks for the informative post. Looking forward to the series.

  37. awaldstein

    Write up a post on them if appropriate and let’s dig in.Building value and acquiring customers is basically all we do all day long and anything that greases thought on this is really useful.

  38. Aaron Klein

    You have to choose one side of the network to seed…whether it’s waiving fees or paying them to engage for a time.It’s hard, hard work…and I’ve seen it fail more than I’ve seen it succeed. But when it works, it’s a killer business.

  39. kidmercury

    there are some companies i’m already a customer of, and i go visit their web site for a customer support issue. then i get the ad retargeting stuff forever. they’re wasting money — i’m already a customer!

  40. Robert Dewey

    Totally agree. While I think re-targeting is a step in the right direction (as far as getting rid of the annoying “I’d never EVER buy that” type of ads), it needs to be refined. Now they should be targeting you with notebook cases or something else related to notebooks, but not notebooks themselves. Heck, maybe people who purchase notebooks are 30% more likely to be going on a vacation soon, so they could throw a couple cruise ads your way. 😉

  41. Michael Brill

    The funny thing about intent is that everyone believes it’s the holy grail yet no one is doing much to improve capturing intent online. Search terms, retargeting, FB Open Graph… quite small steps. What if we could actually express our intent as explicitly as we wanted… for all to see? Seems like life would be a lot easier and “advertising” would be very, very different.

  42. William Mougayar

    ….with a $100 coupon?

  43. ShanaC

    they do that all the time….

  44. Michael Brill

    “Lenovo XYZ800” is a pretty weak signal… *but* it’s a start. So is “pizza” or “vacation.” Make it easy to get from that weak signal to fully-expressed intent and then be able to communicate that to the universe. I think we’d have a very different world.

  45. ShanaC

    we do, it is called behavior. talk vs behavior is very different

  46. Michael Brill

    Not sure I get your point… behavior is not intent. We can try to infer intent from general online behavior, but the signal quality is so weak and devoid of context that no matter how much fancy big data predictive analytics you throw at it, you only end up with a stream of banal, false-positives.OTOH, behavior can involve quite a bit of communication. I do think it’s correct to make a distinction between free-form natural language discussion (“talk”) and more focused, transactional communication. Talk is probably not a useful mechanism to express intent to a potential counter-party except for very high-value transactions. But there’s $16 trillion of retail transactions globally – most of which could probably be improved if there were better communication of intent and capabilities.Seems pretty easy to imagine that Charlie could express that he wanted a laptop by Saturday with a handful of attributes and then see what the world has to offer… intuitively that feels better than him poking around the internet and being followed by Overstock ads.

  47. Steve Palmer

    Love to be part of this discussion. We are moving toward becoming a business that arranges transactions and we have the two-sided marketplace issue @ccrystle:disqus mentioned.