Some Thoughts On Burn Rates
The startup and venture capital businesses are based on a general idea that you can and should invest heavily into your business in order to increase value creation, amplify it, and accelerate it.
These investments mostly take the form of operating losses, driven by headcount, where the monthly expenses are larger, often much larger, than revenues.
These losses are known in the industry vernacular as “burn rates” – how much cash you burn on a monthly basis.
But how much burn is reasonable?
I have been thinking a lot about this in recent years.
Instinctively I feel that many of our portfolio companies, and the startup sector as a whole, operate on burn rates that are too high and are unsustainable.
But it is hard to talk to a founder, a management team, or a Board about burn rates objectively.
There are no hard and fast rules on burn rate so you end up getting into an emotional discussion “it feels right vs it feels wrong.”
That’s no way to have a conversation as important as this one.
So I’ve been looking for some rule of thumbs.
One that I like and have blogged about is the Rule of 40.
The rule of 40 makes an explicit relationship between revenue growth rates and annual operating losses. Below 40 is bad. Above 40 is good.
But the issue with the Rule of 40 is that it is oriented toward businesses (like SAAS) where there is a well-understood relationship between value and revenues and ones that are reasonably developed.
So I’ve been deconstructing the Rule of 40 in hopes of trying to get to a more fundamental truth about burn rates.
And what I have come up with is this:
Your company’s annual value creation (valuation at the end of the year minus valuation at the start of the year) should be a multiple of the cash your company has consumed during the year.
That seems simple and obvious and that is a good thing.
But in order to make this work you need to lock down two things;
- how are you going to objectively measure valuation absent a financing event?
- what is the multiple?
The latter one is easier I think. The multiple should be large. 1x is clearly not enough. I don’t think 2x is either. 3x is borderline. I like 5x. I would want a 5x return on my annual burn.
I think annual value creation should produce a 3-5x return on annual burn. That feels like a good solid range to me.
The first question is trickier. If you have revenues, then using a revenue multiple to establish value is a good way to do this. You can look at comparable company financings and acquisitions and also public company valuations to figure out what revenue multiple to use. But you should be careful to understand that revenue multiples are a function of revenue growth rates. The faster your revenue is growing the higher they are.
So let’s do an exercise here to flesh all of this out.
Let’s say you are a $10mm annual revenue company in 2017 growing to $18mm in 2018.
And let’s say that you look around at public comps and companies similar to your are trading at 6x revenues.
So you can estimate that your business is worth $60mm this year and $110mm next year.
So there will be $50mm of value creation in 2018.
If you want a 5x return on your burn, you should not burn more than $10mm in 2018.
If you are willing to accept as little as 3x, you should not burn more than $16mm in 2018.
That’s how this rule works.
I like it because it is objective and will lead to rational discussions about burn rates vs emotional ones.
It does break down in pre-revenue companies where it is harder to objectively measure value creation. You can use financing valuations as a proxy in pre-revenue companies but then you quickly get back into emotional territory as the end of year valuation will be an aspirational number and unreasonable aspirations/expectations are what lead to unsustainable burn rates in the first place.
Comments (Archived):
Fred, that simultaneously was seriously one of the best and worst comments on this topic I’ve ever seen.First, you of course have the big topic nailed; the VC (or similar) business is about how much you’re willing to lose for how long in hopes of a payday.Second, “acceptable burn rate”, however you calculate it (and once again you of course have done a great job describing that), IS the best single metric for adjusting terms on the fly.But then, this kind of falls down.With all respect to the many and varied intricacies in the idea of adjusting terms (or issuing calls, or whatever), the idea that you need to impress changes midstream upon founders is frightening. As in … : anyone worth investing in to begin with NEEDS to have been and to have remained pragmatic enough to justify investment.I.E., without actually saying it you seems to be saying this becomes about ego at that point.And yeah, by nature folks in all parts of this biz have large egos, but either the math (and future projections) work, or they don’t. And when those conversations, uncomfortable though they are, take place … well, who has hand, anyway?Or maybe that’s your point: term sheets aren’t really addressing what matters?
I think that’s set in the relationship you build with the entrepreneur prior to investing. Term sheets are about the terms. Burn rates are about managing the company. A good VC will talk consistently with a CEO about management, culture etc. That way when the board meeting comes and they don’t hit their numbers you can talk about the real stuff and not just numbers
That was kind of my point. Fred touched on and explained two important issues WONDERFULLY. But then he conflated them when they don’t really collapse that way naturally … and the trick here, as with so many more general business issues, comes down to expectations management …at every step.
As you say, it’s tough to wrangle this into a good rule of thumb, partly because of the difficulty in measuring (and forecasting) “value.” Revenue growth might work sometimes, but as Pets.com and others have shown, it’s not always an indicator of value.
Hype and shallow understanding factor in to a lot of bad decisions – reactions vs. responses.
Seems like you are looking for a version of Return on Invested Capital (ROIC) for companies that have negative net income. Makes sense but one thing to consider in the example above is the long term margin structure of that incremental revenue. As an investor I would accept a higher burn rate for the same revenue generation at a 40% EBIT margin vs 20% ETC. The “quality” and recurring nature of the revenue is critical.
I am a total outsider to venture (so please forgive/enlighten me if this is off-base or naive), but it strikes me that 300-500% would be a lot of return to ask for over just 1 year on many types of spend/burn. Obviously, Fred you have way, way more experience here so I’m sure you see that range all the time.Still, I would think the multiple you target should be tied to the likelihood of the outcome. So, for example, if it’s a consumer-facing startup with a decent amount of history/growth already funding further customer acquisition, it might be a good idea to invest even if you’re getting a <2x multiple on that spend. 1.5x in 1 yr is still a very handsome return as long as you’re highly confident in the outcome, and believe that the business is sufficiently sticky/interesting & that building scale is worth it).Obviously if the source of burn is to fund salaries for a team working on the creation of a new product line that would deserve a much higher multiple (~3-5x+) since there’s presumably a good chance it won’t work as expected.I would think you could make a table of the required multiple for a given type of burn if you knew a factors about a business & burn:1) Likelihood of hitting expected ROI on burn (presumably should be tied to stage of business and amount of data you have on customers / the ROI of your prior spend). 2) Hurdle rate of your most recent investors (obviously seed/series A presumably require ~2-3x even for highly certain ROI spend, but if you just raised a series E from Fidelity and you’re thinking about an IPO, you might be willing to spend at a ~1.2x multiple provided you’re confident in your business model. Generically I would assume in each funding stage the “burn multiple” required declines)3) Downside for the business if the spend fails to produce expected result. Assuming you already have a viable business, it seems like regardless of the expected value of the “return on burn” it would be folly to burn at such a rate that if the spend fails to produce its intended result you would be out of business. If you are google you are free to fund moonshots, if you have 6 months of runway and a load of venture debt then regardless of your burn’s return expectations you need to get profitable at all costs (or raise more 🙂 ).
who coined the term ‘burn rate’?
I’d like to coin the term creation rate … angled from a positive light.
+ rate, which sounds more like building something up rather than burning something down. i’m not sure the VC industry has a + positive lexis.
VC as co-creator?
um, perhaps depends on which VC – as co-conspirator?
Fair. There are bad people still in the world.. weren’t loved or supported enough as children – serious.
sadly true, and i’ve come to the conclusion that many such people are at work in the startup ecosystem. misanthropic sociopaths are not that hard to spot, but are difficult to stop.
They’re even getting into presidency these days!Yin-yang cycle… it’s a fucker.
🙂
Burn rate is a good term. Even if spending is the right thing to do now, it is still the opposite of running a healthy business long term. Burn rate helps you remember you are playing with fire.
Yes, calling it anything else than burn pre-supposes an outcome. When you burn cash, it is burnt. Poof. Gone…If it was burnt the right way, it might create something in the future that can produce more cash than what was burnt. If not, what is gone is gone.
.Nero?Just guessing.JLMwww.themusingsofthebigredca…
Nero invented the term “burn rate”! Wow!
you have to start somewhere and as long as you are transparent about your method, the conversation can ensue. I do know that when I hear in a pitch that typically a company in such and such a sector is bought for 8x-10x top line revenue I want whatever they are smoking. Agree 5x feels like the right number for venture based on liquidity, risk, etc.
There has been too much money pooled to people who have no more than a shallow understanding of something. Capital is too easily available, yet the processes don’t seem to allow for the best, most thoroughly thought ideas to easily get enough seed funding – because these decision makers aren’t sticking to founding principles and metrics.
Burn rate should soundly be relative to the amount of proper preparation work. Do you know what you can use that next $10mm for? Can you execute it efficiently for what different parts of the funding are used for?An acceptable rate of capital expenditure will be at odds with feeling, as you mention, however the Captain of the ship hopefully has a good idea and understanding of what’s going on – if they’re being genuine and honest in where their effort will lead them and the organization; good shipmates will help, and it’s the teams role to educate the BOD about future plans as clearly as possible – however projects are always an evolving process, so spending $X on Y right now might not feel like an acceptable burn rate unless the BOD really understands the value being created.I imagine a high burn rate isn’t overly questioned the more trust people have of the leader, as long as everything is aligned knowing with certain certainty that you’ll have resources to maintain enough runway. Elon’s going to lead the creation of a colony on Mars, and with his current success he’ll have essentially infinite resources to continue on the paths he’s passionate about. If he doesn’t have the money yet, he just puts it on his roadmap. He was lucky of course he was able to start these large capital expenditure projects from payout from his previous successes – I think what we’re missing are giving people with a solid holistic plan the larger amount of money needed to give them the resources necessary to get all of the cogwheels working together. I’m sure it is why big markets and industries like housing and health are relatively slow moving giants, there’s no concentrated (to a single entity) fuel going towards visionaries.
“I like it because it is objective and will lead to rational discussions about burn rates vs emotional ones.”I applaud this post as a good step towards making these discussions more transparent to both investors and entrepreneurs because as you imply Fred, money is a mood changer. Having these discussions at the start or at least as early on as possible, should ideally lead towards more transparency between parties, as well as common language and bottom line. But what happens next when things go awry and either investors or entrepreneurs are not able to execute on agreements, or there are- as always seems to be, some intervening extraneous factor which creates a crisis for the company? Does that just written off as too much risk or bad luck?
Re: “It does break down in pre-revenue companies where it is harder to objectively measure value creation.”This is why I’m a fan of the idea of Convertible Notes et al with a discount and interest – and no cap. I’m not sure I’ve seen a discussion of what a fair interest and discount would be in different scenarios. For myself, I’m certain if I had $10mm I’d have enough runway to do the different projects and create revenue streams and get systems going where I’d never need another round again. Giving investors a minimum of 20% would be fair, however what if what I’m doing really takes off and I can find the next money at $100mm+ valuation – potentially turning the notes into ~10% equity. The company being able to leverage their success for more investment at a higher valuation is great for the first investor(s) if it’s a legitimate valuation and the funds will be used to increase efforts of the organization.
Social capital’s new public fund will let start ups delay this type of reality maybe forever. At some point you have to face the public market scrutiny
So the metric is a measure of how efficiently the company is converting spent cash into market value. It would be interesting to see how public companies in the tech space stack up against this metric. Tough with a sagging stock price.
I’d think you should expect a higher multiple in the earlier stages (to compensate for the higher risk). No?
Thnx.Be interested if you might take a stab at this from the ICO side of things when the money is already raised yet there is an obvious need to create a budget from the flip side.
I know; I know; the startup may have an expected ROI much higher than nearly all other investment opportunities, so, invest the money. Call that the first rule of investment (FRI).To follow the FRI, we need to know the expected ROI, but we don’t!So, instead, we play with simplistic rules (SRs). But the rules are (A) horribly bad approximations to the FRI and (B) hide what is crucial: We can’t use FRI. We know that whether we admit it or not. With the SRs, we’re not even approximating FRI. In case we didn’t know that, we need to know it and admit it.Instead, what is crucial is, and may I have the envelope, please [drum roll] first cut we’re making a bet much like other bets we make in life, business, finance, etc. We’re not doing FRI or approximating FRI; instead it’s a bet. For more than a bet, more than just betting on New England for the Super Bowl, we are getting on the team and actively trying to make the bet good. So, it’s a bet but an active one. But again, it just ain’t FRI or even a reasonably accurate approximation.So, f’get about FRI and try to see how good betting goes:One approach to betting, if we are going to use quantitative SRs, numerical data, models, maybe from a spreadsheet, etc. is to take several such inputs and look for projects where all or nearly all of the inputs say it’s a good bet. Or, we know that each of the inputs is unreliable, maybe even garbage, but somehow we talk ourselves into being influenced by those inputs when most of the inputs promise a good bet. Astrologers used to do something similar with the planets! That approach sounds like an attempt at CYA in front of others.But, investors don’t actually use SRs for very much. We know that. Instead they look for more: (A) They look at the target market. Is it, or will it soon be, really big, big enough for great ROI? E.g., when Bezos started Amazon, the data rate on the Internet wasn’t nearly fast enough to permit his sending the many image files he needs to send for his current success; the data rates are plenty high enough now! (B) How well fixed is the startup for attacking and taking the target market? (C) Is the product, service a “must have” or just a “nice to have”? What is there in barriers to entry? (D) What about the plans to take the market? (E) In those plans, what about the crucial, core, enabling technology, secret sauce, e.g., for pleasing the users/customers and being a good barrier to entry? (F) What about the team? E.g., do they look ready and able to do this? (G) What about the marketing, enough publicity, virality, cost of user/customer acquisition/retention? (H) For the solution/product, is it scalable? Reliable? Low, medium, high in cost, that is, what are the gross margins? (I) Are there great economies of scale? But is getting to big enough for those economies just forbiddingly expensive? (J) Are there some valuable network effects? On and on.My Point: It just ain’t ROI or SRs, not even SRs approximating ROI. We very much wish it were ROI, but we just don’t have anything like that much data. And in fact, as in questions (A)-(J), we don’t make much use of ROI or SRs, and we know that, know it very well.Bigger Point: We’re talking, call it, project evaluation. Business didn’t invent sex (@JLM), and VCs didn’t invent project evaluation. However good VCs are at sex, from what is known about project evaluation, just from average ROI we can see from the recent past, VCs look poor at project planning/evaluation.Poor? E.g., from KPCB, Google Ventures, and many more, how about investing $118,000,000 in a kitchen counter-top juice machine for $700 when anyone could do as well with their hands and grocery stores, especially Whole Foods, commonly have the juice business all in easier/better form anyway?VC average ROI? Well, there arehttp://www.kauffman.org/new…http://www.avc.com/a_vc/201…Project planning/evaluation are old.E.g., there was a story from the Middle Ages of taking an obelisk in Rome and moving it maybe 100 yards. So, there were lots of laborers, ropes, horses, wooden equipment, a big audience, etc. In the end, they were successful! Gee, great project planning, evaluation, and execution, right?Hmm, not quite so fast! How’d the obelisk get there in the first place? Well, it came as a solid piece from somewhere in the headwaters of the Nile. It was cut, loaded on a barge, floated downstream on the Nile and across the Med, and then on to Rome. There it was moved from the water to the place the Middle Ages effort found it. The work was done by Caligula’s slaves, ballpark 1500 years earlier (Caligula, Emperor AD 37–41). So, Caligula’s slaves were able to be successful planning, evaluating, and executing the project.Point: Project planning, and good project planning, go way back.Gee, by the time of Caligula, the Great Pyramid of Giza was already, what, 3000 years old? So, apparently the Egyptians did well with planning and execution of that project.But now for projects we have some advantages and more options than that project in the Middle Ages, Caligula’s slaves, or the builders of the Great Pyramid of Giza.So, here is an outline of what is promising in information technology (IT) projects now:(1) Find a problem. Want the problem to be so that, with the first good or a much better solution, enough people, and with average revenue per user/customer, will be high enough with good margins, etc. to make a sufficiently successful business.(2) Get the needed solution. Commonly we can’t do that. So, then, we have to return to (1) and pick another problem. So, really, we have to pick a coordinated pair, the problem (1) and the solution (2).(3) Do the work to build the solution, take it to market, and get the revenue and earnings.Now, (1) has us pick a sufficiently good problem. For that we first have to guess the problem, but we can do some quantitative analysis, even if just obvious on the back of an envelope, to try to add evidence that it is a good problem. In (3) we have to “go to market” with all the publicity, virality, marketing, selling, etc. that implies.That leaves (2), the solution. There we want (A) “the first good or a much better” solution. And we want (B) a barrier to entry. Well, one way to get (A) and (B) is to use technology, e.g., have some technology that is new, especially powerful, protected at least as a trade secret. We get to exploit current computing and the Internet; those two are not yet fully exploited and are increasing significantly in power weekly or so; but those two are readily available to all companies and not protected as trade secrets.So, for (2) the solution, we want “the first good or much better” and protected as a secret. Okay, it happens that we have had since 1942 or so the spectacular, astounding, unique, all-time, world-class grand champion of getting just such solutions. Right, here we’re talking just (2) and not (1) or (3), but it’s fair to say that (2) can be the crucial core.That grand champion? And may I have the envelope please [drum roll], “RIP”, and the winner is, nope, not one but two, (i) STEM fields in the best US research universities and (ii) the US DoD. Thankfully for US national security, (i) and (ii) provide by far the best examples of doing well on (2) the solution via technology.Right, some approaches to (2) are too expensive for a startup. Okay, then we need other approaches to (2). If (2) is too expensive or, as we mentioned above, too difficult, then we need to pick another pair of (1) and (2). But (i) and (ii) again stand as the A+ way to proceed for (2) the solution.For how to do project planning and execution for the (i) and (ii) approach to (2), we are just awash in examples.Really, the key is the first solution, just on paper, commonly with a lot of applied math as is common in the STEM fields. With good work on that paper, (i) and (ii) show that the rest of (2) is routine and low risk.uh, all across applied math, physical science, and engineering, project planning, especially the technical parts of the good solution, are done, and, again, may I have the envelope please [drum roll], “RIP”, on PAPER.Good news, super good news:In particular, sure, a key to the business is good work on (2) the solution, and, with that paper, success on (2) is routine and low risk. E.g., with the paper, there was no question about the atomic bomb, the hydrogen bomb, the nuclear submarine, inertial navigation, GPS, the SR-71, Keyhole (Hubble, but before Hubble and aimed at the earth), RSA (Rivest, Shamir, Adleman), stealth, TCP/IP and the Internet (from DARPA), and much more. Given the paper, all those projects were successful, routine, low risk.So, for project planning and evaluation, we can’t do much on ROI before good work on (2) as in (i) and (ii), and then we have a decently good shot. And for SRs without good work on (2), again we don’t have anything toward good project evaluation.But good project planning, evaluation, and execution go way back, e.g., to the Great Pyramid of Giza. We SHOULD be able to do better, at least as well, now.Alas, apparently what the IT VC world wants from higher education is a flood of desperate immigrants who will hack routine Linux code 16 hours a day. So, that’s junior college or trade school work.No one on the faculty at Cornell believes that such trade school work is the value of Cornell.E.g., with professor E. Dynkin, student of Kolmogorov and Gel’fand, Cornell has long been a world star in high end stochastic processes and stochastic optimal control. Maybe Dynkin never wrote a single line of C++ in his life. To Dynkin, stopping times, the strong Markov property, etc. are baby talk.Indeed, likely the computer science professors haven’t hacked Linux code in a long time if ever and are not proficient at Web site development, database administration, server farm planning and management, software project management, etc. They may never have touched a high end Cisco or Juniper router or LAN switch.And the US DoD, DARPA, ONR, etc. won’t be impressed with trade school technology either.Bluntly it appears that nearly no VCs anywhere in the country can evaluate technical work just on paper. The people doing (i) and (ii) can; the VCs don’t even try.Q. Does it matter? Uh, if a VC sees Snapchat, plays with it, sees the network effects and the rapid growth, why is a paper on technology needed?A. It’s not, and generally for projects like Snapchat, HotMail, and more, it’s not.Of course, that means that Facebook, Google, or Microsoft could also compete with Snapchat. E.g., Android could include something like Snapchat with each smartphone.But doing only projects like Snapchat would be severely limiting.Q. Why should a VC not just wait for startups to get traction?A. One reason is that there is little or no funding before the traction which reduces the number of good projects to be funded.A second reason is, by the time a project has traction, etc. enough to get a VC to write a check, it may be that the startup no longer needs or wants the check and the term sheet and BoD that come with it.For this second answer, likely the history of VC shows very few such cases, but in the last 10 years there have been some big changes: For a lot of projects, a solo founder can fund the effort out of even a thin checkbook, get the traction, and say to heck with a term sheet, check, and BoD.Q. What about equity funding for the big got to market, build out, rapid growth phase?A. Some projects need equity funding for such things, but some other projects do not. So, really, an entrepreneur should try to pick a project that doesn’t need equity funding for such things.For a lot of projects, equity funding for the rapid growth is not very important because (A) the project may have such high margins that there is plenty of cash for growth and (B) for some of the old capex, e.g., for the server farm, could use, say, AWS.Q. This thread is about burn rate.A. Yes, but with projects well designed and managed, that should be of little concern.For one, if the project is really good and that’s fully clear, say, from a good paper, then the ROI of the project should easily cover reasonable, anticipated, planned burn rates, including for unexpected external problems.Second, with just good ordinary planning and budgeting, burn rate should never get out of control or be a big surprise.For a good project, there should, of course, be good planning and budgeting. Big burn rate surprises for the BoD should be nearly unheard of.
Really interesting post. You almost never see this kind of info from the VC point of view.I would assert that burn rate is at least loosely correlated to exit rate, i.e. how the IPOs/mergers are flowing. When they are VCs think less about burn rate. When they are not, as has been the case the last couple of years, they think more about them in order to hopefully provide more “runway”. This is another VC term, i.e. the the time from now to the point at which the investments end before the company is viable.It might be interesting with all your data over the years to actually do this correlation. You can probably look at the history of your investments and pull it out. Perhaps you might even find that exit rate is a leading indicator that you can use to plan with based on your correlated figures.
This is a great discussion.One part is left out. What happens when you don’t hit your goals and when do you measure them?You are going from $10mm to $18mm in revenue.That means you can burn $10mm according to the 5x rule (now I’d rather see profits meaning you don’t need to burn)But lets agree to the $10mmThat means you can burn $2.5mm a quarter.Ok.But if you aren’t increasing revenue at $2mm a quarter, and I’ll give you a one quarter lag, you better adjust the burn.
For traditional small businesses with precious working capital, people adjust cash burn based on incoming cash flow. This is how business has been done since time immemorial.Now with startups that are venture funded, the promise is in increased valuation growth (not necessarily mapped to immediate revenues, profits, or cash flows as the hope is for monetization to unfold over time). Thus, there is more space to argue for more spending disconnected to short term cash flows.Fred’s model is a good one for post PMF startups with high growth, but for later stage businesses with stable, not crazy growth, being cash flow positive in a traditional sense is the best possible advice.
One of the questions that all companies, public or private, profitable or unprofitable face is – What is the value of the growth itself ?Growth can destroy value just as it can create value.Assume a profitable company that is reinvesting some of its post-tax income on growth.If the Return on capital on incremental investments for growth > Cost of Capital, its value will increase with the growth.If the Return on capital on incremental investments for growth < Cost of Capital, its value will decrease with the growth.The question of quality of growth applies to startups at the cash burn stage. In the example you have cited, you have not discussed the quality of the revenue growth while applying the same 6X multiple to $10 million and $18 million revenue.Now *I think* that you may be making the implicit assumption that the quality of revenue growth question is addressed via the burn rate itself (and therefore the multiple of burn rate to valuation growth is sufficient).But that is not necessarily true. As a simple example – (A)- The revenue growth is coming primarily from enhancing realizations from existing customers (recurring revenue; higher margin; little/no CAC) versus (B)- the revenue growth is coming primarily by new customer (high CAC) /geography additions.In these two scenarios, you can have a differential in other costs such as higher headcount /other investments for future strategic expansion in scenario A as a result of which the net burn rate for (A) and (B) are the same. But the quality of revenue growth in A is much higher, while B is masking a issue (inablity to retain / farm margins from existing customers and over-reliance on new customers).In this case, a 4X Burn rate to Valuation increase in scenario A would be better than a 5X Burn rate to Valuation increase in scenario B (applying your same valuation methodology of using a revenue multiple).p.s. For the same reason, your statement that P/E multiples should increase as growth rate increases is true from a “value” standpoint only if the return on capital invested for that growth is greater than the cost of capital. I say “value” here instead of “valuation” because “valuation” is whatever the market (or a few VCs in a room) decides on a given day…but over the long-term valuation tends to converge toward value.
The biggest sin I have seen is people deluding themselves on the amount of growth and spending money to try and achieve that goal.I agree that just focusing on growth can be toxic. But if you are in a VC business that is what you have to do (unfortunately public markets as well)I would disagree that having growth come from existing customers is necessarily better.Jeff Bezos says it best your margin is my opportunity. I personally have witnessed and disrupted companies that milked their existing customers. Really easy to steal.
In what I referred to as scenario A above, you are demonstrating the ability to retain customers which is a signal about the health of your underlying value proposition. By Farm existing customers, I didn’t mean “Milk” them. I meant increasing your share of wallet from the same customer from adjacent revenue streams (“Hunting” and “Farming” were two terms we used in my old company). That increases the total margin you can make at a margin $$ level, not a margin % level (in fact you might get more aggressive at a margin% level). But to Farm the customer, you have to retain them. And your margins overall are better when you aren’t spending on customer acquisition costs. Finally I didn’t mean that as ‘instead of’ growing new customers…rather like I said, it gives you the headroom for strategic expansion which will then get you new customers. Value growth from existing areas supports Volume growth from new areas. But you don’t get there if you are stuck in the vicious cycle of scenario B.I used to say two things to my team -1. You cannot open a bank account with margin %. We take margin $$ (delivering total opinc $$$) to the bank.2. It costs more to regain a customer that you lost than to hold onto them in the first place. And a century of free markets has demonstrated that excess margin % results in lost customers.
Yes, everybody uses the terms Hunters and Farmers.And I agree, providing additional value to existing customers at a lower rate is much better than finding new ones. And the key point to this is at a lower rate because it is less expensive.HoweverWhat I have found is that when you tell your team we need to grow revenue at X% this year, the easiest way to achieve that when you are behind (and it seems most companies are behind their rose colored projections) is to “milk” existing customers and it works short term.I give you Sears as the poster child.
Yes, it is a short term strategy and it usually comes to bite you at the end. Which is why many companies separate out the hunting and farming teams. They have different quotas and incentives.When in high growth mode, I believe it is super important to do both. Focus on one cannot come at the expense of the other. But this does leads to high burn rates… which are justified when there is growth in revenues and valuation and the underlying market hypothesis is valid..If not, bad things happen.Wonder if there might be case studies of companies who chose to focus only on current customers? It is basically a choice to run a lower growth, higher profitability business vs. chasing market share. In today’s world, it can be a risky strategy unless you are playing in a concentrated industry.
We agree my point is simply this:Understand that the more margin you take from one customer the more risk you have, unless you have a huge investment in infrastructure that was done in a really frugal way. There are two types of moats really high investment costs, or really low margins. Shockingly Bezos has done both.Yes your CAC can be high but you should know how long the return on the CLTV on that is going to take to recover the CAC In my simple mind if it is more than a 18 months you are deluding yourself, and that is generous.If you can’t kill your CAC and get to breakeven with the CLTV with the money you have in the bank you are in a precarious position. Just know that and if that is ok, ok.Because if things don’t go as planned and only once, once in my 30 year career with five different startups have I said: Holy shit I didn’t think we were going to grow that fast, and the growth on that didn’t last as long as I thought it would.
True. Great point about the twin moats of Amazon. They probably have the strongest barrier to competition among GAFA even though that seems non intuitive at the outset.
Today’s new customer is tomorrow’s existing customer. When you are growing market share in a high growth market, it is important to do both even though they have different levels of margins. It is short sighted to do just one not the other, because the real equation to monetize is CLTV: CAC over the life time, not over a quarter or even a year.This is exactly why we see SaaS companies willing to suffer huge losses (because of high CAC) to acquire new customers and not just focus on installed base. The mode of the market (high growth, moderate, low, or declining) matters a great deal as to what combination of hunting and farming investments are appropriate.
It is not just one or the other, that was not the point. It was about what the mix tells you about the quality of growth. Maybe I was not clear.
Yes, but my point is that while you can define the quality of the revenue and the mix in a particular period, it does not tell you the impact on the future. A new customer with high CAC who bought little in this period can do a mega deal next year significantly expanding his user base.I believe that for high growth markets where market share is at a premium, margins in current period are illusory. It is important to grab as much market share as possible. The life time value becomes more critical than optimizing for the current financial quarter or year. Of course, if there is no product market fit and if the market is not high growth this will result in a bad outcome.
Sure. But the CAC – LTV argument works both ways. One can justify high CACs based on customer LTV. That LTV can be a “plug-in” number in a excel worked backwards to justify the CAC. The LTV does not materalize while the CAC hit is taken upfront.Valuation is at a point in time.I think I did a poor job in my original comment and I don’t disagree with what you are saying here.What I was saying :1. In one scenario, you have demonstrated evidence of the ability to extract LTV (which is a signal about the health of your underlying value proposition) and in the second scenario you don’t. Wherein both are delivering the same revenue growth.2. Valuation is at a point in time. At that point in time, the valuations assigned to these two scenarios of revenue growth should be different.Fred’s example had not discussed that element of quality of revenue growth when assigning the revenue multiple to arrive at the valuation increase.
I get your points and agree.Also should be noted that especially in private financings, valuation is more a function of demand vs. supply for the stock at that point in time vs. a reflection of intrinsic value. Hence any justification of a burn rate based on notional increase in valuation can be a flawed premise to begin with.
Jeff Bezos says it best your margin is my opportunity. I personally have witnessed and disrupted companies that milked their existing customers. Really easy to steal.Bezos identifies and spends time on things where margin can be stolen. For example I don’t think he would have stood a chance trying to take money out of Microsoft’s hands in the 80’s and 90’s. I don’t think he would be able to make money from any business which relies on personal relationships that form the basis of why they are able to charge the premium that they do. Or luxury products. He is not out selling against STIHL, right? Black and Decker, yes.Not that your point is wrong. I have always thought (along the same lines) that companies make a big big mistake when they start out charging as much as they can because it’s what the market will bear and because of lack of competition.. All that does is attract attention and people who do napkin calculations (which might even be wrong) and decide “I can do that cheaper” or “look at the margin there wow let’s do that”. Plenty of ideas and businesses start that way.
Gillette is feeling that pain right now.
Ha, I get DM virtually every day of late either supporting or thwarting Nelson Peltz’s P&G board seat initiative. Can’t imagine how much money is being spent on both sides, all w/ the requisite spin. One P&G mailer isolated the ROI on Peltz held companies vs. P&G’s within a narrow period of time so they look strong, until you read the fine print and see their shenanigans.
I got one of those for PG as wel. I marveled at how poorly the mailer was designed. Not something widow and orphan money can figure out. [1]Without knowing anything at all about the situation, and not owning enough P&G to really even care, my gut instinct was to go with Peltz.I also marveled at the oh so typical board makeup many chiefs, no Indians (didn’t dig into this so I could be wrong but I don’t think so..)[1] For a consumer product company that makes their money by branding and shelf space I was amazed at how they send out the proxy (looks like circa 80’s) that in no way my mom would be able to figure out what to check off.
Agree, the P&G proxy design is horrendous and confusing! (I presumed this was perhaps due to SEC requirements.) I too voted for Peltz. I think P&G’s historic approach to brand building is antiquated, particularly in the digital age. They still have enormous clout at retail (re: distribution, shelf space), but P&G’s success w/ new product development, despite spending gobs on R&D, has been pretty bad. They are masters of the line extension, though. I figured Peltz would drive share price. He wants to break the company up to achieve greater value.
Don’t count out the tremendous number of TV advertisements touting the 1,200 jobs they have in MA.Let’s not go political but this is the exact thing that gets people fired up. Wall Street Billionaire pressuring you on one side, startup technology on the other, and outsourcing jobs overseas.These are the really hard discussions.
This is very true in customer relationships, employee relationships or even any personal relationship. If you take advantage of someone by over charging them, or taking them for granted, they will eventually go to someone else who treats them better. It is really just a question of time.
Yes that is a really good way of putting it. I have always argued this. I have had people say well we have to pay X to get new employee. Ok that is right the market says X. But we have proven employee that does the same thing that is making 90% of X. So to hire new employee we need to give current employee a 15% raise.No, no we don’t do out of cycle raises. Bullshit. Then we will lose that employee. They will never know the new person’s salary. If they tell we should fire them.You know it always comes out. Somehow always comes out. A mortgage application, a spreadsheet, or they tell it.But you know what is the most important? Not that I believe it always comes out but it is the right thing to do.
.A wise and thoughtful comment. All growth is not created equal and there are companies who have grown themself into an early grave.JLMwww.themusingsofthebigredca…
I agree that quality of revenues matter and all growth is not the same.But shouldn’t the valuation take this into consideration? Ideally, any investment is primarily optimized for long term valuation growth. A correct valuation, especially for a later stage company with more transparency, will assign a higher revenue multiple to components (and thus corresponding burn rates) that have higher quality revenues (subscription based, higher gross margin, low CAC, etc.).For example Amazon has wildly different businesses with varying levels of profitability (AWS, Retail, Devices, etc.) and the market does take this into account, valuing each component at a different multiple to revenues based on current (and) future profitability and growth rates.Also, I wonder if cost of capital is really relevant in the startup world in developed markets where the cost of capital is very low and startup returns are typically boom or bust?
As just an entrepreneur don’t kid yourselves: A VC is giving you a lot of rope. That is great. Seriously. But if you are using hope as a strategy, that rope is what is going to hang you.And don’t cry about it you took the rope.
Current biz growth is largely the result of prior investments and expenditures. I am not sure they are a fair indicator on whether you are spending money properly right now. When it comes down to understanding burn rate, I still feel a bottom-up approach works better than top-down guidelines.
This makes a lot of sense of course, but try to sit down an ICO’ed company and talk about it with them in such rational terms… Good luck.But I’m rooting for some of them who want to change the world without any constraints being imposed on them.
but try to sit down an ICO’ed companyThat makes sense. After all anyone who is even involved in the startup world (let alone ICO’s) would not be thinking rationally. They are not baking bread in Lancaster PA. Almost a cliche but that is both the good and the bad.
You are saying of course that a balance sheet is a restraint of course 😉
If they know what it is. Their balance sheet is their tokens worth.
Yup and that is why this is so hard to get your head around.Thanks
.What you describe is akin to going away for a 4-day weekend and telling your Airedale, “I have left enough food and water for you for 4 days. Pace yourself.”JLMwww.themusingsofthebigredca…
Humans are animals. And here is what I have noticed with my cat. I am not a cat or an animal expert.Used to be we would give cat (recent kitten) 1 big bowl of food all day and that’s it. He would eat the food when he was hungry I guess and there was typically some left over. My wife read something that you should titrate the cat food a few times a day. So now we give a scoop in the AM a scoop in the PM let’s say. But wait. There’s more. She gives the scoop when she leaves for work but then I give him a scoop when I leave for work because he sucks up to me and appears to want more food. Then the same thing happens at night he pays attention to me because I give him food and he sees value in that. Prior to the change in ‘distribution’ he didn’t have much of a need for me. Since I was not the one giving him the food at all. Now I’ve turned it into love and attention. Can even get a slight ‘trick’ out of him so that he appears to beg. And he seems to be much more appreciative of what he gets then when he was dumped a big load one time per day. It has given him something to get excited about and probably causes a rise in cat dopamine.The cat, like you should be doing in business, is always saving for a rainy day. So any opportunity to get me to give him a scoop he will do so. Really. Even if there is still food left over in the dish. If he spots an opportunity he will try to make it happen. (Analogy is raising money..)
Now that is one f(el)ine analysis!
Cats are good with humans beyond belief. Clearly, that’s why there are so many cats, most of them living lives of luxury!
Dude, you really rock!If you make the scoops random or conditional, I am sure it will result in even more extreme behavior.I guess this is how people train dogs, cats or even killer whales.Humans behave not that differently at all. Just that they may also be plotting to knife you at the same time.
Great post! – does “cat dopamine” come in a bottle?? I bet it’s frickin amazing stuff 😉
I will post a video of Rebel – our 4 month black Lab – peeling out of my son’s bedroom, downstairs and Iinto his food dish.And then I will change his name to Juicero.
Oh, there will be constraints.This time is not different.Oh yes, this times the terms are different.But when the end comes just like somebody said about Equifax I will root for the scum sucking lawyers that will pick the meat from the bones worse than any vulture.
>but try to sit down an ICO’ed company and talk about it with them in such rational terms… Good luck.This is the danger with these companies. If they cannot sit down with their advisors and not mentally follow a reasonably objective calculation they are more likely than not going to be in some serious trouble at some point in their company’s lifetime.
This advice is not relevant to the ICO companies :-)In most ICO situations, there is no product only the promise of it, let alone product market fit.No VC will fund most of these ICO companies. Many will die, but some might just produce something magical and game-changing. We need room for innovation to also operate in a wild west model.
Exactly right.
Rule of 3:- use the ratio for SaaSy plays- use the multiple method ( and evaluate after funding events ) for territory based land grabs – wing it on world changers.Would you have wanted to throttle FB?
.One may agree or disagree with your assessment — I agree — but what is even more important is the assertion of discipline into the thinking of the shape and expenditure of the money.I recently assisted a CEO in creating, what was for him, a very useful tool — a monthly graph of three different KPIs. Looking at it instantly created an awareness of their linkage and made a powerful statement as to how to behave.Entrepreneurs and boards need to focus on “break even” and “profitability.” Getting there has to be a significant objective. There is a lot of delusional thinking focused on not wanting to create those ironclad objectives.JLMwww.themusingsofthebigredca…
Until you are breakeven you are a slave.Sorry for the harsh term but you are a slave.Your master is the one providing the money. And while your master might seem nice right now, when things go wrong or the times are different….You are tied to the whipping post.https://www.youtube.com/wat…
.Wonderfully illuminating and dramatic. Well played.Amongst my various useless talents is I can crack a bull whip. I used to be able to split a styrofoam cup at ten feet.One is going to want a “cutter” for such a task.http://noreastwhips.com/bul…You have to be careful because you can hurt yourself if you don’t have good training.JLMwww.themusingsofthebigredca…
As a veteran of the dotcom crash, I fully agree. When you do get in that $10M revenue range, shouldn’t your goal should be to lower burn to the point where you are self-sustaining?Shouldn’t the investments you’re taking be considered with an X factor attached to them in terms of revenue and value creation (ie. take a $10M investment, which will allow you to drive $20-30M in added revenue and Y amount in added profit (assuming you are reinvesting that profit in the company) thus creating $ZM in enterprise value.From my standpoint, the VCs do want you to take more money to “accelerate” growth but if you do so instead of trying to attain self-sustainability, you are taking a substantial risk when markets turn sour (and they always do at some point). Investing in growth from revenue generated may be a slower path to growth but it is one where you not only gain more control but also can select when to take investments and not.Remember that the minute you take $1 from an investor, you’ve agreed to an exit timeline that shrinks every day.
Taking that first $1 basically takes you back to being an at-will employee again.Millions of sole proprietors are making good money but never took even $0.01 in equity funding,Compared with the businesses of nearly all those sole proprietors, a project in information technology should be an advantage. And for a good project, there should be plenty in retained earnings for growth fast enough that cash is not the “tight constraint”.
Agree. My goal has always been to know that I can get to breakeven before running out of money.I can see needing funding if you at $10mm and you are expanding geographically, horizontally, or really marketing.But those should have a return on investment.And the return will lag from the investment.But you should know how much and you should never lie to yourself don’t say oh it’s going to hockey stick up if you are not on the hockey stick.So if you haven’t raised money nine months out and you can see that you will only have three months of money and your return is six months. You take your foot off the gas and live.And as somebody that has bootstrapped companies I agree with you time is a very important thing.Now sometimes if you don’t act quickly somebody else will and you lose.Sometimes it takes much more time for things to grow than you expect.Fred has been very forthright in saying most companies are better off without VC. If you take VC money you are signing up for the first.
They are delusional by design! It is tempting to focus on easier metrics that do not involve hard decisions…
.There is an apocryphal story about a zealot founder, an out of control burn rate, a lack of basic due diligence, and a bunch of VCs who wanted something to work more than they wanted to evaluate its efficacy.Juicero.http://themusingsofthebigre…Juicero blew through $118.5MM in less than 4 years. Of that money, $98MM was raised in Mar/Apr 2016. Money came from some of the wisest of the wise — Kleiner, Google Ventures, etc.At the end, it was the burn rate of $4MM/MONTH which finally caught their attention.There was the added problem that their bit of high tech wizardry was not needed. They made a high tech juicer.Today, they are out of business and all the money is lost.JLMwww.themusingsofthebigredca…
In plain, simple English, isn’t this all just accountability. One can debate the appropriate metrics, but first one has to buy into the notion of being held accountable. There are far too many “get out of jail free” cards in distribution w/ out being subject to a magistrate’s gavel.
The relationship beteeen burn and rev will obv vary by biz model. If you’re strictly an e-commerce biz you’d expect to see greater immediacy vs. a freemium model where there likely will be a rev lag as you aggregate data and convert visitors/users via an upsell to a paid subscription, for example. Any correlation between acceptable burn and rev needs to appropriately reflect a company’s biz model, which I guess should be somewhat obv.
I really like this model as it makes it clear to all involved roughly what a company’s burn should be.I do however feel that this model may not work for Seed and some Series A companies. For the most part, these companies should be extremely frugal at all times as it always takes longer to get that next round of funding than anticipated.
I like the calculation very much, but think an additional obvious constraint is necessary. If you are spending $10m to create $8m additional revenue, are you sure that you have added value? E.g no additional $5m burn required next year to keep these newcustomers/revenues.
Great post! I’ll be putting this info *to work* this week, so thank you 🙂
This is great guidance. But it is also important to realize that this works best only in scenarios where the burn rate is seen as a means to drive growth post product-market fit. But there are many other scenarios involving high burn rate. For example:Pre-PMF:1. Company is creating something very new that can disrupt a large market, but needs a lot of zig zagging and iteration to get it right.2. Or the idea needs a lot of initial burn or capex for validation. (Say a wireless charging OTA product)Portfolio:3. Company has multiple products in portfolio at different levels of maturity. Cash burn needs to be seen not in aggregate but individually. This can be hard to do in some cases.Post PMF:4. Company was doing well, but is now encountering a major player. Company needs to spend money to strengthen product line or alter pricing to stay relevant.Figuring out burn rate through formulas may be good to take emotions out of the equation. But it is not a one-size fits all rule and needs to be used wisely. Every company is unique and different (Market opportunity, market dynamics, team, state of product, proof, etc.). Some situations deserve a longer rope than others.
Fred, In the formula, do you mean the incremental cash burn relative to the last period (or) the actual full burn? Ideally, it should only be the incremental burn to deliver the incremental valuation increase, but I am not sure…
Thank you, this is a helpful framework for evaluating burn levels. You touch on this in the article, but the problem of evaluating value creation for early stage companies (without a financing event) is difficult given the lack of relevant comps.You mention public comps as a base. Since it’s not common for public companies to be 1:1 on earned revenue models with a startup, this becomes a exercise in filing review and reconciliation.You also mention like-companies and their financing multiples. Again, the drawback here is valuing current revenue against financing that is likely largely influenced by future growth and product development.If we take the framework of burn as an investment in value creation, I think it would be relevant to simply keep burn tied to revenue growth (losing the valuation problem). This would free up the framework to apply to earlier stage companies (which burn doesn’t necessarily lead to a 5x increase in value based on revenue multiples).The role of the CEO and Board would be to apply discipline in maintaining the burn:revenue relationship, and evaluate the rates going forward.
I miss these MBA Monday-style posts. A genuinely new concept, explained through the lens of unique experience and insight, and broken down academically. With a formula to boot!I know you won’t arrive at posts like these everyday, but maybe you can make them happen a wee bit more often, eh? 🙂
When you are bootstrap and you start burning it really feels like a forest fire! I know it is investment but man does it ever create some stress.
Fred, just to point out how hard it is to apply a data-based algorithm to this problem, read these words:”There are no hard and fast rules on burn rate so you end up getting into an emotional discussion “it feels right vs it feels wrong.”That’s no way to have a conversation as important as this one.”…followed by…”The multiple should be large. 1x is clearly not enough. I don’t think 2x is either. 3x is borderline. I like 5x. I would want a 5x return on my annual burn.I think annual value creation should produce a 3-5x return on annual burn. That feels like a good solid range to me.”I do like the approach because it forces a discussion, and adds another angle to that discussion. Thanks for pushing the thinking forward.
Thanks for the guidance on such a murky and anxiety-inducing topic set – leadership by Fred and community taking first real quantified stab at this!TruBrain is at 17x or 13x for L12, depending on which angle of the methodology. Perhaps with digitally native vertical brands (your own branded widgets DTC), it’s difficult to get into stratospheric levels with inventory requirements on speed and cash flow.
Apparently, this calculation can be extended, isn’t it? If you have zero revenue, and you expect to create a company worth 50M, and you need 2M to create that company, then you can offer 20% to the investor, because at the end he will own 20% of 50M, equal to 10M, which makes 5x ROI.This assumes that the burn rate and the value creation are spread evenly through the years, which is true only on average.Also the 40% rule can be extended. If this year you have revenues of 100k because you’re just starting off, and next year you’ll have revenues of 1M, that makes a 1,000% growth, and following the 40% rule you can burn 960% of 100k, or 960k. By extension, if you have no revenues, an acceptable burn rate should not exceed next year’s revenues. Sweet.
Great post.As a founder it is hard to stand back and objectively assess your burn vs value creation, as so much of what we do is building our businesses to scale which in the short term is intangible however in the long run is essential. This is a great sense check!I have made a little calculator on a google sheet please feel free to use it.https://docs.google.com/spr…Tom
a good ceo (and steward of capital) also factors in:- the stability of the market and future revenues- a conservative view on the size of the market- a realistic perspective on unit costs / scaling- and what the ongoing CAPEX is needed to support market positionCouple additional thoughts:- EBITDA (especially adjusted adjusted EBITDA) needs to stop being confused with being profitable- For 99% of businesses if you didn’t have the option to be profitable in 3-years your business is not being run well or its not a business
Is the pig’s stomach really in it?