The unprofitable SaaS business model trap | Hacker News Reader

The unprofitable SaaS business model trap

blog.asmartbear.com

223 pointsaaronbrethorst98 comments

Companies in this space are also in an arms race where there are very finite amounts of scalable channels. (Scalable means, in this instance, any way they can convert money into customers in a predictable fashion.) One would naturally expect that the channels largely go to whomever is interested in investing most in acquiring them. However, since many of the participants in the auction (some channels, like AdWords, are more-or-less literally auctions) have extra-economic reasons to prefer growing, often the auction's clearing price will be non-profitable.

Or, to put it another way: the marketing company which has done best from the investment in marketing companies is Google. (Second best, probably "sales guys." Successful sales reps at enterprise firms are some of the best compensated people in software.)

That's what's interesting about this "arms race" to acquire customers: when startups aren't worried about profits, CAC can be arbitrary ("because we'll make it back later!").

The bi-product of all this is exactly as you describe: startups are artificially driving up the cost of AdWords (and maybe other channels, like salespeople).

Big brands do the same thing. Terms like "Diamond Ring" can't be profitable online since the big brands are willing to lose money to make sure they keep brand equity.

"Companies in this space are also in an arms race where there are very finite amounts of scalable channels. (Scalable means, in this instance, any way they can convert money into customers in a predictable fashion.)"

Just wanted to say - a very smart comment - and something that most people and frankly even VCs don't seem to get: that what works at $5m in bookings won't likely scale to $25m in bookings. This is why you see so many companies moving up to selling to the "Enterprise" - the Adwords driven, low-touch sales approach may scale for a handful of companies who were early and in extremely horizontal business - but for most more niche SaaS offerings, the writing is on the wall - that approach will get you only so far.

I love Jason's blog but I'm having some trouble understanding this post.

It's okay to spend $X on customer acquisition if $X is less than the lifetime value of a customer (where X ends up being rather high for enterprise customers). But if it takes (pulling this number out of the air) two years to recoup that initial $X, then each customer is unprofitable for the first two years. And if you're a growth-minded SaaS firm, it's going to feel like a lot of customers are in those first two years: but once your initial batch of customers pay off their debts, so to speak, their profit can be invested back into customer acquisition -- it's not like your profits have to be funneled outside of the company, or that your growth has to be rampant and unchecked (with enterprise sales, you're more or less determining your own rate of expansion by the quality and quantity of your sales fleet). Acquisition begets acquisition.

Besides the fact that you need a cash reserve (either through your own savings or outside investment) and patience, I don't see what's particularly wrong with this strategy.

I think he implies that the two years you mentioned are longer than the average customer life cycle.

The tl;dr version is that your company will never get to a size where the executives will suddenly go "oh, well, we're big enough" so you'll always be in that customer acquisition phase, and thus you'll never be profitable.

huh?

If the average customer brings in $500 and the cost to acquire the customer is $200 then you'll be profitable as long as the provisioning* cost is less than $300.

*everything else associated with a customer

If the cost to acquire a customer is $200 and taken in the first 6 months of finding a customer lead and the average customer brings in $100/year and stays for 5 years and starts becoming a paying customer after an initial 6 month sales period then you will not be profitable easily as long as you continue growing. It'll take 30 full months to amortize the net cost of acquiring a customer down to 0. That's a long time.

Also, if your growth accelerates you'll just keep digging deeper and deeper into a hole.

The problem is that the costs are all front-loaded. By the time your first batch has come through, you already have another batch costing you. If you're going to become positive, your per-customer expenses have to go down over time, or you'll never catch up to the debt.

I agree with you. It's perfectly reasonable for a firm to pour money into its customer acquisition machine so long as the return is greater than the cost of capital plus a risk premium.

Failure to reinvest every dollar under those circumstances is a Type I error.

This risk premium is an important part of that equation.

"So out of the original $4R, we’re left with $0.1R in profit. That’s 1/40th of the revenue making its way to actual bottom-line profitability, and even that takes 4 years to achieve", Jason Cohen.

That a very tight profit margin, but still could be valid business mode. Especially when you hope that over time 'brand' grows in strength and average customer acquisition costs may lower, conversions can be optimised, R&D costs will be shared across a larger user base. You might even 'max-out' the customer base.

The issue is that revenue is at risk. You might spend $300m acquiring customers for that $0.1R profit 4 years down-the-line. 2 years into that 4 year, a competitor suddenly innovates and steals the customer before you've realised the required revenue.

It doesn't even need a massive innovation. A margin that tight is very sensitive to very small changes. A competitor enters the market and your annual retention drops from 75% to 66.7% and that will probably be enough to destroy any hope of profitability.

I still don't know a single SaaS B2B company - at scale - that is actually profiting right now.[0] The general assumption is that the cost of sale will continue to go down (and thus become profitable), but once the market gets saturation (it is in CRM-SaaS for example), then the cost of sales goes right back up.

> The other company has to bust ass for measly 20%/yr maintenance fees.

This is a funny assumption. On-premise maintenance fees for the likes of SAP, Oracle, etc sit on a nice ~80% gross profit margin.[1] This model is obviously broken today, but has been a major source of revenue for traditional software.

Here is what I think I will happen to SaaS models (note - it isn't pretty): http://www.techdisruptive.com/2012/11/28/how-are-we-going-to...

Anyone know a large successful Enterprise SaaS that isn't selling jackets?

[0] - http://finance.yahoo.com/q/ks?s=CRM+Key+Statistics

[1] - http://blogs.forrester.com/duncan_jones/13-02-14-saps_mainte...

37Signals? (Perhaps not a coincidence that they are private.)

When 37signals started coming out with products, they were really early to the "self-serve saas web app built by cool web devs just like you and me" model. In many ways, they practically invented the industry. Fast forward 9 years or so and the market is very crowded with plenty of investor money allowing businesses that will never turn a dime to stay in business too long (sell lots of jackets)--on the premise that they can be like 37signals.

For startups from the last 2-3 years (or ones currently entering that business), given the competitive reality + the massive downward pressure on saas prices (not to mention upward pressure on user expectations), I don't think the analogy works as well.

Also, 37S never has published numbers, so who really knows what their growth/profitability has looked like over the past several years? I wouldn't be surprised if their growth has slowed significantly due to steeper competition and an inevitable cooling of their brand's coolness.

I used to work for one, LivePerson. Nobody in the valley knows or cares about LivePerson, but it's quite the little performer.

http://finance.yahoo.com/q/ks?s=LPSN+Key+Statistics

What does "selling jackets" mean in this context?

From the end of the article:

>It’s like the old Jackie Mason joke — A man is selling jackets at cost. The customer asks “how can you sell at cost, how do you make any money?” Answer: “I sell a lot of jackets!”'

Except jackets have no recurring revenue.

I think one point of the article was that recurring revenue is more complex than that. It all depends on pricing, time to recoup customer acquisition cost, and a few other things.

If it takes 18 months to become a "jacket" (ie break even), then whether or not it's worth it depends on factors such as churn, cost of account servicing, cost of IT, etc.

then whether or not it's worth it depends on factors such as churn, cost of account servicing, cost of IT, etc.

Sure, but that's stating the obvious for anyone in the SaaS business.

Any SaaS B2B company with network effects does this easily because the revenue from each customer is a function of the value they derive from the software, which increases with the square of the number of customers.

Think Excel (network effects derived from its communication properties), CoreLogic (derived from having all the data), Addepar, etc.

These companies also typically have high fixed costs across all of their customers once they are at scale.

they exist, i know of at least one example. but they are not public (yet), hence their financials are not widely known. soon though.

your techdisruptive article is very much focused on the SAP take on this, transforming themselves into SaaS.

there is big demand in SaaS for real turn-key solutions, rather than the classic generic enterprise software that needed an army of consultants to even boot up. customers want this and won't buy non-SaaS anymore.

> there is big demand in SaaS for real turn-key solutions, rather than the classic generic enterprise software that needed an army of consultants to even boot up.

"classic enterprise software" is turn-key....

The challenges for SAP to transform themselves into a profitable SaaS are no different than any enterprise SaaS trying to be profitable.

no. classic SAP or ORA is generic, they have certain industry templates but applying them is far from easy and fast. there is no way to simply turn on SAP CRM and start working with it say as a biotech.

This kind of feeds into a theory of mine (not well expressed I fear):

The earthquake so far has been Google's effect on Sales and Marketing:

- each person online is just one click away from every other (person?) web presence (business sites mostly, but other people increasingly)

- So if you were the most attractive business on the web you would get all the customers. The mechanism through which your attraction was discovered was search / linking.

The person with the most attractive business process will win next - as businesses make their next step in a chain open to all comers.

So any business process that can be digitised, and can be given defined interfaces, will find itself eviscerated from internal to a company and placed in a network of auctionable providers.

So, that's bookkeeping, accountancy, most of HR, calendering, scheduling, travel, hmmm....

The old idea of outsourcing all your non-core activities is looking like it really will come true.

Wish I had written that book on ebXML now.

For a company that is doing HR/office/mood stuff very well, like the company I work for (Prezi), these activities could not be outsourced because they are part of delivering the core values and feeling of the company and are instrumental in getting the right people on board (HR) and keeping them (office/mood).

Edit: following downvotes.

There is a seemingly sudden rush of SaaS companies at IPO / major growth levels in the B2B marketplace - how do people track them, or know about them? Is there a news outlet I am missing?

Add to that, the underlying sell for SaaS companies is either ease of implementation (which is a non-differentiator) or it is a genuine new activity (cross enterprise, co-ordinated 3rd party cookie tracking to massively increase campaign targeting / feedback) - so is there a discussion area on what these guys are doing underneath?

Basically - what am I missing?

My original for posterity:

I have never even heard of these two companies - where on earth does one find all these suddenly growing companies?

And frankly, is there a wiki page on what they are really doing under the skin (marketo / eloqua look like glorified dashboards for third party cookie tracking)

(Not that there is anything wrong with a glorified dashboard, I just like to know what people are really doing)

Here's a list of SaaS companies sorted by popularity: https://starthq.com/apps/

Popularity is calculated using a combination of the number of likes of the profile page and Alexa reach, so the first few pages aren't particularly useful, but once you get to page 15 or so you'll start seeing major SaaS players in descending order.

Edit: Marketo is on page 32

Thank you.

www.getapp.com. Search any business term (eg, Procurement) and add "Management" + "SaaS" or "Cloud" or "Software" and you will find a wide array of companies that I bet no one here has heard of.

And a lot (including my app) that plenty of people will have heard of. I do agree though, there's a lot of apps in that directory and elsewhere that will end up being well known (or at least acquired) in the next 5-10 years.

The network (cloud) is where you sell your consumer or enterprise application now, not Windows.

The majority of these aren't B2B, really, and of those that are, very few of them are "enterprise" players like Eloqua or Marketo.

Having put the list together myself, I'm certain that 99%+ are B2B. It's true though that there aren't that many "enterprise" players, although Marketo is there.

I just started scraping that list to do some brief analysis / categorising myself - if I come back here with the results later will you object?

Not at all. Let me know via the feedback form (available via the dropdown in the upper right when you sign up) what format you need the data in and I'd be happy to provide it as JSON.

Also, if there's some data missing that you want us to collect, check out our extractor API: https://github.com/starthq/extractor#starthq-extractor-api

It is often really difficult to hear about some of the enterprise players, unless you happen to work in a department they're targeting, simply because the vastly lower user base (usually orders of magnitudes below B2C or even SMB companies) generates a lot less buzz and mainstream coverage.

But you probably know about a few of them. If you're a developer, something like JIRA is probably roughly equivalent to Marketo or Eloqua in terms of size, market share and visibility. You've almost certainly heard of JIRA, but a good number of marketing professionals haven't. The reverse is true for Eloqua and Marketo.

Both of these tools (Eloqua and Marketo) do quite a bit more than just dashboarding, by the way. They both offer integrated marketing platforms, so they'll be sending e-mails, tracking visitors, running people through automated marketing programs like newsletters or drip onboarding campaigns, and a ton of other stuff.

(Disclosure: I'm an ex-Eloqua employee)

so if I described it as a dashboard-of-patio11's-brain would it be so wrong?

I just get the sneaking feeling that an awful lot of the companies are simply trying to land-grab what half a dozen good open standards would solve to the 80% level

I think by that I mean we are going to see a desire to build your own walled garden, but unless people succeed to the facebook level, there will be standards and consolidation, then a rapid rise of competing open implementations, and bingo, no more market.

"so if I described it as a dashboard-of-patio11's-brain would it be so wrong?"

Frankly - yes, you would be wrong. Not that these companies won't face commoditization like the rest of software, but they aren't simply dashboards - they're tools for managing all aspects of online and offline marketing. That includes email, landing pages, on-site forms, reporting, Salesforce integration, notifications/alerts, lead scoring, et al. There is a lot going on.

Where as in the consumer world you get rewarded for attention, in the SaaS business the most profitable guys try to be quiet to prevent a flurry of competitors.

It doesn't matter what business you are in, if you aren't turning a profit on each sale, you won't make it up in volume. In fact, volume will kill you fast.

I remember when Sony was selling the PS3 at like a $200+ loss at launch. I was surprised that Microsoft didn't take a couple billion dollars and buy PS3's. It would have cost Sony hundreds of millions of dollars and would have made the PS3 a money sink hole for even longer. Microsoft had enough money to probably put Sony out of business doing this.

Obviously, Microsoft could have got in a lot of trouble for attempting such a strategy, but the point is simple - an unprofitable business model makes you vulnerable, especially a growing unprofitable business.

If it don't make dollars, it don't make sense.

You can sell razors at a loss if people buy your razorblades at a substantial profit.

You are of course right. The problem is that software companies end up doing the exact opposite more often than one would expect. We do the equivalent of selling razors at a profit, and to make the deal sweeter throw in the commitment to provide "as many razors" as the customer can possibly demand, and replacement razors at heavy discounts... for as long as there's an entity called McRazors, Inc.

That's one problem. Another problem is bad estimates of how much the razors cost and/or how much profit they're actually making on the razor blades.

Printers and ink cartridges is the other example that comes to mind.

Same with ebook readers and ebooks. There are multiple examples of this. This is true even in software - sell the software for cheap, but charge heavily for maintenance and upgrades.

What are you talking about? This is standard practice. Microsoft also sells its consoles at a loss: http://www.neowin.net/news/report-microsofts-xbox-division-h...

The idea is that once you get your console in peoples' homes, you can make money off of the accessories and games.

This was covered in the post, but under the heading of "Undoing the effect of cancelations"

> Undo the effect of cancellations through up-sells/upgrades. Salesforce.com and ZenDesk charge more for every person you add, and more per person when you increase the features in your plan. Their customers grow (on average). Thus, their revenue over four years is not 4R, but rather it might be R on the first year, 1.5R on the second, 2R on the third, etc., so perhaps 7R in four years.

MS, Sony, and Nintendo hope that once you have the console, you will continue to buy games. And the console maker will collect a license fee from each game sold. If MS bought a ton of PS3 consoles, then the general public would have to buy a lot more games to make up for the loss. So the strategy does work, but it would work against any of the companies, not just Sony. But I think the PS3 had one of the biggest losses of any console at $240 - $300 depending on the version of the console purchased.

Also, somehow keeping your competitor's console from being available to real buyers will dramatically reduce the sales of actual games, since all of the sold-at-a-loss consoles will be sitting in a warehouse somewhere.

The really sneaky thing would be to figure out some non-game thing to do with the "enemy" consoles, like turning them into a supercomputer, or scrapping for parts, or whatever.

Herbert Dow would be proud: http://www.quora.com/Business/Whats-the-shrewdest-smartest-m...

Sony didn't see the PS3 as a console. It was a heavily subsidized Blu-ray player, that also happened to play games. So they lost that generation's console war, even though they were losing money on every sale, but won the much more strategically important video format war.

Also, thankfully for them, the PS3/Xbox360/Wii generation lasted twice as long as previous generations, so they had plenty of time to recoup this initial loss.

Interestingly, now the roles are reversed, with Microsoft trying to push a post-disc content delivery system that plays games, and Sony releasing a console hardcore gamers are already raving about.

If that's the case, then Microsoft isn't going to be very happy with how this generation is going to turn out for them. I mean, if they're banking on post-disc, then they don't have a competitive advantage over PS4 in real terms. The complaints gamers had about the DRM system is that it turns your discs into digital but doesn't really give you any actual benefit from doing so.

Yeah, from what I understood, Microsoft's original plan was to have the disc as a convenient way to prevent a lengthy, multi-gigabyte download and to save space on the console's hard drive, but they failed to sell this vision to the developers.

Did Sony lose the generation's console war? On a global basis it looks[1] like a stalemate between Xbox 360 and PS3. MS clearly won the US and maybe just about the UK but Sony won in many other markets. Sony burnt cash on it as an early loss leader and MS on warranty replacements. I haven't checked whether they both eventually reached profitability.

[1] http://uk.ign.com/articles/2013/01/10/report-ps3-surpasses-x...

Of course the Wii was the sneak attack from the previous generation that made Nintendo possibly the real winner although they are looking vulnerable now.

> I was surprised that Microsoft didn't take a couple billion dollars and buy PS3's.

> Obviously, Microsoft could have got in a lot of trouble for attempting such a strategy,

Umm, I think you answered your own question:)

And to go a bit further, the console strategy that Microsoft and Sony employ is to lose money selling the console to grow the user base and make it up on games sales.

Interesting given Get Satisfaction's recent approx. 10x price increase: http://blog.getsatisfaction.com/2013/07/16/the-latest-about-...

I just exchanged tweets with them this afternoon saying it would be interesting to see how this change affects their revenues (maybe if/when they IPO).

This also explains why we're seeing services like KISSmetrics, Dropbox for Teams, etc. moving to annual, pre-paid pricing rather than monthly with free versions

Yes exactly - churn rates too high, free versions not converting to paid, et al. And for companies like KISSmetrics and Dropbox, those free customers come with an infrastructure cost that goes up fairly linearly with the number of customers you add. Tough going if you can't convert from free to paid. And then you add high monthly churn rates and, ugh, you're in a world of hurt.

Wow. I hadn't seen that at all. That's simply reflective of the enterprise-focused path they've been on for the last few years. Micro-ISVs and SMBs don't pay real money for community software only the big guys do.

They are adding Success Services / access to their customer success team. Any idea what this looks like ? A consultancy service or general support ?

I disagree - there exist SaaS companies that have successfully executed and are on the path to successfully executing the profitable B2B SaaS model. Take Constant Contact, Responsys, and Blackbaud. While not all of these companies are highly profitable, they are all proof (and will be increasingly so) that B2B SaaS companies can reach profitability. The names we all know (e.g. Salesforce, Workday, etc.) simply need more time to reach that point of profitability, but they're on their way.

Looking at Jason's example specifically, I have a couple issues:

1) Assuming a fairly strong churn rate (~20%/year), the base of customers for which CAC has been repaid will make up an increasingly large portion of the user base as the company grows (in later stages). Forgive me if I'm wrong, but it seems much of Jason's argument is based around the assumption that acquiring new customers (S&M) in conjunction with ongoing R&D and G&A will always outweigh the gross profit generated by the existing customer base. Maybe if he defines "healthy growth rate" as 50%+, then yes, sure, it will always be outweighed, but let's be reasonable.

2) If Jason is going with 30% COGS, his LTV metric is off. No startup business in its right mind would continue operating with a CAC/LTV of 2.53. We're talking double that in most cases with a bare minimum of 3.

Finally, while this is a good discussion to have, I think we're all a bit naive to think that a bunch of small-scale startup entrepreneurs have enough knowledge, experience, or expertise to questions the decisions of many large, long-standing investment firms and successful individuals that all have supported these unprofitable companies with expectations of their eventual profitability. Having worked at a late-stage investment firm, I looked very closely at 100+ of the leading, big-name SaaS companies (we're talking 1000+ pages of diligence in aggregate). From experience, I can tell you there is plenty of work, far more than just a short article and some speculation, that points to the fact that these companies will reach profitability.

While the general assessment is correct, I think the core argument confuses marginal and fixed costs.

Say the average customer represents R dollars in annual revenue. That’s:

$4R of revenue over the lifetime of the customer. But: $1.5R is spent to acquire the customer (the pay-back period). $1.2R is spent in gross margin to service the customer (4 years times 30% cost). $0.6R spent on R&D (15% over 4 years). $0.6R spent on Admin (15% over 4 years).

The last two items strike me as decidedly fixed. That is, that until some critical mass is hit, there is no difference in cost for R&D and things like HR between supporting one customer, five customers, or fifty customers. Therefore it isn't appropriate to allocate a set percentage to each customer as once you've established an R&D department, each incremental customer is not contributing 15% of its margin to that cost.

Additionally, there is an inherent assumption that no matter what, as long as the company is growing it will necessarily be unprofitable. This is only true if you can assume that there is no point that your customer base is large enough to overcome customer acquisition costs. In reality, the pace of growth is probably going to level off at some point whereas the churn rate of the customer base could be low enough to turn a profit.

I know that the assumption was 70% retention but this seems largely speculative and unfair considering the considerable R&D spend. If new developments are made, one might assume higher retention is a possibility.

This bit also strikes me as double counting:

"And that is without any growth at all. But you need to grow enough to keep up with cancellations at minimum, so that consumes the last notion of profitability."

Growth to keep up with cancellations is covered by the $1.5R acquisition cost.

Although to disagree with you slightly I would say that Admin at least would be partially proportional to the number of staff (which is likely to be related to the number of customers) although improving the system so that less support was required (better documentation, easier to use software) should have a knock on effect here. Some parts would be fixed costs though.

An interesting question to ask is why do unprofitable SaaS companies get bought for very large amounts of money by other enterprise companies? The answers to that question are going to be my goals when building a B2B tool (because my ultimate goal is to make myself rich, let's be honest), not profitability (why would it be, when it's so hard to do and acquirers seem to not care).

Primarily for talent (the big guys are dying for innovation and they are paying for it) and client lists (prove your end user customers actually like using your software versus just being a business need). Sometimes it indicates the potential of an "untapped" market and I would see these acquisitions as getting a horse in the race (when Gamification/Social Enterprise blew up for a bit and you saw everyone buy a horse).

Prove people love using your software and you can get acquired if you want.

Actually if you push the reasoning in the article a bit it can make sense.

When they buy the company, they buy:

- the users and the revenue (which may be losing money for growth as explained)

- and also a product that should be easier/cheaper for them to sell to their existing customers. So they can have a very cheap batch of customers that grows their revenue a lot without increasing their upfront cost too much

-> it's like the up-sells/upgrade in the article: it's much cheaper to acquire but makes the same money

( eg:

- you have 1M clients

- you lose money because it costs you 1.5R to acquire a customer and you're growing

- the company buys you with 10M clients, and it's only 0.1R for them to sell it to these clients because it's an up-sells/upgrade

-> they'll make money with your product even if you can't! )

I was going to launch a SaaS in Spain oriented to the education market, but when studying my potential customers locally, did make me almost leave immediately by the expectations.

I did get a nice job offer in other country and finally did stepped out of my draft/idea.

Still, I've the feeling that there are SaaS to be created and be successful. But maybe it's just a minor representation of a big no-no.

I'm curious about that retention rate - most companies that I know that have gone with Marketo have made a long term beat - not just the software but in terms of implementation, training, data, et al. I'd be surprised if Marketo has a 75% retention rate - I'd expect it to be much higher. Now, they also sell at both the SMB and the Enterprise level, so his numbers maybe right on average.

Now that being said, I do think Marketo's business model seems to be upside down - and I'm a Marketo customer. What I expect to happen: crappy number over a number of quarter drive the price of the stock down to where they are worth about $500-600m (vs. about $1b now) - then someone buys them for $800m (about the same price as Eloqua went to Oracle). $800m would be close to what Eloqua went for (multiple-wise).

I would have guessed closer to Jason's number, but they quote a retention rate close to 100% in SEC filings.

http://investors.marketo.com/secfiling.cfm?filingID=1047469-...

Subscription Dollar Retention Rate. We believe that our subscription dollar retention rate provides insight into our ability to retain and grow revenue from our customers, as well as their potential long-term value to us. Accordingly, we compare the aggregate monthly subscription revenue of our customer base in the last month of the prior year fiscal quarter, which we refer to as Retention Base Revenue, to aggregate monthly subscription revenue generated from the same group in the last month of the current quarter, which we refer to as Retained Subscription Revenue. Our Subscription Dollar Retention Rate is calculated on an annual basis by first dividing Retained Subscription Revenue by Retention Base Revenue, and then using the weighted average Subscription Dollar Retention Rate of the four fiscal quarters within the year. Our Subscription Dollar Retention Rate was approximately 100% for each of 2011 and 2012.

Or, if you're wondering how that accounting maps to operationas, "Marketo's upsells to customers in year N+1 almost exactly cancel churn in yearly subscriptions since year N, when aggregated."

[Edit: Whoops, now that I think about it, they're sort of juicing that metric by construction. In a growing company with 1 to 3 year payment terms, quarter-to-quarter churn could be very close to zero even without churn being near-zero.]

As a public market investor, this sort of thing drives me batty. Many companies won't release an actual churn rate, but only a "dollar retention rate," as above. Yelp is one of them.

While the dollar retention rate is certainly nice to know, knowing the actual customer (logo) churn is critical to evaluating the cost of acquisition and overall profitability.

As Jason points out, if it costs too much up-front to acquire each customer, you can still go broke even with a nice-looking dollar retention rate--you still have to pay to acquire the customers you lose.

Marketo's costs for enterprise level service requires a long term bet. The company I work for has invested millions in Salesforce integration but 8 months later I've received 1 lead which was dead wrong.

Our industry doesn't lend itself well to social marketing so that may be part of it, but I wonder how many businesses will actually see a return from using Marketo. Executives stuck in the sunk cost fallacy will keep them going for a few years even in their worst accounts. Their market presence is so recent that their retention rate has nowhere to go but down.

I have to ask: what did they spend millions of dollar on in terms of Salesforce integration? The software is integrated with SFDC right out of the box (it is, as they say, a native SFDC app - whatever that means). It took us a total of 6 weeks to launch and about 1-2 weeks of work on the SFDC side.

Are you counting just their services, or other things?

Just my opinion, but if your company spent that much money on integrating Marketo with Salesforce, something is deeply, deeply wrong.

Some food for thought in here. But I'm not sure it's necessarily a problem with SaaS businesses. His conclusions depend on the arbitrary numbers in the example analysis.

It seems to me that his criticism is with the pricing rather than the actual model. If you could have charged $100k up front plus 20% maintenance fees, and are only charging $5k per month, your pricing probably has issues.

You could also charge setup fees for customers who need high touch introduction / initial setup to help recoup those costs.

And the sliding pricing scale (eg Salesforce, where you pay more as you have more success with their software) can help you grow customer revenue per customer, over time.

Still, some good points to think about. I just don't think it dooms the SaaS space to crash. It just depends on the value you offer, and how you structure your pricing.

This reminds me a lot of the example that Bezos used in his April 2005 Amazon shareholder letter (http://www.scribd.com/doc/111244354/Amazon-Shareholder-Lette..., page 27).

I'd love to hear from people who know finance better whether the comparison is apt.

The analogy seems to be that when you have upfront expenditures that get paid back over time, you may not have positive free cash flow if you're continually paying up front for later pay back, and you end up with a business that paradoxically generates higher cash flow with slower growth.

Marketo was founded in 2006. Eloqua was founded in 1999. They're now only $10 million in revenue apart. If you were asked to it wouldn't be too hard to put together an argument supporting that Eloqua was too conservative and allowed Marketo to overtake them by being overly aggressive.

Your job, if you're running these companies, is to make your shareholders happy and create a large return. i) Did this happen for the VC's involved? ii) Is this now happening in the public markets? Those are the (only) questions that matter.

Businesses that don't make money will always fail. For some reason, the tech industry finds new and interesting ways to delude ourselves (and investors!) into thinking this law doesn't apply. But sooner or later, like gravity, it exerts its pull.

Make something people want. Sure. But more importantly, make something that costs less to make than people will pay.

Instagram.

There are businesses that can succeed at it, but they rely on not making money until acquisition, where the acquirer bets that they will add other value down the line.

That said, it is not an avenue I necessarily want to follow.

Instagram hasn't proven it's ability to make any money for anyone yet, and I doubt it ever will.

I don't doubt that sometimes there's a greater fool you can scam into buying your money pit, and even provide return to your investors. It's crazy exceptions like this that keep the good-money-after-bad merry go round turning.

But it is not now, nor has it ever been, a viable path to building a sustainable business.

But Instagram didn't make money as a company for itself, Instagram itself was sold and made money for the founders/investors. The two are not the same.

I'm not making that argument necessarily. If you only define success as making money in general revenue, it doesn't fit the bill.

But to me, selling for $1bn (at the outset), is definitely a success.

For the founder and the seed investors and the VCs, certainly it was a great success. From my perspective, that's a type of success is exceedingly rare, akin to winning the lottery.

That mindset doesn't fit with my own personal goal of building a business that returns consistent profits, preferably as soon as possible after launch.

I'm curious if Yammer fell into this trap or if they were able to reach profitability with any of these techniques. Seems like they had a bit of viral growth but some initial searching implies they hadn't turned a profit yet.

Also, the press seems to only want to talk about their revenue growth.

Yammer was not even close to a profitable business at the time of acquisition, not was that ever a goal.

I found this pretty accurate. He might have overstated churn retention is closer to 85% but I think the author is pretty much "on" with this post.

Until the public markets start valuing these companies on earnings and cash flow instead of sales, expect more of the same.

This is a real world example of the Sustainable Growth Rate, which is a financial concept that applies to any business.

http://en.wikipedia.org/wiki/Sustainable_growth_rate

A seriously bad article. I don't know the company in question, but there are always companies in industries that either a) are very profitable, b) not profitable yet, because they invest in growth (profit = sales - (marketing) - R&D) or c) are unprofitable. Many software businesses don't earn money or low rates of ROEs - so what? It would be much more interesting to look at some cross-sectional data than to speculate based on sample size of n=1 ?! Btw, looking at the website of MKTO I'm not surprised, although the public market puts a tag of 900M$ on this company, which in 2012 made 60M$ in revenue while having netincome of -35M$. this has nothing to do with the SaaS model in general IMO (working for a quite successful one myself). Constant market-fit, zero-cost deploys, world-wide customer-base - what more do you want? Compare that to 5 -10 years ago, or making your living with custom software.

If your core business isn't profitable, SaaS won't get you there, but it is a great model if you can get big. If done right, the value from not having to re-implement at each customer site is huge.

This doesn't seem so much about profitability as cashflow.

Cash is, as they say, king.

You can be unprofitable for a long time if your cash position can bear it.

If your cash position is bad, then no amount of projected profits will save you.

Great post but the 75% retention thing was odd. All of the SaaS companies we know would flip their shit if annual retention dipped below 98%. Even 98% would be very painful.

98%? Can I ask who your SaaS companies are selling to? If you are reaching a diverse business audience, I would think > 2% of your customers each year would implicitly cancel by going bust.

That would be surprising. If more than 2% of my customers every year were going out of business, I'd come to the conclusion that I was selling to the wrong audience.

Annual retention is honestly considered fairly strong <20%.

For SaaS? That means you're turning over your entire customer base in less than six months.

You said annual?

Yes. Are you saying that annual retention (not churn) of <20% is considered strong?

Sorry I meant churn, but >80% annual retention is very good. 98% annual retention is unheard of in pretty much any decently sized business I've seen.

At a certain point customers will find your product without the SaaS company having to put money into it. This is where you become profitable.

Reply on news.ycombinator.com