Cloud Strategy

SaaS Failures | The Recurring Revenue Mirage

There is an argument that has been passed around the SaaS community that software-as-a-service requires a long runway to profitability due to high up-front infrastructure and customer acquisition costs coupled with the long payback period of a subscription-based revenue model. While this seems reasonable, my personal belief is that it is complete nonsense that usually just provides an excuse for poor management. Here is why.

1) Infrastructure investment is really not that much.

While infrastructure can be significant for a single customer, the cost per customer drops dramatically once you have reached a few hundred accounts. Most SaaS companies don’t spend more than 10% of revenue on infrastructure. This is the whole point of SaaS, lower TCO. While infrastructure cost may be significant, it is not enough to justify the huge losses that are typical of many SaaS vendors (including some prominent, ostensibly successful public companies). The vast majority of costs of any software company is labor. This is still the case for SaaS. And usually, most of that labor is in sales and marketing.

2) Acquisition labor costs are fixed, but not easily avoided.

The common misconception that SaaS companies can and should recoup acquisition costs over several years of recurring revenue comes from a textbook investment model where a fixed up-front investment is paid for over time by a variable income stream. Like buying an expensive machine that produces lots of inexpensive widgets. Or more similarly, spending heavily on direct marketing to sell high margin magazine subscriptions.

Under this theory, as SaaS vendor can justify high acquisition costs compared to an annual subscription price, because these costs will be recouped over the full lifetime of the customer…right? Wrong. The fallacy is that these are short term analogies applied to a long term problem. So, they distract us with fixed and variable costs, when our eyes should follow the avoidable costs. As the short term turns into the long term, each new fixed investment cost is avoidable. If it does not meet ROI requirements, the companies in the examples above simply have to avoid buying another machine or stop sending out direct mail respectively.

But, most of the acquisition costs in a SaaS or software company are labor-related, i.e., avoiding acquisition investment implies firing sales and marketing staff. Not likely. What is more likely to happen in an innovative, growing software company is that these resources will be reinvested ad infinitum, whether or not the new investments match up to earlier ROI. Or, they will be redeployed into account management and support roles as the company matures. Either way, any historical plan to recover early acquisition costs will be long forgotten. At least until a layoff forces the repressed memory back to the surface.

Enterprise and Web infrastructure software companies grew very rapidly throughout the 90’s, because they were expanding into a supply-constrained technology vacuum with deal values that justified high acquisition costs resulting from an outbound sales model. Most SaaS companies are expanding into demand-constrained markets like hard-to-get-to SMB segments, new unproven application sectors and competitive replacement markets…all of which can be tough going and are better addressed by an inbound sales model. Over investing in customer acquisition is like pushing on a string. A better strategy for the vast majority of SaaS companies is to establish a stable, profitable cost structure as early as possible, then grow profitably by accurately matching acquisition capacity to the market demand they can capture or create. And, build a company culture grounded on technology innovation and a lean, efficient operational approach.

Comment on Facebook!


  • Joel – Thanks for this article! It is a point I’ve been wanting to approach for a long time.

    On infrastructure – add to your arguement the point that if the infrastructure is cloud-based (which it certainly should be in my opinion) it is flexible in both directions. It can expand to meet demand growth and retract to meet slowing transactions.

    On customer acquistion – if the marketing group is smart, they are investing in ways to leverage social media and the network the application rides on – the Internet! I see so many leave this on the table for a competitor to grab it is just plain unbelivable.

  • Joel – I agree with what you are saying. I owned a SaaS oriented business from 2002-2008, raised over a million dollars in private equity on an a weak sales forecast, then invested too heavily in infrastructure and personnel, assumed the profits came over the lifetime value of the customer, fought for every sale, had much higher support costs than expected and ultimately ran out of funding.

    The net result was my loss of the company to the investor and my dismissal. A bitter lesson, but one that supports EXACTLY what you are saying about mismanagement. In the end, I like to call it “the perfect failure” because of the lessons learned. I was fortunate enough to gain incredible insight/experience into the space and am contemplating going back for more, hopefully more successfully and definitely wiser…

    For those reading this who are in the SaaS space, what he is saying is absolutely true.

  • Thanks for these insights, Joel. I agree that the cost of acquiring customers, while important with on-premise applications, is doubly important in the software-as-a-service model. There are, indeed, some prominent SaaS companies who are making a big bet that their customer base and revenue stream will grow fast enough to cover their sales & marketing spending.

    For some, with deep enough pockets, this may work out. Unfortunately, others may end up like WIle E. Coyote in the cartoon. The too-clever predator runs off the cliff, keep his feet moving, and stays airborne… until he finally looks down at which point the floor of the canyon quickly comes up to meet him.

  • Hi Joel. I found this view quite interesting. We are a Company that is helping IT Companies to set up and manage their channels in EMEA and in the US (we are just starting there) and we have noticed that many SaaS Companies have lost time and money in trying to approach traditional IT resellers/VARs to resell their solutions. Obviously the channel value proposition did not match and this has caused many disappointments.
    As you somehow suggest, I think SaaS requires a new go-to-market with a much more flexible operational approach and new categories of partners.
    Should you be interested we could elaborate on this.

  • Joel,

    Thanks for this post. If the fundamental problem is over-investing in client acquisition and your salve is an inbound sales model, can you describe what that looks like for an early stage or startup company that is, in many respects, attempting to create a market for its offering?

    In my experience, creating the catalyst in an enterprise for the type of annual financial commitment that a typical SaaS sale represents (i.e., >$20k/year) usually requires some significant outbound sales effort, even with robust marketing air cover.

    • Hi Tommy,

      Great question! At a price point of 20K per year, you are likely starting to cross over from a transactional sales model to an enterprise sales model. To achieve a transactional model at high price points, there are two fundamental roadblocks that must be removed (SaaS Sales Tips 7 & 8). First, marketing cannot simply provide “air cover.” It must provide a steady flow of qualified inbound leads. Half of this equation is standard inbound marketing techniques. However, the other half lies in the product itself. It would be essential to have online trial and purchase AND a product that is designed such that online trial is EASY and a true first step to adoption (not a demo). In short, you must remove complexity from the purchase. A good recent example of this is Zendesk, which has achieved an exceedingly high revenue velocity with a minimal outbound sales effort i.e., none in the early years.

      This does not mean that you will do NO outbound sales. Given the higher price point, you can certainly afford to. The relevant metric is your payback period on the individual customer, which will be shorter the lower your acquisition costs. And, the higher the mix of inbound : outbound leads, the lower your acquisition cost.


  • Hey Joel,

    Definitely with you on the 1st point re: infrastructure. Not there yet on growth. In my experience you just can’t build a SaaS business of any scale unless your churn is sub 5% (even that’s bad). Meaning your customers stay with you for 2 years or more. Given that reality, why does it not make sense to invest in acquiring more of these 2 year customers?

    The business model should be profitable every step of the way (i.e. per customer economics should always be positive), but I do believe in investing behind scalable growth channels to acquire more of these profitable customers.

Leave a Comment