Why project-based businesses generally don't make great search fund deals

Each week my team ploughs through thousands of company records in search of those that might be a good fit for a search fund. 

The process can be tedious, and it starts with the obvious. Disqualify the nonprofits, restaurants, hotels, and property developers. Eliminate the gun shops and brothels. Remove any company that has a foreign HQ or has gone out of business. Take out GE and Little Joey’s Lemonade Stand.

Then we eliminate certain industries in which we’re uninterested for various reasons. 

Then comes the more interesting part of the screening process in which we evaluate the business model of each company. In particular, we look at revenue quality, in broad terms. How likely is it that a customer who pays us $x this year/month will pay us at least $x next year/month?

The goal in this exercise is to find predictability. Even if that new searcher-CEO doesn’t do a whole lot in the first 6-12 months post acquisition, the business should still be chugging along fine, generating about the same level of revenue as the year prior. This beautiful wheel keeps turning thanks to long-term customer relationships or contracts, an enduring need for the company’s products or services, and a management team that knows how to keep the engine running.

Then once the new CEO understands the business and is ready to step on the gas, her efforts will only add to the existing base of reliable revenue and EBITDA.

Not all businesses lend themselves to this blissful outcome. If the revenue is less predictable, the deal is less likely to play out as above, and the searcher is more likely to devote her efforts to replacing past revenue than adding to it. 

A number of factors can contribute to the predictability of revenue - cyclicality, seasonality, sales cycle, contract value, contract structure, macro trends, and even interpersonal relationships, all of which should be considered when evaluating a deal or industry.

In addition to the above, we look to filter out what we term project-based businesses. In our lexicon, a project-based business is one that sells a customer a product or service once or twice without a reasonable expectation of future revenue from that customer. 

Consider an architect who custom designs homes for high net worth families. There is no denying that this can be a lucrative profession for that architect, thanks to the high price elasticity that often comes with luxury goods, and the high margins that may result. One project may be worth millions to the architect. But that revenue is generally neither recurring nor predictable, because the architect has little assurance that the homeowner will have another project to design in the future. If the architect sells $5m of projects every year, the business is not growing. She is not adding to her revenue base, just replacing it.

The shift in the software industry from installs to SaaS is a reflection of the power of recurring revenue. I’ve been listening to How I Built This with Guy Raz a lot lately, and two examples come to mind. 

The origins of Mailchimp are in web design. Founder Ben Chestnut was doing one-off web design projects for ad agencies, and then one-off email newsletter projects for SMEs before creating their own software. Even then they started by charging customers per email sent before moving to a subscription model, which is when the business really started to take off. The one-off projects were bringing in $10-30k each, and the email software subscriptions were bringing in tens of dollars each per month, but it was the latter that proved to be the fuel for their engine.

Similarly, Atlassian started as essentially a third-party help desk for enterprise software. They would charge per problem solved for their clients, and founder Mike Cannon-Brookes laments what a terrible business that was. Fortunately, they built some internal tools to solve those problems, and these tools would eventually evolve into the subscription-based software that would bring Atlassian to the $65 billion market cap it boasts today.

Both companies moved from a project-based solution to a recurring revenue model, and before making the switch both suffered from the burden of having to replace every earned dollar with new business just to keep afloat. 

(Now of course recurring revenue isn’t the only thing SaaS companies have going for them. They also often have amazing economies of scale, talented people, and brand caché that make them more expensive than other companies with equally predictable revenues.)

If the idea of a search fund is that the new CEO is landing into a company whose wheels are turning reliably without much effort from that CEO in the early months, then a company with predictable revenues, ideally in the form of long-term contracts, is better suited to a search fund than a company that subsists on one stand-alone project after another and will therefore likely require the CEO’s immediate attention to that continuous effort to replace historical revenue with new business. 

And when that CEO does turn it on, each dollar she sells will endure for years and allow that CEO to make 1+ 1 = 2, rather than 1 - 1 + 1 = 1.