Distribution First

This is one of those lessons I have known about in the abstract, but recently gained a lot more definition on.

When we started Glints, we cracked our heads trying to figure out product-market fit for several years. We generally languished and barely grew, especially in revenue. We have this narrative that our inflection point came after an investor advised us to figure out our sales engine. The predominant lesson I drew from that narrative is that systems and positive unit economics are key.

Now that I have heard about the stories of a few more experienced entrepreneurs, I have a slightly different read of that inflection point. Yes, the systems is a key piece of the puzzle, and so is the product, but what really happened was that we cracked a scalable distribution channel, which is B2B sales through a sales organisation. We found channel-model fit.

David Sacks talked about this in one of his talks with a Draper University class. In response to why the Paypal mafia was so much more successful at creating unicorns compared to say the Google or Facebook alumni, his hypothesis is that Paypal's founding team was steeped in treating distribution and product equally. Whereas Google and Facebook almost lucked their way into growth with the right channels, Paypal had to cultivate that distribution mindset more deliberately. He gave the classic example of Paypal giving $10 for every referral, as well as the compelling hook for recipients of Paypal transfers to sign up and receive their funds.

Now, I'm not entirely sure about that comparison with Google and Facebook - I don't know enough of their early days to judge - but at least in Sack's own case, that distribution mindset was very obvious. In Yammer, he pursued a strategy of bottoms-up enterprise sales, with a chat product that could spread easily among employees to gather critical mass, before he sends his enterprise sales folks to close the CIOs. There were some deliberate choices I could see going into that, including embedding virality mechanisms into the design of their chat, and the design of a freemium pricing plan to promote virality, while supporting the CAC of an enterprise client (i.e. the salaries and commissions of a enterprise sales team). In fact, he believes in this bottoms-up enterprise sales distribution strategy so much that he called it out as one of his investment theses in his own VC firm, Craft Ventures.

The second example is Jack Abraham of Atomic.vc, a company-builder that has seen pretty decent successes as far as company-builders go. In his interview with My First Million, he gave some colour into their ideas testing process. He talked about testing an allergy medicine company. Even though the idea was very promising, and the problem extremely painful, somehow the unit economics did not pan out. Specifically, the CAC + COGS is way too high relative to an acceptable consumer price point. From the sound of it, the conversion rates for paid marketing might have been too low.

In his experience, the gold standard for CAC payback is 3 months, and LTV:CAC ratio is 5x. Triangulating with another article about one of the successes from Atomic.vc, hims, they mentioned that they tested the idea with landing pages that were so crappy they almost didn't launch it. The landing pages also collected credit card information as validation proof. Apparently, the tests did well in spite of the design.

Of course, testing distribution first is scarcely a novel idea. Ramon Van Meer, a friend of My First Million interviewers, tested an idea for Soap Opera website by creating Facebook groups, posting content and measuring engagement. He did this way before he built the website, and was almost completely agnostic on what the product should be. Tim Ferriss also recommended cheap testing of muses through Google Ads way back in Four-Hour work week.

So I have always known about this practice, but till now, has not practised this in my building of Glints. We were dead focussed on user interviews, figuring out customer pain points and iterating on the product in our founding days. Nowadays, we have full marketing and growth teams setting up and optimising our distribution channels. So I admit this is a domain and skill set that I am very weak on, and have the briefest of intellectual understanding in. Yet, it has organically come up in so many of my reading materials on company building that it's an insight worth at least an experiment.