A portfolio of small apps

I'm running three Shopify apps at the same time. sonarID flags the notable people buying from a store. OrderEdit lets customers fix their own orders before they ship. OrderSurvey does post-purchase surveys on the thank-you page.

None of them is the company. All of them together aren't the company either. That's the point, and it's a deliberate change from how I've operated for the last decade.

Why not one thing#

Every company I've built before this was singular. All the attention on one product, one market, one bet. That's the standard advice and there's a good reason for it — focus produces better products and concentrated effort is how anything great gets made.

It also means the outcome is binary and the variable that decides it is frequently not you.

Exposely was right about the category and three years early. Vigilance was working and a file got deprecated. In both cases the quality of my execution was largely irrelevant to the ending. I don't say that to excuse anything — LiveRecover and OrderBump worked, and execution mattered there. But I've now had two companies end for reasons that were structural rather than performance-based, and the pattern is hard to unsee.

Three small products means three independent exposures. One dying is a bad quarter, not a shutdown.

What actually makes this possible#

This structure would have been absurd in 2018. A few things changed:

The platform does the distribution. All three are in the same app store, discovered through the same search, billed through the same system. I'm not building three go-to-market motions. I'm building three listings in a channel I've spent seven years learning.

The customer is identical. Same merchant, same store, same problems at different points in the order lifecycle. What I learn selling one is directly usable for the others, and a merchant who trusts one of them is a warm audience for the next.

Small products can actually be finished. Each of these does one thing. There's a real state where a narrow product is done and needs maintenance and support rather than a roadmap. A company always needs a roadmap, because a company has to grow into something. A product doesn't.

I've done this long enough to move fast. Sixth company. The scaffolding is not a research project anymore.

The honest costs#

None of them will be large. A narrow app for a specific problem has a ceiling and I can see it from here. Three ceilings added together is still a modest number compared to what a single focused company could theoretically be. I'm trading the top end for the floor.

Attention divides worse than linearly. Three products is more than three times the context switching. Some weeks each one gets a third of a person and none of them get momentum.

It's a bad story. "I run three small apps" doesn't recruit anyone, doesn't raise anything, and doesn't sound like ambition. I notice myself explaining it defensively, which is usually a sign that I care more about how it sounds than I'd like to admit.

Where this actually sits#

The real answer is that this is the shelf, not the plan. Internet Research Unit is the thing I think could be large, and it's the one with no distribution and a long, slow first act — the same painful stretch I spent three years in at Exposely.

The apps fund the patience. They're built on distribution I already understand, in a market I already know, doing work I can do well without a great deal of thought. That buys time for the thing that needs time.

That's a less inspiring description than a portfolio strategy and I think it's the accurate one.