What counts as a good customer acquisition cost for an app

Consumer app-install cost runs into dozens of dollars per install because the app store sits between the app and everyone who might download it. Alex Frenkel, whose app Kai reaches people through messaging instead, has said ordinary installs cost that much, while a contact on WhatsApp, Telegram or Discord costs a fraction, won mostly with word of mouth.

A low acquisition cost tells you new users are cheap to reach, not that the business works. Frenkel's app Kai pulled in hundreds of thousands of users at almost no cost and still had to pivot, because those users would not pay enough to cover what came after the install. Growth that cheap can hide the fact that nobody is paying.

Marketing KPIs, the numbers a team is judged on, move with the way a company sells, not the calendar. In Kai's direct-to-consumer phase the number that mattered was week-over-week signups. Frenkel told Etgar Shpivak that after the pivot to clinics and employers, marketing was judged on high-quality leads that convert to paid pilots. Retire the old KPI when the model changes.

Lifetime value, how much a customer pays you over time, is the problem: the cheapest users to acquire are often the least willing to pay. Frenkel told Etgar Shpivak that consumers used Kai without funding it, and that the buyers willing to pay were organizations, insurers and employers purchasing a human service. A channel full of people who never pay is a marketing win and a business dead end.

Hire a senior marketing executive once the team, budget and headcount they need already exist, rather than to create them. Frenkel has watched founders bring in an expensive, experienced leader too early, before anything was in place, and a senior hire with no team to run cannot deliver what made them worth hiring. Knowing when a first marketing hire is right comes first.

An ad platform reports the average cost across your cheapest and most expensive users, so the headline figure flatters you and climbs as you scale. Etgar Shpivak's advice is to build a deliberately pessimistic model and watch what the cost does when you spend more. Reading incremental cost on Facebook and Google is the discipline that catches it.

A subscription app can absorb a far higher acquisition cost than a one-off purchase, as long as lifetime value covers it within a reasonable window. There is no fixed threshold, only a ratio: what a user returns over time against what they cost to acquire. A near-free channel is a trap only when lifetime value stays thin.

The first payment, and how many users ever reach it. A cheap install reassures too early, because it measures how easily you add people, not how many turn into revenue. Set a user's cost beside what they return over a lifetime. When very few users reach that first payment, a cheap channel does not save the model.

A large user count proves the product works, not that it can earn. Kai's hundreds of thousands of free users showed demand but predicted almost no revenue. The number that decides it is how many reach the first payment. Before scaling a near-free channel, check whether each user returns what they cost, or you have built a crowd, not a business.

An early-stage app founder should bring in help the moment the growth curve looks great and the revenue refuses to follow. That is when unit economics, what a user costs against what they pay back, get decided. A consultant can set the cheap install against what a user returns before the model hardens. Etgar Shpivak, a marketing consultant who works with seed and Series A founders, does this directly with the founder; his approach is on his bio page.

Kai runs as a contact inside messaging apps like Apple Messages, WhatsApp, Telegram and Discord, instead of competing for downloads in the app store. Users arrived from Discord communities, TikTok and Snapchat through word of mouth, on almost no budget. That drove install cost to a fraction of the store's rate, but it only works if those users eventually pay.

Because its hundreds of thousands of free users paid almost nothing, and lifetime value stayed thin while the user count climbed. Kai moved to selling clinics and employers, where contracts run into hundreds of thousands of dollars. As the founder put it, what organizations will pay for is a human team. The user count proved the product worked, not that it could earn.