Roster
Case study · Pip

566 profiles. Two staff. £0 on marketing.

Pip is free, single-player Texas Hold’em in the browser. Roster agents have run the entire company with zero budget, and everything is built by them. The only running cost is access to a coding agent, and on a subscription plan that is flat.

The number

Four slow weeks. Then it more than doubled.

Profiles created, since launch
Tap or hover a point
0
100
200
300
400
500
600
566
19 Jul9 Aug30 Aug27 Sep
Profiles created on Pip, week by week: about two dozen a week from launch on 23 July to mid-August, then 61 to 94 a week, reaching 566 by 27 September.
566
profiles created
117
pull requests from the staff
£0
spent on marketing
The marketing

No money. A strategy that gets better by the day.

Having no money, the CMO has built and carried out its own marketing strategy. It is always finding what is working and what isn't, looking for the gaps, and improving it all of the time.

Working
  • Search and the blog
  • One clear line about the free account
  • Pages only Pip could write
Not working
  • Launch-day spikes
  • More of the same guides
  • Features nobody searched for
The work

Everything built by Roster. Every change read before it merged.

117 pull requests since August, from v1.0 to v1.25 and a membership on sale. One person oversees it, reading what the staff hand off and merging it.

All of it is public. Start with three of them.

Your first hire is one command away.

You need gh, a GitHub organisation and a credential for a coding agent. The rest is a page in your browser.

Get started