Outgrowing Us Is the Plan

Most agencies pretend the journey ends with them. Ours doesn’t, and we’d rather say it out loud than have it become an awkward conversation in month nine.

918 builds MVPs in the $25k–$80k range, in 6–12 weeks. That’s the right shape for the founder who needs a working v1 in front of users, investors, and themselves. It is not the right shape for the company that v1 turns into.

If your MVP works — actually works, with real users doing real things in it every day — you will outgrow us. That isn’t a failure mode. It’s the plan.

What “outgrowing” actually looks like

The clearest signal is hiring. Once a founder is interviewing engineering candidates seriously, the relationship with us changes shape. We stop being the team that builds the thing and start being the team that holds the thing while a real engineering org gets stood up. That’s a different job, and it has a different end date.

Other signals we watch for:

• A roadmap that no longer fits inside a fixed-scope SOW. When the work coming up next quarter looks more like “platform decisions” than “ship this feature,” fixed-scope billing stops being the right structure. You need a team you can redirect mid-sprint, and you need to own that team’s calendar.
• Infrastructure decisions that outlast us. The day the question “should we move off Supabase?” stops being theoretical, you need an in-house person who’s going to live with the consequences. That person should make the call. Not us.
• A second product or business line. We’re great at the first one. We’re explicitly not the right team for the third one — by then you should have your own people who know the codebase from the inside.

What the handoff actually is

We do not throw a tarball over the wall and wish you luck.

A clean handoff has four moving parts, and we do all four.

1. The codebase. We’ve been writing yours for the engineers you haven’t hired yet. That means real PR descriptions, runnable tests, README files that assume the reader doesn’t know us, and architecture decisions documented in the repo instead of in a Slack thread. Nothing about your codebase requires a 918 person to translate.

2. The accounts. Supabase, Vercel, GitHub, Resend, Twilio, OpenAI, the domain registrar. All of it gets handed over to your account, billed to your card, with seat-level access for the new team. We remove ourselves cleanly. No dependency, no awkward “log in as us” workaround.

3. The partner team. When you tell us you’re past us, we don’t make you find your own next agency or first hire from scratch. We have a short list of partner teams we trust — bigger shops, fractional-CTO networks, design studios that specialize in scale-stage work. We make the intros, sit in the first meeting if you want, and make sure whoever takes over has the full context.

4. The runway. We make ourselves available for code review and architecture questions for 30 days at no extra cost, then on a clearly-priced retainer if you want to keep us in the loop while the new team gets up to speed. Most of the time, the retainer isn’t needed. Sometimes it is. Either way, you decide.

Why we don’t try to keep you

Plenty of agencies do. The math is obvious — keeping a happy client is cheaper than landing a new one, and there’s always a next phase you can sell into.

We’ve watched what happens to clients who get talked into staying past the right exit. The work gets soft. The accountability erodes. The agency starts charging hourly because the milestones stopped making sense. The client starts wondering if they’re being managed.

We’d rather lose the LTV than be that agency. The studio reputation we want is “they shipped what we needed and helped us scale past them when it was time.” That’s the case study we want on our home page in three years. The version where we held on too long is the case study we’d never write.

What this asks of you, the founder

Two things.

One: tell us early. The handoff is much cleaner when it’s planned six weeks out instead of triggered by a hiring offer that closes Friday. If you can see the inflection coming, name it. We’ll start documenting harder, structuring our PRs more, and pulling our patterns into something a new team can read in a week.

Two: don’t apologize for it. We’re not hurt. We’re not going to soft-pitch you on staying. The whole reason we built the studio this way — small, senior, scoped — is that we wanted clients who outgrow us. That’s the customer base of a studio that does its job.

The bridge is the product

When we describe what 918 does, we usually lead with “we build MVPs in 6–12 weeks.” That’s true. It’s also incomplete.

The full version is: we build MVPs in 6–12 weeks, and we build the bridge to whatever comes next, and we hand over a thing that doesn’t need us to keep working. The bridge is part of the deliverable. So is the day we wave goodbye.

If your idea works, you’ll outgrow us. We’d love that for you. Plan on it.