Every Agency OS engagement has the same maximum length. 14 weeks, from kickoff to shipped. We do not quote 16. We do not quote 20. The work either fits, or we change what the work is until it does. Here is why the cap exists, what we cut to keep it, and what we refuse to take on at all.
Why a hard cap exists at all
- Open-ended timelines are how good projects die. The first month is productive, the second is fine, the third drifts, and by the fifth nobody remembers what shipping looked like.
- The cap forces a brutally honest scoping conversation up front. If we cannot fit your project in 14 weeks, we would rather find out on day one when we can cut features together, not week ten when we can only apologize.
- Constraints are a forcing function. They push the team toward sharper trade-offs and away from “we will figure it out later.”
What we cut
- “Just one more thing” requests after week 8. They go on a phase-two list, not into the current build. Phase two is real and it gets shipped. It just does not get shipped this quarter.
- Custom admin tooling that nobody outside the company will ever see. Use a CMS. Use Retool. Use a spreadsheet. Build the customer experience first.
- Edge-case features with usage projections under 5%. They eat 30% of the timeline and add 0% of the value. We will build them if you actually need them later.
- Speculative configurability. Build for the use case you have, not the 12 use cases you might have someday.
What we refuse outright
- Multi-vendor coordination roles where we own none of the code. If we are not building, we are not the right team.
- Critical-path dependencies on other vendors. If our progress depends on another team’s release schedule, the timeline stops being ours to keep.
- Anything where the success metric is “the client is happy.” Happiness is downstream of shipping. Shipping is downstream of clear metrics.
How we keep the promise
- We scope the smallest thing that ships. Then we add. Never the other way around.
- We use the same stack on every project. No reaching for the new shiny thing in the middle of an engagement. Tech choices happen in the proposal, not in week 6.
- One senior engineer owns the build. No subcontractors, no offshore relays, no hand-offs. The person who starts the project ships the project.
- We ship something every week. A staging URL, a new flow, a working prototype. If a week passes without something to show, we have a problem we need to address that week.
- Week 12 is “stop adding.” The last 2 weeks are for hardening, not for new features. No exceptions.
The compounding effect we did not expect
Hard timelines have a side effect we did not see coming when we started. They attract the right clients. People who have been burned by 9-month projects recognize the discipline immediately. People who want infinite scope self-select out before we even take a call. The deals that close are the ones where the constraint is a feature, not a bug.
That is the rule. Fourteen weeks, every time. The constraint is the product.