Why We Run Every Client Project on Basecamp
Every client we work with gets one Basecamp project on day one. Messages, to-dos, documents, files, and the schedule all live there. There is no shared Slack workspace, no side email thread that only two people saw, and no "where did we decide that?" hunt three weeks later. If it matters to the project, it is in the project.
We have shipped 50+ products over 8+ years, and the tooling lesson we keep relearning is that the problem is rarely the tool. It is the number of places a decision can hide. Below is how we set up a client project, which ideas we borrow from Basecamp's Shape Up method, and which parts we deliberately skip.
One place per client, and the client sees it
Each client project in Basecamp has the same layout: a Message Board for decisions and updates, To-dos grouped by scope, Docs and Files for specs, designs, and credentials handoff notes, and a Schedule with the milestones we agreed. The client is invited on day one with full visibility. They see the same to-do lists we work from, not a cleaned-up version.
That last point changes behavior on both sides. We write to-dos that a non-engineer can read. The client stops asking for status because they can see it. And when a scope question comes up, the history is right there: who asked, what we decided, and why.
The rule we hold ourselves to is simple. If a decision happened on a call, in a DM, or in a hallway, it does not exist until someone posts it to the Message Board.
Async by default, live when it earns its time
Most project communication does not need everyone online at the same moment. A design question can wait two hours. A scope tradeoff needs a written option list, not a 45-minute call where three people think out loud. So our default is written posts and short recorded videos, and we keep one weekly live call for the things that really do need a conversation.
Async also suits how we build. Our agents and engineers work in focused blocks. A chat tool that pings all day breaks those blocks into pieces. Basecamp lets people catch up on their own schedule, and the client gets a clear daily or weekly picture without having to attend anything.
This is not a rule against talking. When something is stuck, ambiguous, or emotional, we get on a call quickly. Then we write down what we agreed.
Fixed appetite, flexible scope
The most useful idea we took from Shape Up is appetite. Instead of estimating how long a feature will take and watching the estimate slip, you decide up front how much time it is worth, then shape the work to fit. The time is fixed. The scope flexes.
Our offers are appetites. The AI MVP is 15 days for $4,500. Iteration sprints after launch are one week at $1,500. We do not quote "about three weeks, maybe four." We agree on what the product must do for its first paying users, and when something threatens the deadline we cut or simplify scope, with the client, in writing, in Basecamp.
This only works if the client knows the tradeoff before it happens. So the first post in every project lists the must-haves, the nice-to-haves, and the things we have already agreed to leave out.
Want this built, not just scoped?
A live AI MVP on your domain in 15 days for $4,500. Then $1,500 a week sprints.
Scopes and hill charts for honest progress
We break each build into scopes: chunks of work that can be finished and shown on their own, like "checkout flow," "admin export," or "agent escalation path." Each scope gets its own to-do list. That gives the client a map of the product, not a pile of tickets.
Each scope also goes on a hill chart. The uphill side is figuring out: unknowns, design choices, integration questions. The downhill side is execution: we know exactly what to do and are doing it. A scope stuck on the uphill side for two days is a warning sign, and we say so. A percentage bar would show "60% done" on something that has not been figured out yet. The hill shows the truth.
When a client asks "are we on track?", we point at the hill. If three scopes are downhill and one is still climbing, the conversation is about that one scope, and usually about cutting part of it.
Cool-down after each cycle
Shape Up builds in a cool-down period after each cycle for bug fixes, cleanup, and deciding what comes next. We use a smaller version. After the 15-day MVP launch, we take a few days before the first iteration sprint to watch real usage, fix what real users hit, and write the next sprint's shortlist with the client.
The same applies between weekly sprints when a client pauses. Pausing is normal. The project stays in Basecamp, the history stays intact, and restarting means reading the last few posts, not a kickoff meeting.
What we do not use from Shape Up
We do not run six-week cycles. Our clients are founders and small teams who need something live in weeks, so our cycles are 15 days for the first build and one week after that. We also do not run a formal betting table. With one client per project, the "bet" is a short written plan the client approves.
We skip most of the formal shaping documents too. A pitch with problem, appetite, solution, rabbit holes, and no-gos is a good structure, and we borrow it loosely for the first post. But for a one-week sprint, a short list and a quick sketch is enough. Writing a full pitch for a three-day scope is process for its own sake.
Finally, Basecamp is not where code lives. Code, pull requests, and CI stay in the repository. Basecamp holds the decisions about the code, not the code itself.
Key takeaways
- One project per client, with messages, to-dos, docs, files, and schedule together, and the client invited from day one.
- Decisions made anywhere else get posted to the Message Board, or they did not happen.
- Our 15-day MVP and $1,500 weekly sprints are fixed appetites. Time is fixed, scope flexes, in writing.
- Hill charts show what is still being figured out, which a percentage bar hides.
- We borrow Shape Up ideas, not the whole method: no six-week cycles, no betting table, light shaping docs.
Frequently asked questions
No. We run the account and invite you as a client. You get full access to your project at no cost to you.
We can join a shared channel for urgent pings if your team lives there, but decisions and deliverables still get posted in Basecamp. One source of truth is the point.
We tell you early, usually when a scope stalls on the uphill side of the hill chart. Then we agree on what to cut or move to the first iteration sprint. We do not quietly extend the deadline.
Yes. We can export the project or keep you on it, so the decision history and documents stay available to you and your next team.
The 15-Day AI MVP Playbook
Day-by-day plan, a starter CLAUDE.md, the Lovable-to-production checklist, eval and cost worksheets, and 10 prompts. Pure value, no pitch.
Book 20 minutes, or send a brief
A former CTO replies. You leave with a written plan and a fixed estimate.
Prefer LinkedIn? Connect with Harish.