Original Summary

Back when we mostly built MVPs, we took on a fixed price build for a B2B inventory tool. Quoted 10 weeks and it took 19 lol. The founder kept asking for small stuff and we kept saying yes because every request sounded like an afternoon. A filter on the orders page, a Google login since "everyone has it now" When I finally went through the tickets there were 63 change requests. None of them took more than a day on its own which is exactly why we never pushed back. Together they added up to about seven weeks of extra work and the founder was annoyed anyway because the launch kept moving. We've run the agency for 8 years now and that project is why every fixed price build since has had this clause. Here it is close to word for word: "Any change requested after scope sign off gets sized in working days and we send that in writing within one business day. The client then picks one of two options. Swap it in for work of equal or larger size from the current scope or put it on the Post Launch List to be scoped and priced after launch. The price and the launch date stay the same either way." Scope creep almost never shows up as one big request. Big requests always got a proper conversation because no one adds a whole module without talking about money. The two hour requests got a "sure" because pushing back on something that small feels petty. 60 of those is a different product with the same price tag. With the clause, even a half day change means picking what comes out. On a recent build (a booking platform for physio clinics) the founder asked for team accounts 5 weeks in so several people from one clinic could share a login. We sized it at six days, mostly because permissions touch everything. We sent back two options: swap out the in app chat which was about the same size or park team accounts on the post launch list. He parked it. Six weeks after launch, 2 of their first 40 clinics had asked for team accounts. It's still parked. What clinics kept asking for was SMS reminders which no one had mentioned during the build. I didn't see that coming but the post launch list turned into a filter. A lot of what felt urgent in week five, real users just didn't care about. The standard advice is to bill change requests as change orders. We tried that after the 19 week mess and I'll probably get pushback for this, but I think it's worse than eating the cost. Every small change turns into a mini negotiation and an invoice, so the founder feels nickel and dimed and the agency starts getting paid for delays which flips the incentive. If an MVP launches 3 months late, no one cares that the extras were billed properly. On an MVP the launch date matters more than the money for both sides. Most founders end up liking it once they've been through it. One told me it was the first time a dev shop had made him choose and choosing made him figure out what the product was actually for. Not everyone though. The fight

  • 情报分类：服务器与云资源
  • 分类依据：内容涉及服务器、云资源或网络线路
  • 信息来源：Reddit · SaaS
  • 发布时间：2026/10/3 23:31:37