Context
The site itself wasn’t complex: a single landing page on an existing brand identity, no CMS, branding and copywriting both out of scope. The complexity was that sales had already promised a two-week delivery before I saw the project. The real scope (branding alignment, content gathering, concepting, wireframes, design, revisions, development) didn’t fit in two weeks at any reasonable quality bar. Fixed price, no dedicated QA: the team owned quality collectively.
The stakes
If I said nothing and let the team try to hit the sold date, one of two things happened: we shipped something rushed, or we missed the date the client had already planned around. She needed the site live for an industry conference. Either way, the first thing she’d learn about working with us was that we don’t do what we say. That’s a harder hole to climb out of than a delayed kickoff.
What I did
Reset the timeline before a single deliverable was built. I went to the client directly and walked through what the work actually required, week by week. She wasn’t happy. Her reaction was close to “I’d have gone to another agency.” I didn’t retreat from the number, gave her the reasoning behind it, and offered a scoped-down MVP she could still bring to her conference. She agreed to three to three and a half weeks. Nobody was building against a fiction anymore.
Made “what we need from you” the first line of every update. She ran her own business and disappeared for stretches, including during conference prep, the exact reason she needed the site. I shared the full roadmap in Asana on day one and structured every update, by call or Loom, around four things: done, in progress, needed from you, next. When she delayed sending one email address the contact form needed, then asked a day later if the site was ready, that gap was exactly what the update structure existed to prevent. After that, I led every message with what was blocking us.
Planned around her availability instead of reacting to it. I front-loaded as much decision-making as possible so later stages wouldn’t stall. When she still went quiet, the designer and developer shifted to other active work rather than sitting idle, and I sent a follow-up showing exactly what was done and where we’d stopped. That visible stop-point was usually what brought her back within a day.
Protected margin by capping revisions before they got expensive. With branding and copy already out of scope on a fixed price, every extra revision round ate directly into margin. Every design review had one clear objective, with feedback consolidated before it went back to the designer.
Outcome
- Delivered on the revised three-to-three-and-a-half-week timeline
- Design approved in two review cycles; development needed zero additional revisions
- Held roughly 50% margin despite a tight fixed-price budget and client-side delays
- Client always knew where the project stood, what depended on her, and what came next
What I’d do differently
I’d raise a mis-sold timeline this early on every project. Nobody asked me to challenge the two-week estimate before this one. I could see it wouldn’t hold, so I flagged it myself. Waiting for someone else to catch it first costs more than the awkward conversation does.