Scope creep rarely announces itself. It shows up as a friendly message on day nine, a „quick tweak“ that becomes a second design direction, or a stakeholder who was never on the kickoff call and now has opinions. The way you onboard a new client in the first five business days decides whether those requests quietly eat your margin or get handled as visible, priced decisions. Everything below runs in a blank spreadsheet.
What Losing Control of Scope Actually Looks Like
Before you fix it, name it. Scope failures on service projects almost always fall into four patterns:
- Undefined inputs. You promised a two-week turnaround but the client still hasn’t sent logins, brand files or the product list. You wait, then compress your own timeline to hit the original date.
- Silent approvals. Nobody said yes and nobody said no. You built on top of a draft that was never signed off, and now the rework is free.
- Unlimited revisions. „One more round“ has no ceiling because you never wrote a ceiling down.
- The phantom stakeholder. A VP surfaces in week four with feedback that contradicts everything approved so far.
None of these are personality problems. They are documentation problems. A client who asks for extra work is behaving normally; the failure is that you had no mechanism to convert the ask into a decision with a price and a date attached.
Step One: Run a Real Intake Before You Onboard a New Client
A signed contract is not a started project. The gap between the two is where most control is lost. Treat intake as its own deliverable with its own deadline.
Open a sheet called Inputs Register with one row per thing you need from the client. Columns: input name, owner (a person, not a company), requested date, due date, status, and which deliverable it blocks. Ten to twenty rows is typical for a mid-sized project. Assets, access credentials, copy, approver names, legal or compliance contacts, and the name of the single person who can say yes.
Then attach a rule and say it out loud in the kickoff call, not buried in a PDF:
If a required input is more than five business days late, the affected delivery date moves by the same number of days. No penalty, no drama — the calendar just shifts.
This one sentence removes the most common source of late-project resentment. You are not punishing anyone; you are making a dependency visible before it costs you a weekend.
Step Two: Define Deliverables, Approval Gates and a Revision Budget
Next sheet: Deliverables. One row per item you will hand over. For each row, write a one-sentence definition of done, the file format, the named approver, the number of revision rounds included, and the approval deadline.
The definition of done is the part people skip. „Homepage design“ is not a deliverable. „Desktop and mobile homepage design in Figma, final copy in place, one hero variant“ is a deliverable, because you can point at it and prove it exists.
Set a revision budget explicitly. Two rounds per deliverable is a common default for design and copy work. Define what a round is, because that definition is what actually protects you:
- One round is one consolidated set of feedback from one named approver, delivered in a single document or thread.
- Feedback arriving in five separate messages over four days is still one round only if it arrives inside the feedback window.
- Contradictory feedback from two people is not your problem to resolve. You return it and ask the approver to reconcile it.
Finally, define what silence means. Pick one and write it into the kickoff email: either work pauses until approval arrives, or the deliverable auto-approves after five business days following written notice. Both are defensible. Ambiguity is not.
Step Three: Price Change Requests Instead of Arguing About Them
You need a threshold that decides, in seconds, whether a request is absorbed or billed. Here is a decision rule you can adopt as-is and tune later:
- Under 30 minutes and no downstream rework: absorb it. Log it anyway.
- 30 minutes to 2 hours, or it touches something already approved: change request at your published rate. Small, fast, written.
- Over 2 hours, or it moves a milestone: change request with a fee and a revised date, approved in writing before any work starts.
Then build a Change Log sheet: date, requester, description, category (absorbed / billable / declined), estimated hours, fee, schedule impact in days, status, and who approved it. The log matters more than the fee. It is the artifact you open in the end-of-project call when someone asks why delivery slipped.
Worked Example: A Six-Week Brand and Website Sprint
Fixed fee $9,000. Two revision rounds per deliverable. Change rate $150 per hour, stated in the contract.
Day 12. The client asks for a second homepage direction after approving the wireframe. That is a new direction, not a revision. You estimate six hours: $900, plus three days on the timeline. You send a four-line change order. They approve it. Either outcome is a good outcome — you get paid or you get your schedule back.
Day 20. They want a button color changed. Ten minutes. You absorb it and log it as absorbed.
Day 28. A marketing lead who was never in kickoff asks for Spanish versions of five pages. Fourteen hours: $2,100 and a new delivery date. You route it to the named approver from your Deliverables sheet, not to the person who asked. The approver declines it and defers to phase two.
At handoff, your change log shows $3,000 of requested work, $900 approved and billed, $2,100 declined by the client, and seven absorbed items totaling roughly two hours. The project delivered three days late for a documented, client-approved reason. Nobody is surprised, and you have a clean record for the renewal conversation — which is exactly the kind of operational detail worth revisiting in the other workflow walkthroughs on the blog.
The First-Week Onboarding Checklist
Run this every time you onboard a new client, regardless of project size. It takes about ninety minutes of setup.
- Send the welcome email within 24 hours of signature, naming the single point of contact on each side.
- Create the project sheet with four tabs: Inputs, Deliverables, Change Log, Decisions.
- Fill the Inputs Register and send it with owners and due dates already assigned.
- Confirm the approver in writing: one name, one email, with authority to approve fees.
- List every stakeholder who will see the work, including anyone who reviews but does not approve.
- State the revision budget and what counts as one round.
- State the silence rule and the input-delay rule.
- Publish the change rate and the three-tier threshold above.
- Book the kickoff call and the two mid-project checkpoints now, on the calendar.
- Log the first Decisions row: scope as understood on day one, in your own words.
That last item is underrated. When a dispute happens in week five, the day-one summary written in plain language usually settles it faster than the contract does.
Where a Spreadsheet Stops Being Enough
Be honest about the ceiling. A spreadsheet is excellent for defining rules and terrible at enforcing them. It has no signature trail, so a change order approved by email lives outside the file. It handles one editor well and several editors badly. It does not notify anyone when an input goes overdue.
Move to dedicated tooling when you hit any of these: more than about five concurrent projects, two or more people editing the same tracker daily, retainers with rolling unused hours, or a client who requires a shared portal. At that point you want proposal software with e-signature for change orders, a project tool with task-level owners and due-date alerts, and time tracking that feeds your change log automatically.
Even then, keep the sheet for the first two weeks of every engagement. The rules are the asset; the software is just where you store them later.
If you would rather start from a built version than a blank file, the Client Onboarding & Scope Control OS packages this workflow as a PDF guide plus an Excel system with the intake register, deliverables sheet and change log already wired together. You can also browse the rest of the operations workbooks in the shop if pricing or pipeline is the tighter constraint right now.