buildBetterAds
← WritingSTRATEGY15 JUL 2026 · 6 MIN READ

What actually takes two weeks in campaign setup

We timed every step of a manual campaign build. The surprising part isn’t the keyword research — it’s the waiting.

TL
Thea Lindqvist
Performance lead

“Two weeks to launch” is the standard quote for a new paid search and social build, and it is roughly accurate. What it hides is that very little of the fortnight is spent building. We logged the steps on our own builds to find out where the time actually goes.

The pattern is consistent: the hands-on work is measured in hours, and the calendar time is almost entirely made of waiting for information that only the client has.

Where the time actually goes

Keyword research, structure and copy are the visible work, and the part everyone budgets for. In a straightforward single-location build they are an afternoon, not a fortnight. The fortnight comes from the gaps between them.

A typical build, by elapsed time
1Kickoff call and access requests — 1 day of work, then 2 to 4 days waiting for account access.
2Site and offer review, claim gathering — half a day, then a wait for the answers only the client has.
3Structure, keywords, negatives, copy — under a day of actual work.
4Tracking and conversion setup — hours, unless the tag is broken, in which case it becomes the critical path.
5Internal review and client approval — minutes to read, days to schedule.

Almost nothing in a two-week build takes two weeks. The calendar is made of handoffs, and every handoff is a queue.

The waiting is the product problem

This reframes what automation is for. Making keyword research instant compresses an afternoon. Removing a handoff compresses days, because a handoff costs whatever the other party’s response time happens to be.

The two handoffs worth attacking are the evidence-gathering round trip and the approval round trip. The first can largely be pre-empted by reading the client’s own site before asking anything, so the questions that remain are the few a website genuinely cannot answer.

What compresses and what doesn’t

Approval does not compress, and should not. What compresses is the cost of approving: if the reviewer has to reconstruct the reasoning themselves, the review needs a meeting; if the reasoning is attached to the plan with its sources, it becomes a read they can do between other things.

Platform-side waits do not compress either. Account access, billing verification, domain verification and policy review run on someone else’s clock, which is an argument for starting them on day one rather than at the point they block you.

Strip out the queues and a two-week build is a two-day one, with the same amount of thinking in it. The thinking was never the bottleneck.

Want this framework applied to your site?

Build a campaign and read the recommendation with sources. Nothing launches without you.

Create campaign →