Month one: build and learn
What happens: accounts audited or built, tracking implemented and verified, campaigns launched, negative keyword lists started, first creative in market.
What to expect: your worst cost per lead of the whole engagement. Platforms are learning, waste has not been pruned, and volume is uneven.
What to judge: is tracking working, are leads arriving at all, and did everything launch when promised. Nothing else.
Month two: prune and shift
What happens: search terms cleaned, geography tightened, budget moved toward what converts, second creative round, first real optimization.
What to expect: cost per lead down 20 to 40 percent from month one. Lead quality noticeably better as the waste comes out.
What to judge: direction of travel and whether the specific changes make sense to you when explained.
Month three: the baseline
What happens: the account stabilizes and a repeatable cost per booked job emerges.
What to expect: this is your real number. Everything after is incremental improvement on it.
What to judge: is the number viable against your job value, and is the trend flat or improving?
What both sides owe
From the provider: launches on the promised date, changes explained in plain language, bad news first, and a report you can read in two minutes.
From you: a complete brief, assets when asked, one named approver rather than a committee, and someone answering the phone. Half of all underperformance in the first ninety days is on the client side of that list — usually the phone.
When to actually worry
Tracking still unverified after week two. Nothing changed in the account in thirty days. Reports leading with impressions. A missed launch date without a call. Or cost per lead flat from month one to month three with no explanation.
Any one of those is worth a direct conversation. None of them requires waiting for month six.