top of page

How Long Should I Run a Campaign Before I Judge It?

2 days ago
3 min read

Short answer: Longer than feels comfortable, and on more than one clock. Google's own ads guidance says to judge bidding results over at least 30 conversions, usually a month or longer. Google research also finds that an ad's first four months of returns are matched by the following 20. If you judge at week two, you are grading the setup, not the campaign.


The tension: accountability versus accumulation


Every dollar should earn its place, and owners want proof fast. That pressure is healthy. A bad campaign left running for six months is real money gone.


Advertising pays out slowly, though. The conversions that prove a campaign works trickle in, and the sales it creates keep arriving after most people stop looking. The faster you demand proof, the less of the result you actually see.


What the data shows


The platforms themselves say a month is the floor. Google Ads Help recommends measuring Smart Bidding over "longer time periods that have at least 30 conversions, such as a month or longer." For ROAS (return on ad spend), the revenue each media dollar brings back, the bar rises to 50 conversions. The vendor with every incentive to show quick wins still tells you not to read early results.


Half the return arrives after most reviews end. A 2025 Google analysis drew on a Google and WARC report and Nielsen studies across 11 European markets. It found returns in the first four months equal the returns across the next 20 months. The authors conclude that a short-term focus "could be missing out on as much as half" of potential returns. A 90-day review captures the early returns and misses the rest.


Standard measurement undercounts the slow work. A TransUnion and MMA Global white paper tracked Ally, Kroger, and Campbell's for 9 to 10 months. It used 400,000 to over 1 million households and more than 10,000 surveys. Long-term effects were 1.8 to 6 times greater than short-term effects, and traditional methods undervalued brand marketing's sales impact by up to 83%. These are large U.S. brands, so the multiples will not transfer directly to a local business, but the direction will.


The arithmetic


How long your campaign needs depends on budget and CPA (cost per acquisition), what you pay in media for one lead or sale. Here is a worked example.


Step 1: Monthly budget divided by CPA gives conversions per month. $1,500 divided by $75 is 20 leads a month.


Step 2: Conversions needed divided by monthly conversions gives time to a fair read. 30 divided by 20 is 1.5 months, about 45 days.


Step 3: Add your sales cycle. If a lead takes 30 days to close, revenue data lands around day 75.


Halve the budget to $750 and you get 15 leads a month. That is two months to reach 30, plus the sales cycle, or roughly 90 days before revenue tells you anything.


The pattern: the early verdict


I call this the early verdict: judging a campaign in the window where it is guaranteed to look worst. In the first weeks, the full cost is visible and only a fraction of the return has arrived.


That timing quietly rigs every early review. Cheap, fast, bottom-of-funnel tactics survive it, and anything that builds demand gets cut before it pays. The studies above do not use this term. It is my reading of what they show together.


What this changes


Set the review date before launch. Write down the clock and the threshold before the first dollar goes out. A review date chosen after launch is usually chosen by anxiety.


Judge setup early and results late. In weeks one to three, check delivery, tracking, cost per click and landing page conversion rate. Those break fast and fix fast. Revenue does not belong in that window.


Demand-building work needs its own scorecard. Track branded search volume, direct traffic and how new leads say they heard of you, and review them quarterly. Holding this work to a 30-day return target guarantees you cut it.


The resolution


Accountability and patience stop fighting once each metric gets the right clock. Fast metrics answer whether the campaign is running correctly. Slow metrics answer whether it is working.


The fix is unglamorous. Confirm conversion tracking fires before launch. Calculate the days to 30 conversions from your budget and CPA, then add one sales cycle. Put that review date on the calendar, and avoid major changes until it arrives so you are not restarting the read.


Ask yourself a simple question: Did I decide when to judge my last campaign before it launched, or after it started to worry me?


If you want help setting the right clocks for your next campaign, reach out.


bottom of page