The scorecard
| Measure | Why it matters | Set a threshold before launch |
|---|---|---|
| Cost per click | Keeps spend honest | Within your planning range |
| Identified companies | Proves the right companies came | A number per month |
| Companies that fit your profile | Separates reach from relevance | Share of identified companies |
| Meetings | First sign of real interest | Number in 90 days |
| Pipeline created | What the test is for | Value that pays back the test |
Why clicks alone mislead
A published B2B pilot reported 259 clicks at $5.68 each and zero tracked conversions. The clicks may have been valuable; nothing connected them to pipeline, so nobody could tell. Clicks show that the ad was seen; companies and meetings show whether it worked.
A weekly review in 20 minutes
- Spend and cost per click by ad group.
- Identified companies and how many fit your profile.
- Which buyer questions brought them.
- What sales did with last week's companies.
- One change: move budget, rewrite an ad or pause an ad group.
The day-90 decision
- Scale when fitting companies and meetings are growing and pipeline is on track to pay back.
- Fix when the right companies come but don't convert: change the landing page or the sales follow-up.
- Stop when the companies that come don't fit your profile after two rounds of new questions.
A filled-in example
Fictional company, $3,000 a month of media in the US.
| Measure | Threshold set at launch | Day 90 result | Verdict |
|---|---|---|---|
| Cost per click | Under $6.50 | $5.10 | Pass |
| Companies identified per month | 60 | 85 | Pass |
| Share that fit the profile | 30% | 34% | Pass |
| Meetings in 90 days | 6 | 9 | Pass |
| Pipeline | $90,000 | $120,000 | Pass: scale |
Had meetings stayed at two despite good companies, the verdict would have been "fix the follow-up", not "stop the channel".
Who should sit in the review
Marketing owns the numbers; sales owns the meetings; whoever owns the budget makes the call. Thirty minutes, one decision.
Questions
What if results are mixed?
Keep the ad groups that pass, drop the rest and run another 60 days on the winners. Mixed results usually mean some questions work and others don't.
Should we compare with other channels?
Yes, on cost per in-profile company and cost per meeting, using the same definitions for every channel.
What if 90 days is too long for our board?
Report at day 30 and 60 on leading measures, companies and fit, but keep the decision at day 90.
Start here: ChatGPT ads for B2B. More articles in ChatGPT ads.
