Short answer
Measure CIA at three levels. First the question: click rate and cost per click by ad group. Then the account: companies identified, the share that fits your ideal customer profile, and target accounts reached. Then the pipeline: meetings, opportunities and revenue by question. Compare CIA with LinkedIn ABM and programmatic on cost per in-market target account and cost per meeting, not on CPM or CPC.
Why CIA needs its own scorecard
Search metrics judge keywords. ABM metrics judge account coverage. Conversational Intent Advertising sits between them: it's bought by question and valued by account. A scorecard that only counts clicks makes CIA look like expensive search. One that only counts account reach makes it look small. You need both, plus pipeline.
Level 1: question metrics
| Metric | What it tells you | Early reference point |
|---|---|---|
| Click rate by ad group | Whether the ad answers the question well | 0.66% in the one published B2B pilot |
| Cost per click by ad group | What the question costs to win | About $3–5 reported; we plan B2B at $4.50–6.50 |
| Spend share vs budget share | Whether the question is asked often enough | Below half its share for two weeks: widen the phrasings |
| Page depth after the click | Whether the page continues the conversation | Compare across ad groups, not with other channels |
Benchmarks for this channel are thin. Treat your first 4–6 weeks as your own baseline.
Level 2: account metrics
| Metric | Definition |
|---|---|
| Identification rate | Share of clicks matched to a company |
| Companies identified | Unique companies named from CIA clicks |
| ICP fit rate | Share of identified companies that fit your ideal customer profile |
| Target list hits | Identified companies that are on your target account list |
| Cost per fitting company | Media spend divided by companies that fit your profile |
| Question-to-account map | Which questions bring which kinds of companies |
The last one is the metric unique to CIA. It tells you which problems your best-fit accounts describe, which helps sales, content and product marketing as much as advertising. More on company match rates and ICP scoring.
Level 3: pipeline metrics
- Meetings with companies first seen through CIA
- Opportunities and their value, by question and by intent tier
- Influenced pipeline: open opportunities whose account clicked a CIA ad before or during the deal
- Time to opportunity from first CIA visit
Pass the click identifier and the ad group into your CRM with every form fill, so a deal can be traced back to the question. How to set up conversion tracking.
Comparing CIA with LinkedIn ABM and programmatic
Each channel reports in its own language. LinkedIn reports members and engagement. Programmatic reports impressions and account reach. CIA reports clicks and questions. Put them on the same denominators:
| Common metric | LinkedIn ABM | Programmatic ABM | CIA |
|---|---|---|---|
| Cost per target account engaged | Engaged accounts from LinkedIn reporting | Accounts with clicks or visits | Target list hits |
| Cost per in-market account | Needs intent data added | Needs intent data added | Fitting companies from vendor and decision questions |
| Cost per meeting | From CRM | From CRM | From CRM |
| Pipeline per $1,000 of media | From CRM | From CRM | From CRM |
Comparisons in depth: CIA vs LinkedIn ABM and CIA vs programmatic.
The 90-day CIA scorecard
| Result after 90 days | Decision |
|---|---|
| Fitting companies at or below your cost per account on other channels, and meetings booked | Scale: add questions and budget |
| Fitting companies, but no meetings yet | Fix: check routing, follow-up speed and the decision-tier pages |
| Clicks but few fitting companies | Fix: tighten context hints and move budget to the vendor and decision tiers |
| Little spend, few clicks | Check that buyers ask AI about your category; widen phrasings before widening topics |
| None of the above after fixes | Stop, and keep the question map for AI visibility |
These are the rules we use, not industry standards. The full version: how to judge a ChatGPT ads test.
Measurement pitfalls
- View-through conversions. Report them separately from click conversions. View-through conversions explained.
- Modelled conversions. Platform totals can include modelled figures that won't match your CRM.
- GA4 differences. Analytics and the ad platform count differently. Why the numbers differ.
- Dark AI traffic. Many AI-driven visits arrive as direct traffic. How to spot it.
- Unmatched visits. Home and mobile visits are hard to name. Count them as demand anyway.
- Too short a test. Conversion bidding needs tracked conversions before it can learn. Plan for 90 days.
Questions
What's the single most important CIA metric?
Cost per fitting company from vendor and decision questions. It joins what you paid, the intent of the question and whether the account is one you want.
Can we measure CIA without company identification?
Partly. You'd have clicks and form conversions, but you'd lose the account level, which is where most of CIA's B2B value shows up.
How do we report CIA to the board?
Use the same three lines for every channel: target accounts engaged, meetings, and pipeline per $1,000 of media. CIA then sits on the same page as LinkedIn and programmatic.
Facts checked against the sources below on 14 October 2026. AI ad platforms change quickly: if something here no longer matches what you see, tell us and we will update it.
