There is a structural conflict at the centre of account-based advertising that almost nobody states plainly: the metrics the ad platform gives you reward the opposite of what your strategy requires.
Click-through rate rewards broad appeal. It goes up when an ad speaks to more people. Account-based marketing is a bet that a small number of specific companies are worth disproportionate attention, which means your creative should be speaking to a deliberately narrow group — the four or five people at each target account who own a particular expensive problem.
An ad that does that well will have a mediocre CTR, because most of the people who see it correctly conclude it is not for them. If you optimise on CTR, you will systematically kill your best ABM creative and replace it with something that entertains a large audience of non-buyers.
The metric conflict, made concrete
A real shape of result, from a programme running five creatives against a 210-account list:
| Creative | Share of clicks | Accounts moved a stage | Cost per movement |
|---|---|---|---|
| Ungated benchmark report | 44% | 6 | $1,840 |
| “What a failed migration costs” | 12% | 19 | $290 |
| Architecture guide (document ad) | 9% | 14 | $355 |
| Customer quote card | 21% | 4 | $2,110 |
| Product tour video | 14% | 7 | $720 |
Illustrative figures from a composite programme.
The benchmark report was excellent at what it did. Data engineers everywhere clicked it, most of them at companies that were not on the list and were never going to buy. It behaved exactly as a good ungated resource should: it attracted an audience.
The migration-cost ad was uninteresting to almost everyone. It named a specific, expensive, slightly embarrassing problem. The handful of people at each account who own that problem stopped, and several of them at the same company stopped within the same fortnight, which is the event the whole programme exists to produce.
An ABM ad should repel most of the people who see it. If yours does not, check who is clicking before you celebrate.
What specificity actually looks like
"Be specific" is unhelpful advice on its own. Four concrete moves that produce it:
Name the cost, not the benefit
"Cut deployment time by 40%" is a benefit claim, and every competitor makes one. "A rolled-back migration costs about six engineer-weeks and the trust of your ops team" names a cost the reader has personally paid. Costs are more specific than benefits because they are remembered.
Use the vocabulary of the role, not of the category
Category vocabulary — platform, solution, streamline, unlock — is legible to everyone and therefore selects for no one. Role vocabulary is the language that appears in the internal document your buyer wrote last month. If your ad could be run by a competitor with the logo swapped, it is written in category vocabulary.
Address the internal argument, not the purchase
At most target accounts the real obstacle is not "should we buy something" but "can I convince the three other people who have to agree". Creative that arms your champion for that argument — a comparison, a cost model, a migration risk breakdown — works harder than creative that pitches you.
Make the offer proportionate to the stage
A demo request in a first-touch ad asks for a commitment the reader has no basis for. A demo request in a bottom-funnel ad, shown to an account where four functions have engaged over six weeks, is exactly right. The same offer is wrong and right depending on position.
Testing against movement
If you take one thing from this: change what you A/B test on. Not CTR, not cost per click, and not landing-page conversion rate. Test creative on accounts moved between engagement stages, and cost per movement.
Practically, that means:
- Run creatives long enough to accumulate account-level movement, which is a slower signal than clicks. Two to four weeks minimum, and be honest that you will have fewer data points.
- Attribute movement to the creatives an account was exposed to in the preceding window, not just the last one. Accounts see several ads; a strict last-touch rule will overstate whatever runs most.
- Accept coarser conclusions. You will be able to tell that one creative is three times better than another. You will not be able to tell that it is 8% better, and you should stop pretending otherwise on click data too.
Keep one broad ad running anyway
None of this means broad creative is worthless. It means it should be honest about its job and funded from a different line.
A high-CTR, broadly appealing asset does real work: it builds category recognition, it earns organic reach, and it gives sales something to send. What it should not do is take budget from an ABM campaign and then be defended on the basis of the engagement metrics it is naturally good at.
Put it in a separate campaign, measure it on reach and organic amplification, and stop comparing it to the ad that is quietly moving nineteen accounts a quarter.