A bad target account list is the most expensive mistake in ABM, and it is also the least visible one.
Your CPMs look fine, your CTRs are acceptable, but the pipeline never materializes because you are spending budget warming up accounts that will never buy from you.
Worse still, LinkedIn’s algorithm learns from your audience.
It starts optimizing toward the bad-fit companies in your list, which means future campaigns become less efficient even after you fix the list.
This post covers Bilal Ahmad’s (GTM and RevOps specialist at Userpilot) list-building workshop at the ZenABM ABM Bootcamp 2026, where he walked through the full process of building an ICP-accurate target account list using Clay, from the “truth list” approach to AI-powered scoring to automated audience updates.
You can watch the full session on YouTube here.
Short on time?
Here’s a quick rundown:
This is Bilal’s central point, and it deserves to be front and center:
There’s one boring way of wasting budget on LinkedIn that you’re not going to notice. And that’s through having a bad list.”
Here is what happens when 30% of your target account list is a bad fit:
LinkedIn’s algorithm receives engagement signals from those bad-fit accounts because they see your ads, and some even click.
LinkedIn then learns that these companies are your audience and optimizes delivery toward similar companies, which compounds the problem with every impression served.
Impressions are up, CTR is within benchmark range, and CPM is normal.
Nothing in the LinkedIn Campaign Manager data tells you there is a problem, because the metrics are measuring activity rather than the quality of that activity.
The engaged accounts are not the ones who would ever buy, but the reporting does not surface this connection clearly enough to diagnose the root cause.
You see engagement without a pipeline and cannot figure out why the two are disconnected.
The team concludes that LinkedIn ABM does not work, reduces the budget, or shuts the program down entirely, and the real culprit (the list, not the channel) never gets identified.
The fix is building the list right from the start by using your existing customers as the ground truth and systematically filtering prospects against that standard.
The truth list is the foundation of the entire ABM target account list building process.
It is a set of accounts you know are a good fit because they are already your customers, or because you have manually selected them as ideal targets if you have no customers yet.
You need at least 30 accounts.
Below 30, there is not enough signal for AI to identify reliable ICP patterns, so if you have fewer than 30 customers, supplement with manually selected ideal target accounts to reach the minimum.
For each account on your truth list, enrich every available field in Clay:
This enriched truth list becomes your ICP definition in data form, not a persona document sitting in a Google Drive folder but an actual data profile of what good customers look like that you can score prospects against programmatically.
Once campaigns go live, ZenABM helps you pressure-test whether your truth list assumptions were right.
Its CRM sync, ABM stage tracking, and company-level engagement data let you compare how different account types actually behave after exposure, which helps refine your ICP with live market feedback instead of static assumptions alone.
With your truth list enriched, the next step is to pull a large pool of potential target accounts and score each one against your ICP criteria.
Sources for your prospect pool:
AI scoring in Clay:
Once you have your prospect pool in Clay, write an ICP scoring prompt and run it against every account:
“Based on the following company data, score this company’s fit with our ICP on a scale of 1 to 10. Our ideal customer: [description based on truth list patterns]. Provide a one-sentence explanation of your score. Flag any company with a score below 6 as Tier 3, do not include in our LinkedIn ABM campaigns.”
With a well-constructed prompt, AI classification achieves 90 to 95% accuracy at a fraction of a cent per account.
Run your 20,000-company prospect pool through this using Clay’s AI column, and the output is a tiered list where Tier 1 (scores 8 to 10) and Tier 2 (scores 6 to 7) go into your LinkedIn campaigns while Tier 3 goes into a low-cost nurture track or gets excluded entirely.
ZenABM becomes a strong feedback layer here, too.
After launch, its company-level engagement reporting can show whether Tier 1 and Tier 2 accounts are actually the ones responding, and its job title analytics can help validate whether the right personas inside those companies are engaging.
That makes your Clay scoring model easier to improve over time.
Your LinkedIn campaign targets companies, but your outbound sequences target people.
After identifying your Tier 1 and Tier 2 accounts, you need contacts, specifically the individuals at each account who match your ICP buyer persona.
Bilal’s approach for contact enrichment in Clay:
Connect to LeadMagic, Prospeo, and Apollo in sequence so that if the first provider does not have a verified email, the system tries the next.
This maximizes coverage without paying for redundant data from a single provider.
Filter by job title, seniority, function, and location.
If your ICP buyer is a VP of Engineering, filter for VP-level and above in engineering functions specifically.
Do not pull all contacts at every company; pull only the contacts who match your buyer persona.
Accounts with fewer than 200 to 300 LinkedIn connections are likely inactive on LinkedIn, which means your ads will not reach them even if you upload them to your audience.
Prioritize contacts with higher connection counts because those are the people who are actually spending time on the platform.
If your website visitor data shows that the people visiting your pricing page from a target account are based in a specific city, find contacts in that city specifically rather than just pulling anyone at the company’s headquarters.
This level of precision compounds the relevance of your outreach.
With your enriched, scored, contact-found account list ready, upload to LinkedIn Campaign Manager as a matched audience.
Two critical requirements:
Without it, LinkedIn matches by company name, which produces mismatches especially for companies with common names.
The URL gives LinkedIn an exact match to the company’s LinkedIn presence. Expected match rate with URL: 90 to 95%. Without URL: often below 70%.
LinkedIn takes time to process company lists, so uploading the day before launch means you start the campaign before your full audience is available.
Give the system at least two full days to process before activating any campaigns against the list.
Target account lists are not static, and Bilal’s recommendation for keeping them current:
That second step is more powerful than it sounds.
ZenABM can surface companies that are already engaging with your LinkedIn ads, even if they were not originally prioritized.
Combined with CRM sync and account-level engagement data, that gives you a live discovery layer for refining your target account list instead of treating list building as a one-time exercise.
For a deeper look at the full Clay ABM workflow, including how to connect your account list to outbound, see Clay for target account list building.
The main lesson from this article is that ABM list building is not just a setup task. It is a performance lever.
If your target account list is weak, your campaigns can look busy while quietly training LinkedIn toward the wrong companies and a starving pipeline.
But when the list is built from a strong truth set, scored carefully, enriched properly, and refreshed over time, the rest of your ABM program gets much sharper.
That is also where ZenABM becomes especially useful.
It helps you move beyond static list logic by showing which companies are actually engaging with your LinkedIn ads, syncing that data into your CRM, tracking account movement through ABM stages, and surfacing new accounts worth reviewing.
If you want your target account list to improve continuously instead of decaying quietly in the background, ZenABM is a strong layer to add.
Try ZenABM for free (37-day free trial) or book a demo now to know more!
Some questions about the ABM target account list and their answers:
It depends on your budget. At $10,000 per month, you can realistically run enough impressions to progress 1,000 to 2,000 accounts through awareness stages, while at $30,000 per month, 5,000 to 7,000 accounts becomes feasible. Use the ZenABM budget calculator to calculate the account list size your budget can support based on your revenue goals and the median 0.58% deal open rate.
With LinkedIn company page URLs included, expect an 85 to 95% match rate. Without URLs, matching only by company name, you will typically see 60 to 75% with significant mismatches. Always include the LinkedIn company page URL column in your upload file.
If you have fewer than 30 customers, manually curate a set of 30 or more ideal target accounts based on your best understanding of who should buy your product, then enrich them in Clay. Use the enriched data to write your ICP scoring prompt, pull a wide prospect pool, and score against that prompt. The quality of your prompt is the key variable here, so spend real time defining your ICP criteria carefully rather than rushing through it.
Quarterly at minimum. Companies change faster than static lists account for: acquisitions, leadership changes, funding events, and product pivots all affect fit. Additionally, add accounts surfaced by ZenABM that are engaging with your ads but not yet in your list, and remove accounts that are confirmed bad fit from sales feedback or that have become customers.
