Bulk account intelligence & ICP scoring

Lists is where a spreadsheet of company domains becomes a worked, scored, prioritised account table. It is built for the moment a rep or a RevOps lead is handed 300 domains and asked which ones matter.

Open Lists from the top bar.

Building a list

1. New list — name it, and paste or upload the domains. CSV upload and plain paste both work.
2. DatIQ normalises and de-duplicates as it imports: https://www.Acme.com/pricing, acme.com and

ACME.COM are one account, not three. The import summary tells you how many duplicates it collapsed.

1. Pick the persona the list is for — it seeds the scoring rules with a sensible starting profile.

Enrichment that never invents a field

Running a list reads each company's public site and returns firmographics: industry, size band, pricing model, whether they publish pricing at all, positioning, proof points, and contacts where a site exposes them.

Every field is one of exactly three things: observed (read directly off a page), inferred (derived from what was observed, and labelled as such), or absent. There is no fourth state. Absent fields are omitted rather than filled with a plausible-looking default — which means an honest run produces a lower coverage number rather than a wrong score.

Runs are durable and chunked. A list of 500 accounts is processed in claimable batches, so closing the tab, losing your connection, or a provider having a bad minute does not lose the work already done. Re-open the list and the progress is where you left it.

ICP scoring — rules you can actually edit

The Rules tab holds your Ideal Customer Profile as data, not as an opaque model. Each criterion is a field, an operator, a value, and a weight — for example industry is one of Software, Fintech at weight 30, or employee count is at least 50 at weight 20. Mark a criterion required and an account that fails it cannot qualify however well it scores elsewhere.

Two rules govern the maths, and both exist to stop the score lying to you:

• An unmeasured field is never scored as zero. If a company's headcount could not be found, that

criterion is excluded and its weight is redistributed across the criteria that were measured. Scoring it zero would punish a company for our failure to read their site.

• Coverage travels with every score. A 70 computed from five of five criteria and a 70 computed from

two of five are different claims, and the table shows you which one you are looking at. An account with zero coverage scores no result, not zero.

Set the qualification threshold (50 by default) and test it live against a sample domain before you apply it to the whole list. Change a weight, watch the sample re-score, then run.

The review queue

Low-confidence extractions land in Review rather than silently entering your table. You confirm, correct, or discard them. This is deliberate: the alternative is a 94%-accurate table that nobody can tell the bad 6% inside, which is a table nobody trusts.

Getting the list out

Export the qualified accounts as CSV or JSON, or push them straight into HubSpot, Notion, Airtable or Slack from the same screen. Scores, coverage, and the source URL behind each field travel with the export — so the rep working the list can see why an account qualified, not just that it did.

Plan note. Bulk lists use your plan's batch allowance, so the number of accounts you can enrich in one
list matches the batch size your plan already includes (see §18). Top-up bundles raise it.