AI can build a list that starts with a clear inclusion rule instead of a random search result.

Write the inclusion rule as if someone else has to audit it.

“Find great SaaS companies” produces a list that cannot be judged.

A usable request says where to look, what qualifies, what excludes a company, how current the signal must be, and which evidence goes in the row.

Strawberry can work from that rule and preserve the source material beside the result. The list then becomes a decision document rather than a pile of names that only makes sense to the person who searched for them.

Search more than one path to avoid a list shaped by one directory.

A single directory often over-represents companies that market to that directory.

Better lists combine the relevant public sources with specialist databases, then make the source of each result clear.

Apollo supports account, company, contact, people, and search operations. People Data Labs supports searching people and companies, while Hunter supports company discovery and domain search, so Strawberry can build several candidate pools before the team chooses the final list.

Deduplicate on the entity, not just the spelling.

The same company can appear as a brand, a legal entity, a subsidiary, and three slightly different domains. A name-only cleanup creates false duplicates and misses real ones.

Strawberry can compare domains, company pages, parent relationships, and the context in each source before it recommends a merge. It should leave ambiguous cases in a review bucket, because aggressive deduplication can remove the very account someone needs to keep.

Treat the first batch as a quality test, not a finished market map.

A hundred rows can feel productive even if half of them do not meet the criteria.

Inspect a random sample before the list feeds outreach, recruiting, partnership work, or a large research project.

Save the approved sourcing and QA rules as a Strawberry skill. A routine can return a new batch at a set cadence, including a fixed sample for review, while keeping CRM creation, email verification spend, and outreach outside the list-building step.

Frequently asked questions