1. Translate the audience into explicit filter logic
Write the audience in one sentence before opening a search tool. “Founders or heads of sales at software companies in France or Belgium” has two role alternatives, two country alternatives and one industry requirement. That sentence is easier to review than a pile of disconnected filter chips.
In Feediqo, alternatives within the same field broaden the match. Different fields narrow it. A record can match Founder or Head of Sales, but it must also match the selected industry and one of the selected countries. Quoted keyword phrases stay together, while separate keyword terms are treated as separate requirements.
2. Separate company fit from role fit
A suitable company does not guarantee that every employee is a suitable contact. First decide which organisations can benefit from the offer. Then decide which roles are likely to understand, evaluate or influence it. Keep those decisions separate so a matching company name does not conceal an irrelevant title.
Company-only records are useful for account research, but they cannot satisfy a job-title condition when no person-role field exists. If the campaign needs a person, use a people search and verify the current company association. If the campaign begins with accounts, save companies first and research contacts as a separate step.
3. Use exclusions to protect list quality
Exclusions are not just a way to reduce the result count. They encode businesses and situations that should never enter the working list: existing customers, competitors, agencies when you need end buyers, unsupported countries, and previously suppressed contacts.
Apply an exclusion when the rule is stable and objective. Do not exclude a broad category simply because one example was poor. A long exclusion list can hide a weak audience definition, so inspect which condition removed a promising record before expanding it.
4. Remember that unknown is not the same as no
A record with no industry label is not automatically outside your industry. A missing field does not describe the business; it describes the limits of the available record. Presence filters and supported numeric thresholds usually remove records whose relevant field is unknown, which can make an apparently precise segment incomplete.
Use required fields only when the next action truly depends on them. If email is essential for the immediate workflow, “has email” is sensible. If a phone number would merely be convenient, leaving that filter off preserves otherwise relevant records for later review.
5. Test the segment with a review sample
Review the first page before revealing contacts. Count clear fits, uncertain records and obvious mismatches. Note which filter admitted each mismatch and whether the problem comes from the audience definition, an alias, or old source data. Change one condition at a time so you can tell what improved the sample.
A segment is ready when another person could explain why its records belong and when the sample produces enough relevant prospects to justify deeper research. Save the search criteria for reuse, but save only reviewed records to a prospect list.