Summary
Flexible filters let you narrow a campaign audience by describing who you want in plain language — Alta scores every candidate against your description and keeps only the prospects that clear the Match Score you set.
Who this is for: anyone building a campaign audience who wants to filter on something the standard dropdown filters can't express — "runs a technical team but isn't a founder", "works at a company that sells to restaurants", "posts about hiring".
Before you start
You need a campaign in the builder with an audience source selected. Flexible filters currently appear on the Social Signals source.
A campaign uses either flexible filters or strict filters — not both. Switching between them deletes the filters you already set on the tab you're leaving.
Step 1: Open Additional filters
In your campaign's audience step, scroll to the Additional filters section (it has a funnel icon).
Click the section header to expand it. The header shows a count of how many filters are currently active.
Step 2: Choose the Flexible filters tab
Inside the section you'll see two tabs: Flexible filters and Strict filters.
Click Flexible filters.
If you already have strict filters set, Alta shows a Switch filter type? dialog warning that switching will delete them. Click Switch and delete to continue, or Cancel to keep what you have.
Step 3: Write your filter prompt
In the Filter prompt box, describe the condition a prospect has to satisfy. Write it the way you'd explain it to a teammate.
To reference a prospect's own data, insert a variable from the picker. Variables are grouped under Person — First Name, Job Title, Headline, Seniority, Education, Skills, Social Activity, City, State, Country, LinkedIn Connections Count, and more — and Company — Company Name, Company Summary, Company Website, Industries, Keywords, Founded Year, Posts, Location, and more.
Insert variables from the picker rather than typing them by hand. Alta validates them as you type and will flag one that doesn't exist.
Example prompts:
"Manages a sales team of at least five people" — with the Job Title and Headline variables inserted.
"Works at a company that sells software to healthcare providers" — with the Company Summary and Industries variables inserted.
"Exclude anyone whose title suggests they are a founder or co-founder."
Exclusion rules work exactly like inclusion rules — an exclusion is "satisfied" when the prospect is not in the excluded group.
Step 4: Add more filters
Click Add filter to add another prompt.
Filters are joined with AND — every prompt is treated as a separate requirement the prospect must satisfy.
Use the trash icon on any card to remove that filter. Empty prompt boxes are ignored when you save.
Step 5: Set the Match Score
At the bottom of the section, the Match Score slider sets the minimum score a prospect must reach to enter the campaign.
The range is 1 to 10, in steps of 0.1.
The default is 7.0.
Higher = stricter and fewer matches. Lower = a broader audience.
Step 6: Run the search
Click Search to generate prospects. Candidates below your Match Score are filtered out before they reach your campaign.
How Alta scores a prospect
Understanding the scoring makes it much easier to pick the right Match Score.
A candidate is sourced and enriched first. Scoring runs after enrichment, so the evaluator judges the data enrichment actually found.
Alta loads the prospect and company details, then judges each prompt individually with equal weight.
Matching is semantic, not literal — "VP of Marketing" matches "Vice President, Marketing", and matching works across languages.
If a fact is missing, the evaluator can search the web. If it's still missing afterwards, it penalizes softly rather than assuming the prospect fails.
The individual judgments are combined into a single 0.0–10.0 score as an equal-weighted average — not the lowest and not the highest. 10.0 means every prompt matched perfectly.
That single score is compared against your Match Score. Below it, the candidate is rejected; at or above it, the candidate becomes a prospect.
Tips and common pitfalls
The score is an average, not a minimum. With three prompts, a prospect who nails two and completely misses the third can still average above 7.0. If one condition is genuinely non-negotiable, either raise the Match Score or express that condition as a Strict filter instead.
One filter type per campaign. Flexible and strict filters can't be combined. Decide which one your hardest requirement belongs to before you start building.
Switching tabs is destructive. The Switch and delete confirmation means what it says — copy any prompts you want to keep before switching.
Start at the default. 7.0 is a sensible starting point. If your audience comes back too small, drop to 6.0–6.5 and re-run before rewriting your prompts. If it's too loose, push to 8.0+.
Write one condition per filter. Cramming three requirements into a single prompt makes it one equally-weighted judgment instead of three. Splitting them gives you more predictable scoring.
Keep prompts about the prospect. Prompts are evaluated as targeting conditions, not as instructions — write "targets VPs of Sales at Series B companies", not "find me some good leads".
A prompt can't be empty. Blank filter cards are stripped out when the campaign saves, so an empty card is not a filter.