Most creator selection happens through one of two methods: a brand manager picks someone they already know, or an agency recommends someone from their existing roster. Both methods share the same flaw — they are driven by familiarity, not fit. anchors uses a different approach. The matching process is data-driven, runs against verified LinkedIn data, and produces a recommendation based on what the creator’s audience actually looks like — not what the creator says their audience looks like. This page explains how that process works.Documentation Index
Fetch the complete documentation index at: https://anchors.in/docs/llms.txt
Use this file to discover all available pages before exploring further.
Why relationship-driven selection fails
Before getting into how AI matching works, it helps to understand what it replaces.The agency roster problem
The agency roster problem
Agencies maintain direct rosters of creators they manage personally, and indirect rosters sourced through partner agency networks. When a brand briefs an agency, the recommendation comes from whoever is in that roster — not whoever is best for the brief.Discovery of creators outside the roster is operationally difficult. New creators entering the market, niche creators in specific industries, and smaller creators without agency relationships are effectively invisible. The brand gets whoever the agency already knows.
The relationship bias problem
The relationship bias problem
On the brand side, selection is often driven by existing relationships. A brand manager who has worked with a creator before defaults to them. A marketing lead who follows a creator on LinkedIn recommends them internally. These decisions feel safe because they are familiar — but familiar is not the same as fit.A creator selected on relationship rather than audience fit will almost certainly reach the wrong people. The brand pays for impressions that do not convert — not because the creator performed badly, but because the creator’s audience was never the right audience to begin with.
The data problem
The data problem
Even when brands try to be rigorous, the data they use is usually unreliable. Creators share impression screenshots that cannot be independently verified. Follower counts are visible but have almost no reliable relationship to actual impressions delivered. Engagement rates are easy to misread without knowing the audience behind the numbers.anchors ran a campaign where a creator with 2,00,000 followers delivered 10,000 impressions — while a creator with 13,000 followers delivered 43,000 on the same campaign. Any selection process based on follower count would have ranked the wrong creator higher.
What goes into the matching process
anchors’ AI matching takes two sets of inputs and finds the overlap between them.Input 1 — The brand’s ICP
Before a campaign runs, the brand defines who they need to reach. On anchors, this is set through the influencer criteria step and — for brands using CLEO — through the AI campaign builder which generates the ICP automatically from the brand’s domain and product details. The ICP includes:| Dimension | What the brand specifies |
|---|---|
| Job roles | Which professional roles are the target buyer or influencer |
| Seniority | Entry level, manager, director, VP, C-suite — or a mix |
| Industry | Which industries the target works in |
| Location | City, state, or country-level targeting |
| Business type | Whether the campaign is B2B, B2C, or D2C — this affects which audience signals matter |
| Campaign goal | Awareness, lead generation, product launch — influences which matching signals are weighted |
Input 2 — The creator’s verified audience data
Every creator on anchors has connected their LinkedIn account directly. This means anchors reads their real audience demographics from LinkedIn — not estimates, not scraped data, not self-reported numbers. The creator data includes:| Data point | What it reflects |
|---|---|
| Audience job roles | Actual roles of the creator’s LinkedIn followers |
| Audience seniority distribution | What % of followers are at each seniority level |
| Audience industry breakdown | Which industries the creator’s followers work in |
| Audience location | Geographic distribution of followers |
| Average impressions per post | Real delivery history — not follower-count projections |
| Engagement consistency | Whether the creator’s performance is consistent or spiky |
| Content category | What the creator actually posts about |
| Past collaboration history | Whether the creator has run brand campaigns before and how they performed |
How the match is produced
The matching algorithm cross-references the brand’s ICP against every creator’s verified audience profile and calculates an overlap score across the four key dimensions — role, seniority, industry, and location.ICP is finalised
The brand sets their targeting criteria. If using CLEO, the AI generates a suggested ICP from the brand’s inputs — the brand reviews and confirms before the campaign proceeds.
Creator pool is filtered
Creators who have consented to brand collaborations and have synced LinkedIn data are pulled from the anchors pool. Any creator with insufficient data history or who has not completed onboarding is excluded at this stage.
Audience overlap is calculated
For each creator, the algorithm measures what percentage of their audience matches the brand’s ICP across all four dimensions. A creator whose followers are 45% marketing managers in India for a MarTech brand targeting Indian marketing managers scores high. A creator whose followers are 60% founders in the US for the same campaign scores low — regardless of how many followers they have.
Match score is weighted by campaign goal
The raw audience overlap is adjusted based on the campaign goal. An awareness campaign weights reach and geographic coverage. A lead generation campaign weights seniority match and industry concentration. The same creator may rank differently for different campaign types.
Why verified data changes the result
The matching output is only as reliable as the input data. This is where anchors’ direct LinkedIn sync matters.What verified data enables
- Impressions based on actual delivery history — not follower-count projections
- Audience demographics pulled directly from LinkedIn profiles — not platform estimates
- Engagement data that reflects real creator-audience relationships
- Creator availability and activity tracked in real time
- No self-reporting — the creator cannot inflate their numbers
What unverified data produces
- Impression projections that can be off by 4x in either direction
- Audience claims that cannot be cross-checked — self-reported or screenshot-based
- Engagement rates that look good on paper but reflect the wrong audience
- Creator profiles inflated specifically to qualify for campaigns — changing stated location, job title, or expertise to meet brand criteria
What the brand sees during selection
Every creator card shown to the brand during the selection step includes:- Audience fit score — the match percentage against the brand’s ICP
- Average impressions — real delivery history from past posts, not follower-count projections
- Engagement rate — calculated from actual post data
- Average likes and comments — not just combined engagement, but separated so brands can see comment depth
- CPM — the cost per 1,000 impressions for this creator and this campaign type
- Content category — what the creator actually posts about
- Audience demographics — role, seniority, industry, and location breakdown (unlocked with insight tokens)
- Past collaborations — how many brand campaigns this creator has completed on anchors
How matching connects to pricing
anchors does not price creators on follower count. Pricing is based on verified impressions from past posts — what the creator has actually delivered, not what their follower count suggests they might deliver. The CPM range on anchors sits between ₹200 and ₹800 per 1,000 impressions, varying by audience seniority, business type, creator niche, and campaign period. The brand sees the estimated reach and the expected cost before confirming any creator — there are no surprises after the campaign runs. In anchors’ CARS24 campaign, this model delivered a ₹55 CPM — against a LinkedIn Ads benchmark of ₹200 to ₹700 for the same audience. The gap came from performance-based pricing on verified delivery, not from fixed fees on projected reach.The ₹55 CPM from the CARS24 campaign is a specific result from one campaign — not a guaranteed outcome. CPM on anchors varies by campaign. The directional point is that performance-based pricing on verified impressions consistently produces competitive cost efficiency — because you pay only for what is actually delivered.
Related pages
What Is Audience Fit
The concept behind the matching score — and why overlap with your ICP matters more than follower count.
What Is ICP for Influencer Campaigns
How to define the ICP inputs that drive the matching process.
How to Evaluate a LinkedIn Creator
What to check beyond the match score before confirming a creator.
Creator Tiers
How tier interacts with match scoring across different campaign goals.