You cannot add two creators' follower counts and call the result campaign reach. Some accounts may follow both creators, and following a creator does not establish that someone saw the campaign. Influencer audience overlap estimates help you assess how much the follower groups share. They cannot, by themselves, tell you how many distinct people will see the posts.
Use overlap to compare possible creator combinations. Keep the combined follower estimate separate from any forecast or report of campaign exposure.
Calculate the shared accounts first
Consider a hypothetical campaign with two creators on the same platform. These are invented accounts and numbers, used only to explain the calculation.
| Follower group | Hypothetical count |
|---|---|
| Creator A's followers | 100,000 |
| Creator B's followers | 60,000 |
| Accounts following both | 20,000 |
Adding the two counts gives 160,000 follower relationships. The shared 20,000 accounts appear twice in that sum.
The combined follower count after removing duplicates is:
Combined distinct accounts = A + B - shared accounts
= 100,000 + 60,000 - 20,000
= 140,000If A is already selected, B adds 40,000 follower accounts outside A's follower group. That is a calculation about membership in two groups. It says nothing yet about who will watch either sponsored post.
With complete, matching account lists, this would be an observed count of distinct accounts. With an estimated overlap input, 140,000 remains an estimate, even though the arithmetic is exact.
Ask what the overlap percentage divides by
The same hypothetical shared group produces different percentages:
| Denominator | Calculation | Result |
|---|---|---|
| A's followers | 20,000 / 100,000 | 20% |
| B's followers | 20,000 / 60,000 | 33.3% |
| Combined distinct accounts | 20,000 / 140,000 | 14.3% |
A report saying only "20% overlap" leaves out information you need. Ask whether it measures overlap relative to one creator, the smaller audience, the combined audience, or another base. Do not compare two providers' percentages until you can reproduce both calculations.
For three or more creators, pairwise overlap figures alone may be insufficient. An account following all three appears in several pairwise intersections. Subtracting each pair's shared followers without accounting for that three-way intersection can remove the same account too many times. Ask for a deduplicated total for the whole proposed list.
Separate followers, viewers and people
The platform's unit matters. Instagram's current insights definitions distinguish follower information from content exposure. Its help page defines Viewers as unique accounts that saw content on screen at least once. Views counts plays or displays. The page also marks certain metrics as estimated and in development.
YouTube uses a different measurement process. Its unique viewers documentation describes an estimate of people watching during a period. The system accounts for different devices and signed-in and signed-out traffic. YouTube also explains that subscribers may not return for every upload.
These distinctions create three separate questions:
- Which accounts belong to both creators' follower groups?
- Which viewers saw both campaign posts during the reporting period?
- Which records represent the same person across platforms?
An answer to the first question does not answer the other two. Even separate creator reports of unique viewers do not reveal their shared viewers. YouTube illustrates this within one channel: a person can count as a unique viewer for each video, then count once in the channel total.
Nor does an Instagram account count establish which YouTube viewer is the same person. The cited platform definitions do not supply that cross-platform match. Matching usernames or similar demographic percentages is insufficient evidence to identify shared people. Keep platform totals separate unless your measurement method explains how it resolves identity and what it misses.
For reporting labels beyond overlap, use the guide to reach, impressions and views in one campaign.
Read the provider's method before its headline number
Modash's audience overlap article describes a creator-list feature that reports unique followers and percentages. The article says this check works for Instagram audiences. That stated scope matters when planning a campaign across several platforms.
Its data methodology page says it collects public creator information and applies machine-learning models to produce creator and audience insights. Those two pages do not give an overlap-specific sample size, coverage rate or error range. Their description alone cannot establish the precision of an individual overlap result.
Use three labels in your planning sheet:
- Observed. A count or field you can trace to a dated source, with its definition attached. A displayed platform field can itself be an estimate.
- Inferred. A provider's estimate or your scenario, with its method and assumptions recorded.
- Unavailable. Information the source does not supply, such as the identity match needed to deduplicate viewers across platforms.
Do not quietly turn unavailable information into zero overlap. If a provider reports too little data, keep the result unknown.
Run an uncertainty check before changing the shortlist
Ask these questions before paying more for a supposedly less-overlapping creator pair:
| Check | What to record |
|---|---|
| Population | Followers, recent viewers or accounts that interacted? |
| Platform | Same-platform comparison or a claimed cross-platform match? |
| Timing | Collection dates and reporting window for each creator |
| Coverage | Complete lists, samples or modeled data; known exclusions |
| Denominator | The base used for every percentage |
| Identity | Account-level matching or person-level estimation |
| Exposure | Whether the measure concerns followers or campaign viewers |
| Uncertainty | Published error range, or an explicit statement that none is supplied |
If uncertainty could change your choice, test alternative inputs. In the hypothetical example, suppose you try shared-follower counts of 10,000, 20,000 and 30,000. The resulting combined counts are 150,000, 140,000 and 130,000. These are planning scenarios, not a statistical confidence interval.
Would you select the same pair throughout that range? If so, obtaining a more precise overlap estimate may not change the decision. If the choice reverses, ask the provider about coverage before treating a small estimated advantage as decisive.
Low overlap also does not establish audience relevance. Two groups can have little in common because one is outside your target market. Check audience fit before follower count before comparing the incremental audience each creator adds.
For a campaign seeking exposure beyond an existing creator's audience, lower follower overlap can be a useful selection signal. For a campaign intended to repeat a message within a defined community, shared followers may be acceptable. Neither choice proves delivery or a sales effect.
Before approving the next creator, put the overlap denominator, platform, collection date and uncertainty beside the estimate. Label the result "estimated combined follower accounts," and keep the campaign reach field separate until you have evidence about actual exposure.
