You set up a campaign targeting 800,000 people. You give it a reasonable budget. After two weeks, you look at the reach data and realise the same 15,000 to 20,000 people have been seeing the ads repeatedly. The frequency is climbing. The reach is flat. A large portion of your defined audience has apparently never seen the campaign at all. You have a same audience problem.
This is one of the more frustrating experiences when scaling Meta ads, and it is commonly misread as a platform problem. Meta must be prioritising certain users unfairly. The audience must be corrupted. Something must have gone wrong in the setup.
The real cause is usually structural, and it is fixable once you understand the mechanism behind it.
Why the Algorithm Concentrates Delivery
Meta’s algorithm does not distribute ad delivery evenly across your defined audience. It optimises toward the users within that audience who are most likely to produce the result you are optimising for. If you are running a purchase conversion campaign, it serves your ads to the users it predicts are most likely to purchase, not to a representative cross-section of everyone you defined.
This creates natural concentration. The algorithm finds a subset of your audience where purchase signals are strong, invests most of the budget there, and reaches the outer edges of the audience only when the budget is large enough to exhaust the high-intent pool first. At modest budget levels relative to audience size, the algorithm may never meaningfully explore the outer portions of the audience at all.
This is not a bug. It is the algorithm working correctly. The problem is that this concentration compounds over time. The same high-intent users are reached repeatedly, frequency climbs on them, their engagement signals deteriorate, and what looks like an 800,000-person audience is effectively functioning as a 20,000-person audience in terms of where the spend is actually going.
Campaign Cannibalisation: When Your Own Campaigns Compete
The situation gets worse when multiple campaigns are running against overlapping audiences. If you have three separate campaigns that all target “website visitors” or that all target similar interest groups, they enter the same auction simultaneously. They compete against each other for the same users, driving up the cost of reaching those users and concentrating delivery among the highest-intent fraction that all three campaigns are bidding for.
This is campaign cannibalisation. The symptoms look like high frequency, high CPM, and stagnant reach, even though the budget is sufficient to reach a much larger audience. The root cause is that multiple campaigns are bidding for the same pool of people without any structure to prevent it.
The fix is audience segmentation with clean exclusion logic. Each campaign should target a non-overlapping audience segment, and exclusion lists should ensure that people who qualify for more than one campaign are served by the most appropriate one. Existing customers should be excluded from prospecting campaigns. People who have already converted should be excluded from conversion retargeting. MOFU audiences should exclude BOFU audience members to prevent mid-funnel content from reaching people who are already in the conversion stage.
Without these exclusions, the structural problem that produces same-audience concentration is compounded by multiple campaigns fighting over the same people simultaneously.
The Role of Creative in Audience Expansion
There is a less obvious contributor to the same-audience problem: creative that generates signals only from a narrow profile of users. The algorithm uses the engagement signals from your creative (who clicked, who watched, who converted) to build a model of who in your broader audience is most similar to those early converters. If the early converters are all from a similar demographic, the algorithm narrows delivery toward that demographic.
This is not a flaw in the algorithm. It is a flaw in what the creative is communicating. A creative that resonates only with a narrow profile generates a narrow signal. The algorithm optimises toward more of the same narrow profile, reach concentrates, and the rest of the defined audience is effectively unreached.
Creative that generates broader signals produces broader delivery. A creative that resonates with diverse sub-segments of your audience, different demographics, different problem framings, different purchase motivations, generates a more diverse signal pool. The algorithm builds a broader model of your potential audience from that diverse signal and distributes reach more widely.
This is why testing multiple creative variations, particularly with different hooks and different audience-framing angles, often expands effective reach even without any change to the audience targeting. More creative variations give the algorithm more options for who to reach with each specific asset, distributing delivery more effectively across the total defined audience.
Broad Targeting as a Structural Solution
One of the counterintuitive findings from recent Meta advertising analysis is that broad targeting, with minimal or no interest restrictions and a large potential audience, often produces better ROAS and wider reach than narrowly defined interest audiences. Analysis comparing broad targeting against lookalike audiences found broad targeting delivering 113% ROAS against 76% for lookalikes in a controlled comparison.
The reason is that narrow interest targeting concentrates the algorithm’s signal-gathering into a pre-defined profile, which limits the algorithm’s ability to discover the diverse sub-segments within a broader population who would actually respond to the offer. Broad targeting gives the algorithm access to a much larger pool of people to find signal in, which often produces better results than the advertiser’s own hypothesis about who the audience is.
The caveat is that broad targeting requires strong creative to guide the algorithm. Without compelling creative that generates meaningful engagement signals across a diverse audience, the algorithm will still concentrate delivery among the easiest-to-find high-intent users. The creative, not the audience definition, becomes the primary mechanism for telling the algorithm who to find.
Audience Sizing and Budget Calibration
The concentration problem also has a budget dimension. An audience of 800,000 people with a daily budget of $50 is severely under-funded relative to its size. At $0.50 to $1.50 CPM (cost per thousand impressions), a $50 budget might reach 33,000 to 100,000 impressions per day. Against 800,000 potential targets, that is a very small daily reach percentage, and the algorithm will concentrate those impressions among the highest-intent fraction.
The audience-to-budget ratio determines how quickly the algorithm can meaningfully explore the full audience. A general guideline: the daily budget should be large enough to reach at least 5 to 10% of the total audience weekly. Below that ratio, the campaign’s effective reach is concentrated by design, not by error.
For businesses where the budget cannot support full audience coverage, the correct response is to reduce the defined audience size to match the budget, not to maintain a large audience and accept concentrated delivery. A campaign targeting 100,000 people with a $50 daily budget has a better audience-to-budget ratio and will typically produce better reach diversity than the same budget against an 800,000 person audience.
The Bottom Line
Meta showing ads to the same people is a symptom of algorithm concentration, campaign cannibalisation, creative signal narrowing, or audience-to-budget mismatch. Usually it involves more than one of these simultaneously.
The fix requires addressing all the contributing factors: clean audience segmentation with exclusion logic, creative that generates diverse engagement signals, audience sizing calibrated to the budget, and a campaign structure that prevents multiple campaigns from competing for the same users.
Before diagnosing the problem as a platform issue, ask yourself:
- Do you have overlapping audiences across multiple campaigns without exclusion logic in place?
- Is your creative generating signals from a narrow demographic, which might be limiting the algorithm’s delivery model?
- Is your daily budget proportionate to your defined audience size?
- Have you checked whether existing customers and converters are excluded from prospecting and retargeting campaigns?
- Do you have multiple creative variations running that might generate more diverse engagement signals?
The same-audience problem is architectural. It has an architectural solution.
Book a free consultation with the SynapseBN team — no pitch, no pressure. Just a straight conversation about what’s working, what isn’t, and what to do about it.
