Not different copy. Not different creators. Not different product shots. Different underlying concepts — ads that are structured differently, that lead with different problems, that use different hook styles, that are clearly intended for different types of viewers.
For most brands, the honest answer is somewhere between three and seven. The rest of the library is variations.
Why This Matters More Than It Used To
Creative diversity has always mattered in paid social. But it mattered in a relatively forgiving way — as long as you had a few different angles working, the system would find audiences for them. The inefficiency of running homogeneous creative was real but not acute.
Meta's Andromeda update changed the tolerance the algorithm has for conceptual repetition. The system now evaluates creative by its underlying concept archetype, not its surface execution. Multiple variations of the same concept are recognized as the same concept. Once the algorithm has learned what your audience does with a given concept, additional variations provide diminishing returns.
The practical result is that brands running what looks like a diverse creative library are often getting the distribution efficiency of a much smaller one. The algorithm is smarter than the spreadsheet.
The Seven Archetypes
Every paid social ad can be classified into one of a small number of concept archetypes. The most common ones that appear in high-performing paid social libraries are: Problem-Solution, Testimonial, Demo, Lifestyle, Before-After, Social Proof, and Education or Tutorial.
A healthy creative library has representation across most of these. An unhealthy one — the most common kind — is dominated by one or two. The most common imbalance is an over-index on Problem-Solution and Testimonial, with almost nothing in Demo, Before-After, or Education. These categories often underperform individually because brands under-invest in them, not because they do not work.
The Hook Style Problem
Concept archetype is one dimension of creative diversity. Hook style is another, and it is often more acute.
Most brand creative libraries are built around a small number of hook mechanisms. Direct problem call is the most common: open by naming a problem the viewer has and positioning the product as the solution. This works. It also gets recognized quickly by the algorithm as a pattern once enough variations exist.
The other hook styles — question hooks, visual pattern interrupts, shocking statements, social proof openers, search-intent openers — are underrepresented in most libraries because they require more creative risk and more production effort. They also represent the diversity signal the algorithm is most actively rewarding right now.
A library with twelve problem-call hooks and two question hooks has a hook style diversity problem regardless of how different the twelve problem-call executions are.
How to Diagnose Your Own Library
You do not need a sophisticated tool to do a first-pass audit of your creative library. Pull your active and recently paused ads. For each one, write down: what archetype is this, and what hook style does it use?
Plot the results. You will almost certainly find that 60 to 80 percent of your library clusters into two or three combinations. That clustering is your diversity problem.
The gaps in the map — the archetypes and hook styles that are absent or underrepresented — are your brief for the next production cycle. Not more variations of what you already have. Genuinely different concepts that fill the gaps the algorithm is not yet rewarding you for.
The Brief Is the Fix
Most creative diversity problems are brief problems. They come from briefing for volume rather than variety, from asking creative teams and creators to produce more of what has worked rather than something conceptually different, from optimizing for efficiency in production rather than diversity in output.
The fix is not producing more creative. It is producing different creative. A brief that specifies the archetype and hook style required — not just the product features to highlight and the audience to target — produces a genuinely different concept rather than another variation.
The brands winning on paid social right now are not producing more. They are producing differently. The audit is the first step. The brief is where the diversity problem gets solved.
Running the audit once is a start. Building a system that monitors your concept and hook distribution continuously — and surfaces the gaps before they become CPM problems — is where the compounding advantage comes from.