Creative analytics · Meta

Facebook ads creative analytics:what kind of creative wins.

See what kind of creative wins on your Meta account: AI tagging, creative DNA, spend-weighted findings and a written report. Included on every plan.

• Performance grouped by actual creative, not by ad
• Six comparable traits, tagged by AI
• Findings weighted by spend, not by count

Start free trialSee pricing

7-day free trial · No credit card

The report

Reporting says which ad won. This says what to brief next.

TRAITS / CREATIVE

6

Format, hook, angle, emotion, stage, tone

TAG VALUES

56

One fixed vocabulary for every creative

ANALYSIS / CREATIVE

1

Analyzed once, reused forever

PLANS

3" of 3"

Included on Starter, Growth and Agency

Source: uplads product, October 2026

The uplads creative report groups your Meta ad performance by the real creative asset, tags every creative with 6 traits from a fixed list of 56 values, and ranks which kinds of creative earn spend on your account. It is included on all 3 plans (from 69 EUR a month) and in the 7-day trial.

Ads Manager answers "which ad spent what." It cannot answer the question a media buyer actually briefs from: what kind of creative wins on this account - UGC or produced, question hooks or bold claims, problem-first or offer-first. Answering that requires two things Ads Manager does not do: grouping performance by the real creative asset, and describing every creative in comparable terms.

The uplads creative report does both, on top of the same bulk launch library you already work from.

First problem: Meta splits your creative's history

Upload the same video into five campaigns and Meta stores five video IDs. Your best creative's track record is scattered across rows that look unrelated. uplads folds them back together - by the creative's actual content and by the upload library's file fingerprints - so one creative is one row: total spend, total results, every ad it ran in, every country it ran to.

That grouping is the foundation. Every number on the page - ROAS, cost per result, hook rate, hold rate, CTR - is per creative, with a one-click country filter and a comparison window against the previous period.

Second problem: "which won" is not "what kind wins"

The analysis

Every creative, read once and described in six comparable traits.

Visual format

UGC, talking head, product shot, testimonial, before/after and more.

Hook tactic

Question, problem callout, bold claim, social proof, curiosity.

Ad angle

The argument the ad makes, comparable across the account.

Emotion

The feeling the creative leads with.

Awareness stage

Who the ad speaks to, from cold to ready-to-buy.

Tone

How it says it - and whether that tone earns spend here.

An AI pass reads each creative once and tags it along those six fixed dimensions: 12 visual formats, 11 hook tactics, 11 angles, 9 emotions, 5 awareness stages and 8 tones. The list is fixed on purpose: a shared vocabulary is what makes "every UGC ad on the account" a filterable, comparable set instead of a vibe. The same pass stores a detailed per-creative deconstruction: the observable hook, the mechanism, the core desire, value proposition, offer, proof, objections, CTA and on-screen text. Videos are transcribed, so the analysis includes what the ad says, not only what the thumbnail shows.

Tagging runs in batches of up to 10 creatives. Each creative is analyzed exactly once and the result is stored and reused, so re-opening the report costs nothing and AI credits are only spent when new creatives appear.

Creative DNA: where the spend sits, and whether it earns its keep

The DNA view splits the account's spend across every trait value and grades each one against the account's own benchmark - not an industry average. A typical read looks like this: most of the spend sits on product shots that perform below the account benchmark, while UGC runs above it on a small share of the budget. Each trait value shows its share of spend and its result against the benchmark, so that pattern is visible on one screen. Filter the entire report down to one trait and compare like with like.

On top of the findings sits a written report: what works on this account, what does not, and what to do next - ranked, in prose you can paste into a creative brief or forward to a client. Shareable read-only links exist for exactly that.

The Launch Cockpit: did what I launched actually win?

The second view answers a narrower question from uplads' own launch history: what did I launch, and did it win? Because it is built from your launches rather than reconstructed from the ad account, it stays honest about what you shipped - including creatives that never earned delivery. It covers uplads-launched creatives; the creative report above covers the whole account.

How it fits the loop

The loop

From launch to a briefed next batch.

1. 01

   Launch in bulk

   Ship the batch through the launch flow - grouped placements, clean naming, one pass.

2. 02

   Let it spend

   The report reads real delivery data from Meta. No pixel to install, no tracking migration.

3. 03

   Read the DNA

   See which formats, hooks and angles earn spend on this account, graded against its own benchmark.

4. 04

   Brief the next batch

   Replace "make more ads" with "more UGC problem-callout hooks for cold traffic." Generate the copy with AI ad copy.

Where it ends

This is post-launch intelligence for Meta accounts. It does not predict a creative's performance before it runs, it does not analyze TikTok or Google accounts, and it is not an attribution suite - there is no pixel and no customer-journey tracking. It reads what your Meta account already knows and turns it into a brief. For the launch side of the loop, start at the bulk upload page; for plans, the pricing page - the report is included on every tier.