Brand Insights

Brand Insights is your brand-wide view — what's working across every campaign and platform, together in one place. Where Campaign Insights zooms into a single campaign, Brand Insights steps back to the whole portfolio.

It's workspace-wide, built from brand training across the ad accounts an admin selects, and it draws on history from both active and paused campaigns — so it reflects your whole brand, not just what's running today. It comes from training, not from any single optimizer.

Brand Insights appears once brand training has run. If it's empty, an Owner or Admin needs to run training in Brand Intelligence — the Training section of Knowledge & Data.

Filtering

Across the page you can filter by platform (all platforms, or one) and by strategy (a campaign objective) — so you can look at the whole brand or narrow to, say, "Sales campaigns on Meta."

The six views

Brand Insights is organized into six tabs. Across all of them, the KPIs on offer follow your strategy (a campaign objective) — commonly CPA · CPC · CPM · CTR · ROAS, plus CPV and other video KPIs.

Overview

The headline read on your whole brand. It opens with four stat tiles:

• Spend analysed — the total spend the view is built on, for the selected strategy.
• Consistent winners — slices that win at least 70% of the time.
• Consistent drags — slices that lose most of the time (win rate 30% or below).
• Coverage — how many rows of data back it, and when it was last updated (or "Not trained").

Below the tiles:

• KPI trend over time — a weekly, volume-weighted line for the chosen KPI, with an optional benchmark line (needs at least two weeks of data).
• Top consistent winners and Consistent drags — your highest-conviction wins and persistent underperformers (top five each), with dimension, value, win rate, KPI, and a confidence level.
• Spend & primary KPI by objective — spend and the benchmark KPI for each strategy.
• Dimension health — per dimension (placement, device, age, gender, geo, hour of week), the count of winning / losing / mixed slices, coverage, and confidence.
• Top movers — what's shifted most recently versus the benchmark, up or down.

Platforms

Every platform side by side on a single objective, each read on that objective's KPI, with the best performer flagged. Because each platform is judged on its own KPI, you don't compare across different objectives.

• A card per platform shows its headline KPI and spend, with the brand-best platform marked "Best."
• A relative-performance table lists each platform's spend, primary KPI, winner/drag counts, and its single top winner and top drag. A "—" means that platform wasn't run, or has too little data.

Strategies

The headline comparison — two strategies (campaign objectives) side by side. The same audience or placement can win for one objective and lose for another, so the view always compares within a strategy.

• Each side shows spend, the headline KPI, its consistent-winner and consistent-drag counts, up to three top winners and drags, and a "scale X · trim Y" recommendation.
• A "Where the two strategies disagree" table flags slices that win for one objective but drag for the other.

With only one strategy in the cube, you'll see a single column and a note that comparison needs at least two objectives.

Dimensions

Break the brand down by a single dimension — placement, device, age, gender, geo, hour of week, region, or position — for the KPI you pick. Each value's row shows:

• a classification — Consistently winning, Consistently losing, Trending up, Trending down, Mixed, or Insufficient data,
• its KPI value and win rate,
• a consistency bar (win / neutral / loss), and
• a confidence level (High / Medium / Low), with the sample size.

Click any row to drill down. When two or more platforms have data for the dimension, a cross-platform matrix compares them side by side.

Each strategy has one primary KPI — its objective's headline metric (ROAS for a Sales strategy, for example). If you switch the KPI picker to a different metric, Zimmer still shows that metric's value, but it can't rank winners and losers or give a win rate or confidence for it — the classification, win-rate, and confidence columns only apply to the strategy's primary KPI.

Creatives

Your ad creatives ranked across the whole brand.

• Summary tiles: total creatives, consistent winners, consistent drags, and total spend.
• A searchable, sortable table (by spend, KPI, win rate, or name) — each creative's classification, KPI, spend, spend share, win rate, and confidence; click through for detail.
• Top performing and Underperforming creatives (top three each, high-confidence).
• Creative patterns that win — which formats (video, image, carousel, and so on) tend to win for the strategy, with counts, average KPI, spend, and winner/drag tallies.

Forecast

Project performance forward, in two modes:

• Trend projection — project one slice's KPI forward, with a predicted value, a likely range, the per-week trend, and a history-plus-projection chart.
• Campaign composition — model a mix of dimension values to see the blended predicted result.

Forecasting needs a few weeks of history before it can run.

KPIs

Every strategy is read on its own KPI — the metric that defines that objective's success — so the numbers you see follow the strategy you're looking at. The common ones:

Video strategies add view-based KPIs such as View Rate and VCR. For the full metric glossary, see Campaign Insights.

(Two labels you'll see throughout: a classification — Consistently winning, Consistently losing, Trending up, Trending down, Mixed, or Insufficient data — and a confidence level, High / Medium / Low, reflecting how much data backs each slice.)

How it connects

• Set up and refresh the data behind it in Brand Intelligence (Knowledge & Data → Training).
• The cube's winner / drag classifications also feed your optimizer agents' recommendations — so what your brand learns improves what the optimizers propose.
• For a single campaign's detail, see Campaign Insights.