The Meta Audience Analyzer's waste score now thinks like the Creative Analyzer's verdict
Each segment is judged on what its underlying ad sets are actually optimizing for, not blindly CPA. A "Female 25-34" segment in a Landing-Page-Views campaign is now scored on cost per LPV — not on a CPA the campaign was never trying to deliver. CPC for link-click campaigns. CPM for awareness and reach. Cost per ThruPlay for video. Cost per lead for lead-gen. Each segment is compared against peer segments in the same optimization-goal bucket, not against an account-wide average that mixes apples and oranges.
The pill now carries a confidence rating. A segment whose call is rock-solid — large peer group, single optimization goal, ad sets out of learning, plus an independent audit flag pointing the same way — gets a solid pill. A segment with a smaller peer group, a mix of optimization goals across its underlying ad sets, or ad sets still in Meta's learning phase gets a dashed border around the pill. Thin data gets 70% opacity. So a glance at the column tells you which scores to act on right now and which to revisit once more data is in.
Two signals agreeing locks the verdict in. When a segment is both cost-expensive vs peers AND already flagged by Pace's Targeting Audit (e.g. an Audience Network placement that's burning budget without converting), the waste score escalates to high-confidence Underperformer regardless of how small the peer group is. Two independent signals saying the same thing is enough.
Zero-outcome handling is per-goal now. A Landing-Page-Views ad set spending $50 in a segment that delivered zero LPVs is correctly flagged as wasteful — with the dollar amount in the tooltip. Previously the waste score's zero-handling assumed CPA semantics regardless of what the campaign was actually optimizing for, so non-conversion segments could slip through.
Hover tooltip explains the call in plain language. Instead of a generic 4-signal score breakdown, you get a single sentence — "Cost per LPV of $12.40 is 280% above peer avg of $3.27 — and this segment is also flagged by the targeting audit, two signals pointing the same way" — followed by the confidence rating. So you can act, or skip, without translating numbers in your head.
⚡ Performance
Meta Creative Analyzer's performance verdict actually means something now
The badge now judges each ad on what its campaign is actually optimizing for. Until now Pace compared every Meta ad to an account-wide CPA average, even ads in campaigns optimizing for landing page views, link clicks, ThruPlays, or reach. So a Landing-Page-Views campaign would get rated against a conversion metric it was never trying to deliver. Pace now reads each ad's optimization goal — Conversions, Leads, Landing Page Views, Link Clicks, ThruPlays, Reach, Engagement, App Installs — and judges it on the matching cost-per-outcome. Then it compares each ad against its peers in the same bucket, not a global mix. So an LPV ad is rated against other LPV ads, a leads ad against other leads ads.
Ads that spent money and got nothing no longer show up as Top Performer. A bug was treating "zero conversions" as "$0 CPA, infinitely good" — so an ad that burned $97 without converting once would sometimes get a green badge. The verdict is now the opposite: "Spent $97 without any conversions" in the underperformer treatment. Ads with too little spend to judge fairly show "Not enough spend to evaluate" instead of being forced into a verdict.
Meta's own quality diagnostic is layered into the call. The Quality, Engagement, and Conversion rankings we surfaced last release now feed into the verdict. When the cost signal and Meta's diagnostic agree, the badge gets a confident, full-strength treatment. When they disagree the badge softens and the tooltip explains why — e.g. "Expensive vs peers but Meta rates the creative above average — could be a targeting or audience issue rather than the ad itself," or "Cost is in line with peers but Meta rates the creative below average — may not scale." You see at a glance whether the issue is the creative, the audience, or noise.
Confidence shows up in the badge styling. High confidence (cost and Meta agree) renders as a solid full-strength badge. Medium (signals partially disagree, or only one of the two is available) renders with a subtle dashed border. Low (not enough peers in the same optimization-goal bucket, or thin spend) renders at 70% opacity. So a glance across the column tells you which verdicts to trust the most.
✨ Improved
Sharper Meta Improvements — Meta's own ad rankings, frequency-aware waste, 180-day lookback
See Meta's own creative rankings inside Pace. The Meta Creative Analyzer now shows three new columns — Quality, Engagement, and Conversion — pulled straight from the same diagnostic Meta uses in Ads Manager. Each cell tells you how an ad stacks up against the others competing in its auctions: above average in green, bottom 10% in red, with the full label in a tooltip. Sort by Quality and the ads quietly dragging your costs up surface to the top, ready to pause or replace. Ads still building up traffic show a dash — Meta needs at least 500 impressions in the last 7 days before it'll rank one.
The waste score now knows when an audience is burnt out. Until now a Facebook placement showing the same person an ad nine times over scored the same as one reaching fresh users at the same CPA. The Audience Analyzer now pulls frequency for every segment, factors it into the waste score, and shows a Freq column on the table. Anything at frequency 4 or above is highlighted in amber so you can spot saturated placements at a glance, and they sort closer to the top of the table because they're the ones where extra spend is genuinely wasted.
Pull half a year of audience data, not just 90 days. A new "Last 180 days" option sits next to the existing 30 and 90-day presets on the Meta Audience Analyzer. Useful for smaller-spend accounts where 90 days doesn't have enough volume to be confident about a demographic, placement, or device-level call. Same waste-scoring logic, just with more history behind every number.
🚀 New
Forecast any metric — clicks, impressions, conversions, CPA, or ROAS
Pick the metric you actually care about and Pace re-fits the saturation curve around it. A new metric switch sits next to the lookback selector on the forecast tab. Choose Conversions, CPA, Clicks, CPC, Impressions, CPM, Revenue, ROAS, or Spend, and the saturation chart, KPI strip, and scenario cards all reshape to that lens.
ROAS now reflects audience-quality decay, not just changing CPA. Pace fits a direct revenue saturation curve from your last 30/60/90/180 days of daily history, so as you scale spend the chart catches both rising acquisition cost and falling per-conversion value as one signal — exactly the kind of decay a constant-multiple ROAS forecast would miss.
Each metric gets its own curve. Behind the scenes Pace fits four independent saturation models per platform — one for cost per conversion, one for cost per click, one for cost per 1,000 impressions, one for cost per dollar of revenue — picking power-law or Hill per metric based on which fits your history best. Switching the chart's primary metric is instant; nothing has to refetch.
The AI overlay reads the right curve for each metric. When the AI adjusts a scenario it now considers each metric's own saturation evidence — clicks aren't pulled along by a flattening CPA, and impressions aren't moved by revenue decay. Each metric gets its own honest answer.
Scenarios you've already saved still open cleanly. Older scenarios default to the CPA view they were built around. Nothing to migrate; nothing changes unless you reach for the new metric switch.
🚀 New
Eighteen new Meta columns at every level — campaigns, ad sets, and ads — fetched only when you ask for them
Add the Meta metric you actually want, not the one we picked for you. The campaign table, the ad-set table, and the ad table on Meta accounts now each offer eighteen extra columns through the column manager: video completion rates at 25%, 50%, 75% and 95%, average watch time, video plays, cost per lead, cost per purchase, cost per landing page view, cost per unique click, cost per link click, link click CTR, outbound clicks and outbound CTR, unique clicks and unique CTR, post saves, and post reactions.
They're off by default. You turn on what you need. Until you tick a column in the column manager, Pace doesn't ask Meta for it — every refresh stays as small as your view. That keeps the Meta API call light on big accounts and means Pace stops paying for fields nobody is reading.
Switching a column on doesn't make you wait for a full refresh. If your data is already on screen, Pace fires a single small call to Meta for just the new metric, just for the rows you can see. A subtle skeleton fills the new cells until the value lands — usually a second or two — then the column populates. From the next regular refresh onward, the metric is part of the standard call and reloads with everything else.
Each table remembers its own selection. Picking Cost / Purchase on the campaign table doesn't force it onto the ad-set table — they're independent so you can tune each level to the question you're actually asking there.
Everything you already had keeps working. Your existing column choices are preserved. Spend, Link Clicks, CTR, CPC, Conversions, ROAS, Reach, Frequency — all unchanged. Cross-platform totals, Sparks, budget pacing and reports are untouched. This is purely the Meta tables getting wider and lighter at the same time.
🚀 New
Meta Clicks now match what you see in Meta Ads Manager
Meta now reports Link Clicks separately from Clicks (All). Link Clicks counts only the clicks that took someone to your destination URL. Clicks (All) is Meta's broader number that also includes likes, comments, and profile clicks. Both are available; Link Clicks is the one shown by default in Meta tables.
Meta CTR, CPC, and conversion rate now use Link Clicks. That makes those figures match what Meta Ads Manager shows in its Link Click columns, which is what most teams reason about when judging creative.
Cross-platform dashboards compare clicks like-for-like. Blended click totals across Meta, Google, LinkedIn, and Microsoft now sum destination clicks across every platform, so portfolio-level CTR and CPC are a fair comparison instead of a mix of definitions.
If you tracked total interaction volume, Clicks (All) is still there. Add it from the column manager on any Meta table, or include it in a report. Sparks and budget pacing now reason about Link Clicks for Meta, which keeps Pace's recommendations in line with the rest of your platforms.