Creative intelligence for performance marketers

See every ad that works. Yours and theirs.

Emberlix scores every ad you run and every ad your competitors run, across all five channels, and tells you what to make next.

Meta, TikTok, Snap, Google and Apple Search Ads. In one place, on one scale, for the first time.

Get a free audit of your account See how it worksNo card. Takes 3 fields.

[data-static] .u-rv,[data-static] .u-stamp,[data-static] .u-pane,[data-static] .u-chip-drift{opacity:1!important;transform:none!important;transition:none!important}[data-static] .u-bar-fill{width:var(--w)!important;transition:none!important}[data-static] .u-plot{stroke-dashoffset:0!important;transition:none!important}

Snapshot 2026-08-12n = 248Vertical UK subscription appsSpend-weighted, 14-day window. Placeholder creative.

This week’s hero rail

Best of the last 14 days, ranked by creative score

#1META2 of 4 ads liveIn-house

Score 87

VID-604yours

Live since 5 Jun 26 (68d, 54d delivering)

Markets London, Manchester

type polishedangle discount-ledhook app-in-handsetting kitchen

“Your first month is on us. Cancel any time — no forms, no phone calls.”

Last 30 days

Spend

£14,800

Installs

2,387

Purchases

143

CPI

£6.20

CPA

£103.50

Install rate

14.26%

CTR

1.35%

Impressions

1,240,000

Reach

486,000

Saturationinstall rate, % of peak, 7 weeks, 100 → 61

Widen

The audience is finished, not the ad

Click-through 94% of peak, install rate 61%. The ad still works; the audience is finished.

Widen the audience. A new cut changes nothing.

Last 14 days: £7,200 spend · 1,161 installs · £102.86 CPA · updated yesterday

Strongest levers: 1.35% CTR, 94% of peak click-through

#2META1 of 3 ads liveInfluencer

Score 72

VID-581yours

Live since 2 Jul 26 (41d, 23d delivering)

Markets London, Birmingham

type ugcangle savingshook direct-addresssetting kitchenstructure caption-hook

“I was wasting £40 a week on food.”

Last 30 days

Spend

£11,400

Installs

1,932

Purchases

106

CPI

£5.90

CPA

£107.55

Install rate

15.08%

CTR

1.42%

Impressions

902,000

Reach

358,000

Holdinstall rate, % of peak, 7 weeks, 100 → 97

Hold

Hook lands, retention drops

72% of people stop scrolling. Only 18% are still watching at the quarter mark.

Recut seconds 2 to 5. Leave the opening frame alone.

Last 14 days: £5,600 spend · 949 installs · £107.69 CPA · updated yesterday

Strongest levers: 72% scroll-stop, 1.42% CTR

#3TIKTOK1 of 2 ads liveInfluencer

Score 54

VID-522yours

Live since 18 Jun 26 (55d, 38d delivering)

Markets Manchester, Leeds, Glasgow

type ugcangle time-savinghook questionsetting living-room

“Meal prep in 10 mins?”

Last 30 days

Spend

£9,600

Installs

2,087

Purchases

CPI

£4.60

CPA

£102.13

Install rate

18.78%

CTR

1.10%

Impressions

1,010,000

Reach

402,000

Watchclick-through, % of peak, 7 weeks, 100 → 52

Recut

Worn-out creative, intact audience

Click-through 52% of peak, install rate 96%. Everyone who taps still converts.

Recut it. The audience is still buying.

Last 14 days: £4,800 spend · 1,043 installs · £102.13 CPA · updated yesterday

Strongest levers: 18.78% install rate, £102.13 CPA

The rest of the wall

The other 9 creatives in this window, scored on the same scale. Every one of them ranks below the rail.

SNAPyours

Placeholder creative, Apple Search Ads, scored 49

ASAyours

Placeholder creative, Google UAC, not testable

n/t

UACyours

METAtheirs

TIKTOKyours

METAtheirs

Placeholder creative, Snap, not testable

n/t

SNAPyours

Placeholder creative, Google UAC, scored 38

UACyours

Placeholder creative, Apple Search Ads, scored 34

ASAyours

Two frames here refuse to score. There was not enough data to be honest about them, so they say n/t instead of a number. That is the product working, not failing.

Why we built this

Your best creative insight is stuck in a meeting.

Creative makes the ads. Performance knows which ones worked. Getting that from one team to the other means digging through dashboards, arguing about what the numbers actually mean, and briefing the next round on a hunch.

So creative builds, waits, finds out, and starts again. Weeks per cycle, and most of what was learned never survives the handover.

Emberlix does that analysis and writes the verdict.

Every ad gets a specific, plain-English read on what worked, what didn’t, and what to change. Your creative team can act on it the same day, without waiting for anyone. Proactive instead of reactive.

The score

Find out exactly where your ad loses people.

An ad can fail in five different places, and each one costs a different fix. A single score out of a hundred tells you it failed. Emberlix tells you where.

Placeholder creative, Meta, United Kingdom, scored 72

VID-581 · META · UKyours

Ranked against the other 214 ads in this account.

Did they stop scrolling?hook

Did they keep watching?hold

Did they tap?click

Did they go through with it?store

Did it cost you sensibly?cost

Worst in your accountBest in your account

The hook works. The next three seconds don’t.

Seven in ten people stopped scrolling for this ad. Fewer than two in ten were still watching by the quarter mark, so almost everyone the opening won was lost immediately after it.

Recut seconds 2 to 5. Leave the opening frame alone.

Measured against your own history

Every ad is ranked against the rest of yours, so a good score means good for you. Switch to your own targets or an industry benchmark whenever you want.

n/t

And it tells you when it doesn’t know

Not enough spend, not enough days, not enough people. Rather than a confident number built on nothing, you get “not testable” and what would make it answerable.

Scored on numbers that hold up

Ad platforms mark their own homework, and the winners they report are not always winners. We check their figures before anything gets scored, so a good score is real. Your attribution tool stays exactly where it is.

Figures shown are an illustrative example, not a specific customer. Creative slots are placeholders, never fabricated ads. The free audit shows you your own numbers, from your own accounts.

A product tour for people who buy media, not software

Your next best creative, decided by what already worked.

Emberlix puts every ad you and your competitors run into one view, scores them on a single scale, and gives you the verdict: what to make next, what to stop, and what a new creative will never fix.

Built by a Head of Acquisition, in anger, against live spend. Every surface below is a real product surface. Creative slots are placeholders, never fabricated ads.

Five tabs become one view.

Meta in one window. TikTok in another. Snap, Google and Apple somewhere else entirely. Nobody has ever seen their creative side by side, because no tool puts these five in the same place.

Emberlix does. Same scale, same rules, one screen.

Most creative tools stop at Meta and TikTok. Snap and Apple Search Ads appear in no creative report, anywhere.

Each channel on its own · UK · 14 days

Every ad scored against your own account.

Not an industry benchmark somebody invented. Each creative is ranked against your cohort on the five things a creative actually controls: hook, hold, click, store, efficiency.

And when there is not enough data to be honest about it, you get n/t instead of a made-up number.

Two frames here refuse to score. That is the product working, not failing.

Top of the rail

#1META2 of 4 ads liveIn-house

Score 87

VID-604yours

Live since 5 Jun 26 (68d, 54d delivering)

Markets London, Manchester

type polishedangle discount-ledhook app-in-handsetting kitchen

“Your first month is on us. Cancel any time — no forms, no phone calls.”

Last 30 days

Spend

£14,800

Installs

2,387

Purchases

143

CPI

£6.20

CPA

£103.50

Install rate

14.26%

CTR

1.35%

Impressions

1,240,000

Reach

486,000

Saturationinstall rate, % of peak, 7 weeks, 100 → 61

Widen

The audience is finished, not the ad

Click-through 94% of peak, install rate 61%. The ad still works; the audience is finished.

Widen the audience. A new cut changes nothing.

Last 14 days: £7,200 spend · 1,161 installs · £102.86 CPA · updated yesterday

Strongest levers: 1.35% CTR, 94% of peak click-through

The other 11 in this window

METAyours

TIKTOKyours

SNAPyours

ASAyours

n/t

UACyours

METAtheirs

TIKTOKyours

METAtheirs

n/t

SNAPyours

UACyours

ASAyours

Not a score. A brief.

A number out of a hundred tells your creative team nothing. Emberlix reads the evidence and writes the instruction, down to the second of the edit.

This is the part that earns its keep: a creative's whole signal history turned into one sentence a designer can act on this afternoon.

Hook is landing. Retention is not. Fix seconds 2 to 5.

Placeholder creative, Meta, UK, scored 72

VID-581 · METAyours

Channel: Meta
Type: UGC
Angle: Urgency
Producer: Influencer-produced
Delivery: Partial
Days live: 23 of 41 meaningful
Live ads: 1 of 3
Cohort: 214 ads

Hold · fix retention · high confidence

Hook lands, retention drops

72% of people stop scrolling. Only 18% are still watching at the quarter mark.

Keep it running if you need the volume. Recut seconds 2 to 5. Leave the opening frame alone.

Percentile in this accountranked against 214 other ads

hook

hold

click

store

cost

Worst in your accountBest

“I was wasting £40 a week on food.”

type ugcangle savingshook direct-addresssetting kitchenstructure caption-hook

Last 30 days

Spend

£11,400

Installs

1,932

Purchases

106

CPI

£5.90

CPA

£107.55

Install rate

15.08%

CTR

1.42%

Impressions

902,000

Reach

358,000

Holdinstall rate, % of peak, 7 weeks, 100 → 97

Last 14 days: £5,600 spend · 949 installs · £107.69 CPA · updated yesterday

The refresh you were about to brief will not work.

When performance drops, everyone assumes fatigue and orders new creative. Half the time the creative is fine and the audience is finished. New assets change nothing and the money goes anyway.

Emberlix separates the two, because they look identical on a dashboard and cost completely different things to fix.

Clicks holding, conversions falling. That is saturation, not fatigue.

4 creatives are saturating, not fatiguing. £18,400 of spend is affected.

VID-604 · click-through against install rate, each indexed to its own peak

100%50%

30 Jul12 Aug

Click-through, holding at 94% of peakInstall rate, down to 61% of peak

Creative · mechanism · evidenceSpend · CPI

87
VID-604saturatingMETA2 of 4 live£7,200£6.20
Click-through 94% of peak, install rate 61%. The ad still works; the audience is finished.
Widen the audience. A new cut changes nothing.
51
VID-733saturatingSNAP3 of 5 live£5,100£4.25
Same split as VID-604: clicks holding, installs falling away.
Widen the audience. A new cut changes nothing.
54
VID-522fatiguingTIKTOK1 of 2 live£4,800£4.60
Click-through 52% of peak, install rate 96%. Everyone who taps still converts.
Recut it. The audience is still buying.
46
VID-497healthyMETA2 of 2 live£3,300£5.50
Neither rate has decayed against its own peak.
Leave it alone.
n/t
VID-612n/tASA1 of 1 live£1,400£3.68
9 days of meaningful delivery, below the 14-day floor.
No call yet. It gets one at day 14.

2 further saturating creatives make up the remaining £6,100. The bottom row is the product refusing to name a mechanism it cannot see yet — it still reports the money, because the ad still spent it.

What to make next, in your words.

Every creative is tagged by angle, hook and type, and those tags follow it across every channel and market it runs in. So you stop asking “which ad won” and start asking “does urgency beat discount for us”.

The answer becomes the brief for the next round.

Your creative team gets a pattern, not a leaderboard.

TagAdsHookStore

Angle

✓ urgency687461

discount544358

social proof615866

problem-first415149

Format

✓ ugc827168

scripted476641

partnership395552

static333844

talent-led236257

Structure

✓ offer by :02576963

Hook and store are PERCENTILES — the median rank inside this account for the creatives carrying that tag, out of 224 tagged. Urgency leads discount on hook by 31 points.

24 creatives carry no tags: Google UAC assets. The channel reports at asset level and exposes no ad-level detail, to us or to anyone.

Generated brieffrom 248 scored assets · 2026-08-12

Lead with urgency. It beats discount by 31 points on hook across all five channels.
Keep it UGC, not scripted. Scripted holds attention but does not convert on store.
Front-load the offer by second 2. Your best performers all do.
Nothing new needed for Snap. Two assets there are saturating, not fatiguing.

Everything your competitors ran, including what they deleted.

Ad libraries only show you what is live right now. Emberlix has been watching every week and keeping it, so you can see the ad they pulled last month and the one they have quietly run for 142 days straight.

That archive gets more valuable every week, and it cannot be bought.

The greyed frame is an ad Meta no longer serves. We still have it.

Activity and creative strategy only. No competitor spend data exists on any channel, for anyone, and no tool has it.

Advertisers

Rival A34

Rival B21

Rival C12

Active ads: 34
Launched this week: 9
Proven, 90+ days: 6
Dominant theme: “free trial”

Format mix71% video of 34 live

video 24static 7carousel 3

Coverage11 of 11 weeks captured, none missed

FramesRival A · theirs · live date from the platform’s own library, pull date from our weekly capture

proven

Live from 23 Mar 26 · 142d

“Cancel whenever you like. We will not make it difficult.”

proven

Live from 16 Apr 26 · 118d

“The only app my whole team actually opens.”

new

Live from 6 Aug 26 · 6d

“Dinner, decided. Plans from £4.99/week.”

gone

Ran 8 May 26 – 22 Jul 26 · pulled

“Half price for life if you join before Sunday.”

The same data runs your reporting.

Once every channel is in one place on one definition, creative is only the first thing you can ask. The same layer gives you channel performance, market by market — never one blended figure with two different stories inside it.

One view. Your creative team and your board read from it.

Same figures in the creative view, the channel report and the board deck. Nobody has to ask which one is right.

United Kingdom23,433 installs · £121,000

ChannelSpendInst.Purch.CPICPA

Meta£58,0009,240553£6.28£104.88

TikTok£27,0005,743287£4.70£94.08

Snap£15,0003,620194£4.14£77.32

Apple£12,0003,270234£3.67£51.28

Google£9,0001,56096£5.77£93.75

All channels£121,00023,4331,364£5.16£88.71

Australia6,290 installs · £34,500

ChannelSpendInst.Purch.CPICPA

Meta£16,4002,510148£6.53£110.81

TikTok£7,6001,53082£4.97£92.68

Snap£4,20094047£4.47£89.36

Apple£3,60088062£4.09£58.06

Google£2,70043027£6.28£100.00

All channels£34,5006,290366£5.48£94.26

CPA is judged against the account’s payback ceiling of £100.00; CPI against each channel’s own. A dotted rule is drifting, a solid orange rule is out of band, and every banded figure says which it is out loud for a screen reader.

Blended, Meta’s cost per purchase reads £106.13 and only drifts. Reported separately, Australia is £110.81 — out of band — and the UK is £104.88. One figure would have hidden the market that needs the work.

Creative detail is filtered to UK, so the UK rows itemise the creatives the rest of this page names and the AU rows report at channel level. 24 Google UAC assets itemise on neither: the channel reports at asset level and exposes no ad-level detail, to us or to anyone.

What we can and cannot see.

Three columns, not one status. Where the data does not exist, the cell is blank and says why. If competitor creative on Google or Apple is what you came for, stop here. Nobody has it, including us.

Scroll the table sideways to see all four columns.

Emberlix channel coverage: your creative, competitor creative, and attribution depth across five channels
Channel	Your creative	Competitor creative	Attribution depth
Meta	full · 5 signals	EU and UK only¹	full
TikTok	full · 5 signals	EEA only, no spend	SKAN-attributed
Snap	full · 5 signals	no library exists	30-day settle
Google UAC	full · 5 signals	no commercial API	black box
Apple Search Ads	3 signals · CPP²	no library exists, for anyone	native

1Meta returns non-political ads only for EU or UK reach, under DSA Article 39.
2Apple Search Ads creative means Custom Product Pages. There is no video unit, so hook and hold do not exist on this channel.

Why this exists

Built by a marketer who got tired of guessing.

This was not built by an engineering team imagining what media buyers need. It was built by a Head of Acquisition Marketing to answer questions that were costing real money, and it has been running against live spend across five channels and two markets ever since.

If you run an agency

Every client in one place, per-client workspaces, and an audit you can put your name on and send to them. Your creative team gets patterns instead of a spreadsheet.

If you run acquisition in-house

The channels your tools skip, the numbers your platforms inflate, and a second opinion on the agency you are already paying.

What you get free

An audit of your account: where your reported numbers and your real numbers part company, how much spend is buying users you already had, and a scorecard of your creative against your competitors’.

What we will not do

Give you a confident answer we cannot support. Every report ends with what it could not work out, and why. That section is the point of the product, not an apology for it.

Find out exactly where your own ads lose people.

We score your recent creative and send it back: where each ad wins, where it loses people, and what to change. You also get the pattern across your channels, how your work sits next to your competitors', and a straight note on anything there was not enough data to judge honestly.

This is a request, not a signup. It emails a person, who reads it and replies. Nothing connects to your ad accounts and nothing runs automatically. If your account is not a fit, we will tell you that and tell you what is.

What it costs

The audit is free and there is nothing to cancel. Product pricing is published in full on this site the day the product opens: every tier, every usage limit, on the page. Nothing below enterprise will say “contact sales”.

See every ad that works. Yours and theirs.

This week’s hero rail

The rest of the wall

Your best creative insight is stuck in a meeting.

Find out exactly where your ad loses people.

Measured against your own history

And it tells you when it doesn’t know

Scored on numbers that hold up

Your next best creative, decided by what already worked.

Five tabs become one view.

Every ad scored against your own account.

Not a score. A brief.

The refresh you were about to brief will not work.

What to make next, in your words.

Everything your competitors ran, including what they deleted.

The same data runs your reporting.

What we can and cannot see.

Built by a marketer who got tired of guessing.

If you run an agency

If you run acquisition in-house

What you get free

What we will not do

Find out exactly where your own ads lose people.

What it costs

Request an audit