On-site/checkout/shippingMobile Safari visitors abandon at the shipping step
$2.1K – 3.4K/mo at riskEvery client leaks
money. You know first.
Anry reads each client's whole funnel, store, ads, every channel, finds where revenue is slipping out, and prices the fix. You bring the answer to the table, for every store on your books.
HomeOpportunities
Paid (12)
On-Site (3)
Paid (12)On-Site (3)Tools
User Path
Heatmaps
Screen recording
Page view
User PathHeatmapsScreen recordingPage view
Home
Ask anything you’d like... type @ to reference a campaign, an ad or a page
Hi Susan, what are we working on today?
Opportunities for you this week
7 newOn-site/checkout/shipping
Paid
Brand Search17 search terms spent $924 last month and never bought
$2.6K – 3.6K/mo wasted spendPaid
Summer UGC videoRepeat viewers stopped clicking — CTR down 31% in two weeks
$1.1K – 1.6K/mo wasted spendScheduled reports
Any (22)On-site (8)Paid (9)SEO (2)AEO (1)CRM (2)
Stream
Reports
Scheduled
Assigned
On-site
Exit rate • /checkout/shipping
Mobile exits before the payment step, against a 30% threshold.
Daily - 8AM
Sofia Mitchel
Paid
ROAS • Google Ads "Brand Search"
Tracked revenue over synced spend, on a 7-day click window.
Weekly - Monday 9AM
Elias BrackettPaid
Impressions vs clicks • Summer UGC video
Whether clicks keep pace with impressions, against its campaign.
Weekly - Monday 9AM
Lana McKenzieOn-site
Bounce rate • homepage
Where homepage bounce sits against the site’s own baseline.
Mon, Wed & Fri - 9AM
Tony LeeCRM
Revenue per send • welcome flow
Anchored on clicks and attributed revenue, never opens.
Monthly - 1st
Marta DelgadoSEO
Do the "gifting" posts bring buyers?
Traffic, assisted sessions and revenue across the three posts.
Monthly - 1st
Peter HalvorsenAEO
Share of AI answers • gift sets
Which assistants send sessions, and which answers leave you out.
Weekly - Monday 9AM
Stephanie VealPaid
Spend efficiency • Summer Sale
Every ad in the sale, ranked by its ROAS against the campaign’s.
Weekly - Friday 5PM
Adrian KesslerPaiddetectedmonitored
Google Ads · "Brand Search"
campaign · monitored since July 26·snapshot #2 · immutable·July 27, 2026
Archive
ROAS
3.3above your 2.5 target
tracked revenue ÷ synced spend · 7-day
Wasted on dead terms
$924/30dup from $848
17 terms · a sixth of the campaign's spend
Their sessions · purchases
932 · 0
the first 14 · ≈23 expected at your 2.5%
Estimated waste
$2,560–3,560/mo
visible floor $924 · Google hides a shareThe ad · responsive search ad
Admaison-lumen.com/collections/all
Soy Candles by Maison Lumen | Hand-Poured | Free Shipping $49+
Clean-burning soy candles, poured in small batches. Gift sets ready to ship. 30-day happiness guarantee.
Search · ad group "Brand · Exact"8 headlines · 4 descriptions$176/dayWhat's happening
Detection found 17 search terms that spent $924 in 30 days with zero purchases, bouncing at 87%. Fourteen were there when this campaign was first read on July 13; three more appeared last week. The campaign itself returns 3.3, above your 2.5 target — which is how a sixth of its spend goes out on searches that were never going to buy without anything looking wrong.
Why do we believe this?How can I fix it?
ROASBounce (affected groups)Dead-term spendchanges you logged
Jul 5 · change · logged by you“Cut gift-set prices 10% for the summer promo.”Brieuc · logged same dayJul 24 · change · logged by you“Raised the daily budget $152 → $176 for the summer sale.”Brieuc · logged same day
Weekly ROAS, with every change logged on this campaign marked on it. The line holds all eight weeks — a campaign this size carries the dead terms without flinching, which is why nobody finds them by watching it. Bounce and dead-term spend are the other two metrics.
What you should know
Add all 17 as negatives — the list is ready to paste. Ten minutes in the same screen where you raised the budget on July 24 — and at $176 a day the same share of dead spend costs more than it did. Google keeps minting close variants, so expect a few each month. The next snapshot reads bounce and dead-term spend on the affected ad groups against today's and tells you whether it took.
SnoozeView fix suggestion + preview
Finding · wasted terms
Searches that never bought
est. $2,560–3,560 / mo wasted| Search term | Spend (30d) | Clicks | Purchases | Bounce |
|---|---|---|---|---|
| “candle making kit” | $152 | 168 | 0 | 91% |
| “wholesale candles” | $124 | 132 | 0 | 89% |
| “candle supplies bulk” | $96 | 108 | 0 | 88% |
| 11 more terms | $476 | 524 | 0 | 85% |
Visible floor only. Google hides a share of search terms, so the true waste is likely higher than the $924 these two tables can see.
The landing page with every other source
CVR · 30 days
Sessions into /collections/all over 30 days, by the way they arrived.
The first 14 of these terms sent 932 sessions and bought nothing, where organic, email and other paid all land near 2.5% on the same page.
2.4%
Organic search4,2102.7%
Email1,1302.5%
Other paid2,4800%
The first 14 terms932What their sessions did on site
These termsSite averageEach measure against the site's own baseline, not a previous period.
Scroll depth (median)18%58%
Time on page6s42s
Add to cart0%11.4%
New since we flagged it
Same pattern, three fresh terms
$76 · 0 purchases| Search term | First seen | Spend | Clicks | Purchases |
|---|---|---|---|---|
| “candle diy kit” | Jul 24 | $36 | 44 | 0 |
| “bulk candle jars” | Jul 25 | $24 | 28 | 0 |
| “candle wax supplier” | Jul 26 | $16 | 20 | 0 |
History of this campaign
Snapshots are what we measured, and they never change once written. Changes are what you shipped, in your own words, logged here so the next snapshot reads before and after around them. One timeline, and the chart above plots both.
- Jul 27, 2026snapshot#2 · this report · 17 terms · $924 / 30d · still zero purchases
- Jul 24, 2026change“Raised the daily budget $152 → $176 for the summer sale.” Brieuc · logged same day
- Jul 13, 2026snapshot#1 · first detection · 14 terms, $848 / 30d, zero purchases
- Jul 5, 2026change“Cut gift-set prices 10% for the summer promo.” Brieuc · logged same day
Every number above comes from a report component, never from the model.
Agents read this exact document via MCP · get_report(report_id, snapshot: 2)
WHY AGENCIES
The audit is done
before you walk in.
Walk in with the answer
Anry watches every client funnel and prices what it finds. You open the meeting with the leak, what it costs, and the fix ready to ship — not a wall of dashboards.
Your work, on the record
Every fix you ship is remembered, along with what it was worth. The value of your work sits in the client's own numbers, where both of you can see it.
Built for a roster
Every client store under one login, and adding one takes minutes, not a project. The smallest shop on your books gets the same watch as the biggest.
WORKING TOGETHER
Two ways in.
Pick per client.
Partner rate
MOST AGENCIESWholesale
For the book you already carry. You buy at a partner rate and fold Anry into the service you already sell.
Partner pricing that leaves room in your retainer
Every client store under one login
Every brief in your branding
You own the account and the client relationship
One invoice to you, not one per store
Referral commission
LARGE CLIENTSCommission
For clients who'd rather hold the contract themselves. They buy from us directly, you still earn on it.
20% recurring commission for as long as they stay
They're billed by us, so nothing sits on your books
Onboarding and support sit with us
We can run the demo with you or for you
Tracked, so you can see what's landed
Not sure which fits? Bring your client list to the call and we'll work it out per store.
Every client store has a leak.
Be first to it.
Bring your list to the call. We'll run the demo with you and work out the rate per store.