Find the money leaking out of your Meta ads
Connect your Meta account read-only and we’ll score every hour, audience and ad set in it — then show you the one change worth making first.
No sales call required. We never get write access.
For agencies, freelance media buyers, in-house growth teams, and owners running their own spend — one ad account or forty.
Trusted by global brands running Meta ads
The gap
Budget follows volume, not return
Split the last 30 days into four windows and read two columns side by side: the share of budget each window takes, and the return it gives back. Afternoon holds 42.4% of the spend at 1.78x. Evening returns 2.11x on 26.6%. Nothing is broken here — the money is simply sitting where the volume is.
Budget by Time of Day
Share vs return · Last 30 days
Night
12AM–6AM
1.80x
−0.06 vs avg
Morning
6AM–12PM
1.74x
−0.12 vs avg
Afternoon
12PM–6PM
1.78x
−0.08 vs avg
OVER-FUNDEDEvening
6PM–12AM
2.11x
+0.25 vs avg
UNDER-FUNDEDReturn is read against the account average, 1.86x. Only evening sits above it.
Budget health
72 / 100
Pacing is on track.
Account health
59 / 100
Placement is not.
Afternoon takes 42.4% of the budget at 1.78x. Evening returns 2.11x on 26.6%.
The dead hours
Subtract waste before you add budget
Today’s spend landed in seven hours out of twenty-four, and $456.71 of it fell between 12AM and 5AM — a window where almost nothing converts. That money is already yours. Taking dead hours back does not depend on any comparison being right.
Spend by Hour
Which hours carried spend · Today
12AM–5AM · WASTE WINDOW
Coverage, not volume. A bar means the hour carried spend at all — the height is not a dollar amount.
Recoverable waste
12AM–5AM$456.71
Spent where almost nothing converts.
Hours carrying spend
TODAY7 / 24
Five of them inside the waste window.
Seventeen hours got no spend at all. The five that should not have, did.
The evidence
Count the conversions behind the rate
Every comparison rests on a number of conversions, and this account has around fifty. Three sit in the night window against twenty-one in the afternoon. Men 45–54 click at 6.64% on $507 with no conversions reported at all — a hypothesis to test, not a segment to fund.
Evidence Behind Each Split
What is converting · Last 30 days
Conversions by time of day
Night
12AM–6AM
3
TOO FEW TO READMorning
6AM–12PM
11
Afternoon
12PM–6PM
21
Evening
6PM–12AM
15
One dot is one conversion. A window needs around fifty of its own before two rates can be compared.
By age and gender
SpendClick rateMen 25–34
1.98%
Men 35–44
4.28%
Women 25–34
1.72%
Men 45–54
6.64%
Men 45–54 click most on the least spend, with no conversions reported here. A hypothesis, not a target.
A rate with nothing converting behind it is a test to run, not a budget to move.
The other axis
Hours are one axis, not the only one
The same test — share against return — runs on who you reach, not just when. This account reaches 1M against an ideal core of 5M: too narrow to give delivery room to optimise. The widest gaps are where the next edit belongs, once this one has settled.
Audience Health
Nine dimensions vs target · Last 30 days
Current vs ideal
Widest gaps
Cost
30 → 70
Audience
40 → 70
Efficiency
50 → 80
Nine dimensions, each scored 0–100 against the target for this objective.
Reaching
1M
Ideal core
5M
Reach sits at a fifth of the ideal core — too narrow to give delivery room to optimise.
The score is not the point. The distance between the two shapes is.
The move
One edit, then leave it alone
Nothing here asks for a bigger budget. Move $403.17 out of morning and into evening — one ad set, on a lifetime budget so the hours can be scheduled at all. Then stop, and judge it on conversions once the ad set is out of learning.
The Move
One ad set · One edit
Out of morning, into evening. The total budget does not change — only where it sits.
What the edit needs
Then stop
Change one thing
Under 20% of the ad set’s daily budget
Let learning finish
≈50 optimisation events in 7 days
Judge on conversions
Not on tomorrow’s cost per result
Projected recovery
$1,957 / month
The only figure that changes is where the money sits.
Agencies and serious operators
don't scale manually
It starts with a spreadsheet, then a script, then a stretched team. These are the three leaks that show up in almost every account we audit — and the reason none of them get caught by hand.
The CPA Creep
Metric: Efficiency Drift
Micro-fluctuations across high-spend sets. Undetectable by manual review until budget exhaustion.
Invisible Fatigue
Metric: Creative Health
Audience saturation triggers CTR decay. Spend remains constant while conversion probability drops.
The Silent Drop
Metric: Attribution Link
Backend tracking failure causes optimization towards ghost conversions. Immediate capital drain.
A real estate agency running 22 Meta campaigns
CPA drifted from $10 to $16 — silently, over 18 days. Here is what the drift looked like, and where it was caught.
The Situation
The creative had not changed. The targeting had not changed. But frequency had moved from 1.8 to 4.3 and hook rate had dropped 31% — signals that were visible in the data the whole time.
AI Chat Assist flagged the drift on day two. The campaign was paused, the creative rotated, and spend was shifted to the two performing ad sets before the client saw a single report.
Flagged before client noticed
Proactive intervention, not reactive damage control.
Estimated spend saved
Prevented from bleeding into a failing ad set.
Client escalations that month
Trust maintained through silent optimization.
What your team sees before money leaks
Most reporting tools tell you what happened yesterday. AI Chat Assist shows you what is starting to go wrong today — with the context to act on it before spend compounds.
Campaign health
RISK RANKEDCampaigns that need a decision surface first.
Creative fatigue alert
DECAY DETECTEDCTR down 28% over 5 days. Frequency moved from 1.9 to 4.1. Hook rate falling.
Recommended action
Pause Ad 3. Refresh the hook. Move spend to Ad 5.
CPA spike — act now
URGENTCPA moved from $10 to $16 overnight. $500 spent since the signal appeared.
Cause identified
Audience overlap with Retargeting set. Budget redistribution recommended.
Every team has Ads Manager.
Few have intelligence
AI Chat Assist vs a manual campaign manager — side by side across the dimensions that actually affect client outcomes and retention.
High retainer, plus ad spend, paid regardless of consistency
Low flat fee, plus ad spend, with the full system included
Set manually, revisited occasionally based on gut feel
AI audience builder refines continuously from live performance data
Fixed split, manually rebalanced when someone notices an issue
Automatically shifts toward what's converting in real time
Runs on a flat schedule, rarely tuned to actual audience activity
Time-slot optimisation based on when the audience is actually active
Depends on staff availability, often hours or next business day
AI chat assistant responds instantly, 24/7
Manual or bulk-blasted, high risk of spam flags and number bans
Paced broadcasting with automatic throttling to protect the number
Reviewed periodically, often subjectively
Continuous, data-driven creative scoring
Monthly PDF or call, format varies by agency
Account audit, campaign, and weekly reports as standard, every month
Generic templates reused across clients
Content adapts to each business's specific USP automatically
Depends on the individual manager's bandwidth and attention that week
Same level of monitoring and optimisation applied every day, automatically
When something breaks, you get the next move
Most tools send an alert. AI Chat Assist sends a decision package — what changed, why it changed, and the specific action to take.
What changed
Creative fatigue detected. CTR down 28% over 5 days while spend stayed flat.
Why it happened
Frequency increased from 1.9 to 4.1 over the same period. Hook rate is falling. The audience has seen this ad too many times.
What to do
Pause Ad 3. Refresh the hook. Move budget toward Ad 5, which is performing at $7 CPA with 3x lower frequency.
Your media buyer reads it, makes the call, and moves on. No dashboard archaeology. One clear signal with the context to act.
What the campaign intelligence layer does
Four checks running continuously against every active campaign, so the response is automated and your time goes to strategy.
Campaign triage
Every active campaign ranked by deterioration and budget exposure. The ones that need a decision today are at the top.
CPA guardrails
Set your maximum acceptable cost per lead. When actual CPA approaches that threshold, the system flags it with context.
Creative fatigue detection
Tracks CTR, frequency, hook rate, and engagement decay simultaneously. You know the cause and the recommended next action.
Budget risk detection
Spend anomalies flagged early, with the campaign context attached. An explanation of what moved, by how much, and what it means.
Lead Qualification
Meta sees a lead click. It cannot see what happens in the conversation that follows. AI Chat Assist routes every new lead from your Meta campaigns directly into a WhatsApp qualification flow. Before your sales team speaks to anyone, the system has already asked the right questions, scored their intent, and separated genuine buyers from tyre-kickers.
- check_circleWhat the lead actually needs and when
- check_circleWhether they have the budget and decision authority
- check_circleWhich objections came up before they agreed to talk
AI Qualification Bot
Online
The timeline your customers experience:
You see which campaigns are wasting budget and why — with the cause and recommended action.
You see which conversation patterns produce your best buyers. Your sales team talks to fewer leads and closes more.
You have a buyer intelligence model built entirely on your own conversations. No competitor can replicate it.
"A competitor can copy software. They cannot copy the intelligence your business has built."
Spot performance problems before they drain your budget
Continuous signal monitoring
Platform uptime reliability
Average time to first signal
See it in three minutes
Watch the system flag a creative fatigue event, identify the cause, and surface the recommended action — before another rupee of budget went into it.
Trusted by teams that needed
predictable growth
Avtar Singh
CEO, AI Soch Studio
“We were running 18 campaigns across three client accounts. AI Chat Assist flagged a CPA spike on a real estate campaign within 36 hours of it starting. We'd have caught it in the weekly review — four days later. That's the difference between a professional conversation and a difficult one.”
Abhishek Dhakoliya
Marketing Head, Best Prop Deal
“Before AI Chat Assist, creative refresh decisions were gut feel and spreadsheets. Now I can see exactly when frequency is hurting performance and rotate before CTR falls. My clients see more consistent CPA and I spend less time explaining variance.”
Monitoring campaigns across every major ad platform
Questions about
the system
Clear architectural answers on how AI Chat Assist protects and optimizes your operation.
How does budget risk detection work?
AI Chat Assist monitors CPA movement, spend anomalies, creative fatigue, and lead quality signals across your Meta campaigns. When budget is exposed, the platform flags the affected campaign and explains what changed.
What is a CPA guardrail?
A CPA guardrail is a maximum cost-per-lead threshold you set. When actual CPA moves beyond that threshold, AI Chat Assist surfaces the risk early so your team can review the campaign before spend compounds.
How does creative fatigue monitoring work?
The platform watches signals such as CTR decline, frequency increases, hook-rate decay, and CPA movement. When a creative starts deteriorating, it shows the likely cause and recommended next action.
What does lead qualification do?
Every lead passes through an AI qualification layer. The platform asks pre-set questions via chatbot, and only routes high-intent leads to your sales team.
Does AI Chat Assist replace my Meta Ads Manager?
No. AI Chat Assist connects to your existing Meta Ads Manager account - it does not replace it. You keep full access to Ads Manager while adding a campaign intelligence layer on top.
Why not just use Meta alerts?
Meta alerts tell you something happened. AI Chat Assist adds context by explaining why it happened, which campaigns are affected, and what action your team should consider next.