Find profit leaks. Approve the fix. See what changed.
Sentinel tells you which ads are actually making money using store orders, not platform ROAS. It stages the next move. You approve. It proves what changed.
Set up in minutes · Free monitoring during beta · Actions follow your approvals and rules
Animated example · Sample data · No account changes
Your next move.
Already in the conversation.
Spot your winners, check the evidence, and approve a budget increase in Slack. Next, the same decision by text. Your dashboard keeps the receipt.
Why the numbers hold up
Your ad platforms claim more revenue than exists.
Meta, Google, and TikTok each grade their own homework, and the same order can be claimed by more than one platform. Sentinel matches ad activity to store orders and keeps the three numbers apart. That is the engine underneath every leak, fix, and receipt.
How Sentinel verifies revenue →$6,690 gap between platform-reported and Sentinel-attributed revenueSAMPLE DATA
Unmatched to any channel: $3,725 (12% of store revenue). Model: last click · 7-day click window. Attribution assigns credit; it does not measure incremental lift.
Memory that compounds
It remembers what worked for your store, and shows you.
Every action is remembered with its outcome. A generic model on platform data cannot know your economics. Sentinel does, and it gets sharper the longer it runs.
Memory you can inspect
Every fix, its outcome, and whether it is worth trying again. Nothing is a black box.
The engine: your pixel, real orders
Connect ad activity to store orders. Inspect matched revenue and keep unmatched orders visible.
Product illustrations · Sample data
Verified profit on every ad, not just every campaign
A 3.1× ROAS ad can still lose money. Sentinel knows which ones do.
Every channel matched to real store orders by Sentinel's own pixel, then scored on contribution after COGS, shipping, fees, and returns. The same numbers your media buyers and the agent both act on.
You stay in control
Your rules. Your approval. Your account.
You approve every fix
Each one is staged with its evidence and waits for a click, in the app or in Slack.
Rules in plain English
"Pause any ad over $100 with 0 purchases." Write it once, Sentinel handles the rest inside that line.
Every action has a record
What changed, who approved it, and what happened afterward. Review the evidence and outcome in the action history.
Works with your agent too
Claude, ChatGPT, or your own agent can ask Sentinel directly.
Ask about your brand in the AI you already use. Connect it to Sentinel through MCP, and it can answer with your store orders, ad performance, and profit data. You choose which brands it can access.
Why did revenue drop this week?
Revenue fell 20%, from $20,000 to $16,000. You had fewer orders, while the average order value stayed the same.
Traffic held steady, but fewer visitors bought. I’d check where shoppers dropped out before changing ad budgets.
Source: Sentinel · Demo store · Last 7 days vs. previous 7 daysPricing
Free to watch. Paid to fix.
Monitoring and verified attribution are free while Sentinel is in beta. Operate is $199 per brand per month at the founding-member rate, after a 14-day free trial.
- Verified revenue, matched to store orders
- Hourly monitoring with Slack and email alerts
- Pixel health and tracking diagnostics
- Everything in Monitor
- Diagnoses and staged fixes, ranked by dollars at stake
- Approvals, plain-English rules, and the audit log
- Outcome tracking and brand memory
- Daily briefs, creative intelligence, and Ask Sentinel
- Everything in Operate, per brand
- Agency overview with clients ranked by urgency
- Team roles and per-brand isolation
- Weekly client proof reports and assisted onboarding
$199 per brand per month is the founding-member rate, and founding members keep it. No card needed for Monitor. Cancel anytime. Billing runs on Stripe.
Questions, answered
Is this another attribution dashboard?
No. Attribution is the substrate, not the product. Dashboards stop at "what happened". Sentinel diagnoses the leak, quantifies it in dollars per day, stages the fix, waits for your approval, runs it inside your rules, and measures what changed at 24 and 72 hours. If you already have a dashboard, it becomes one of Sentinel's inputs.
Does Sentinel change my ads on its own?
No. Every fix is staged with its evidence and waits. It runs only after a person approves it, or when a plain-English rule you wrote explicitly covers that case, such as "pause any ad over $100 spend with 0 purchases". Every run is logged with who or what approved it.
What does "verified revenue" mean?
One order, counted once, matched to the clicks and touches that actually happened. Your store's order feed is the ground truth, not the ad platform's own conversion count.
Do I have to leave my current attribution dashboard?
No. Keep it. Sentinel reads it as one more signal and sits above your existing stack. Your dashboard becomes an input, not the answer.
How long until the first verdict?
Connecting Meta, Google, and your store takes about four minutes. The first hourly check runs within the hour, and the first morning verdict arrives the next day once Sentinel has a baseline for your brand.
Is my data isolated from other brands?
Yes. Each brand keeps its own connections, pixel, rules, targets, and Slack channel. Team members see only the brands they were granted, with owner, member, or viewer roles per brand.
What happens when the beta ends?
Monitor is free during beta. We will communicate any changes before they take effect. Operate is $199 per brand per month at the founding-member rate, after a 14-day free trial. Founding members keep the $199 per brand per month rate. The trial converts only if you choose to.
A clearer verdict for your next move.
Connect your ad accounts and store. Give Sentinel the context to build your first verdict.