Every morning, Sentrum’s diagnostic engine has already read yesterday — spend, efficiency, creative decay, mix shifts, the edits that were made — and written the verdict: what changed, why, what it’s worth, and the first move. A separate automated anomaly sweep runs each day, severity-gated so it only speaks when it matters. An analyst added to your team, not a replacement for it.
The same Meta OAuth you’ve done a hundred times. Sentrum never touches budgets or bids and never changes anything without your explicit approval. From the first night, the account gets read while you sleep.
Soren reads yesterday against your account’s own baselines, in diagnostic order: spend, results, efficiency, the supporting metrics, mix shifts, and the edits that were actually made. Signal separated from noise before the verdict is written.
One sentence on the dollar figure, the do-this-first move, and the top findings with their evidence. Severity gates decide what deserves your morning; everything below the threshold is logged, not alerted. A separate automated anomaly sweep runs later the same day on the same gates.
Each answer follows the same contract: what changed, quantified; why, with the driver named; what it’s worth; and the move tied to the diagnosed cause. No vibes, no “the algorithm is learning.”
The morning verdict names the driver, not just the delta — spend pacing, auction pressure, creative resonance, conversion quality, or a mix shift — and ties the first move to that cause.
The most expensive misdiagnosis in the job. Soren reads the decay curve, frequency, and CPM together, so a delivery shift never gets blamed on the creative — or the reverse.
Every overnight read reconciles performance against the edits that were actually made — budget changes, new creatives, paused ad sets — so credit and blame land on the right change.
A separate automated sweep runs every day and flags any metric that has broken from its baseline — with the driver attached. The gate keeps the channel quiet enough that when it speaks, you move.
Your stack can already ping you. The difference that reaches a client is whether the ping arrives with the why, the worth, and the move — and whether it knows when to stay quiet.
Alert tools in this category fire on fixed benchmark thresholds — a number crossed a line, here’s a ping. Thresholds don’t know your account: they page you on noise and sleep through slow decay. The Pulse reads moves against your own baselines and models each concept’s decay curve, so the flag arrives when the curve bends, not after the revenue line does.
Even the category’s flagship creative-analytics tools lack proactive fatigue alerts, per their own reviewers — fatigue is something you discover in the chart, after the spend. Sentrum’s overnight run checks every live concept against its decay curve, every night.
Spend-tiered pricing is the category norm: caps at $500K–$1M managed and surcharges per extra $100K of spend. Sentrum is flat monthly pricing that never scales with your ad spend — and the overnight run is part of every plan, every night.
Other tools watch your metrics. Soren reads your account.
2 ad accounts · daily diagnostics
5 ad accounts · Ask Soren follow-ups
15 ad accounts · client-ready reports
Monitoring shouldn’t be spend-tiered. No spend caps and no surcharges for a good month — the overnight run is part of every plan, every night, flat.
See full pricingNo. The overnight run reads your data, writes its verdict, and changes nothing. The only changes Sentrum can apply are ones you explicitly approve — a pause or resume, a schedule change, or a new ad created as a PAUSED draft — never budgets, bids, or targeting. There is a standing never-touch-budgets rule in the product itself.
Every night. Soren reads yesterday against your account’s own baselines and the verdict is waiting before the workday starts — what changed, why, what it’s worth, and the first move. A separate automated anomaly sweep runs each day and flags any metric that has broken from its baseline, so the big moves surface on the next sweep instead of in next week’s review.
No — that is the design constraint the whole surface is built around. Every flag passes a significance gate before it speaks: moves of roughly ten percent or more against your baseline, with real spend and real results behind them. Anything below the gate is logged, not alerted, and there is no praise without conversions — a campaign with zero results never gets a “great job.”
One sentence on the dollar figure — what changed and what it’s worth — then the do-this-first move, then the top overnight findings as expandable rows: the evidence behind each one, the driver it traces to, and the action tied to that cause. Every number is the engine’s own output; nothing is written for effect.
The verdict says so, honestly. If nothing crossed the significance gate overnight, the Pulse tells you the account held and shows you the standing watch items — no manufactured urgency, no invented findings. A quiet verdict you can trust is the point of having gates at all.
No — they are two altitudes of the same engine. The Daily Pulse is the operating layer: what changed since yesterday, flagged and explained. The Account Audit is the strategy layer: the day-one scrub and the QBR, reading structure, creative mix, audiences, and budget across months. Mornings belong to the Pulse; the big questions belong to the audit.
A flat monthly price per tier, depending on how many ad accounts you run — see the pricing page for current plans. The overnight run is part of every plan, every night. Pricing never scales with your ad spend: no spend caps, no per-$100K surcharges.