THE CAMPAIGN DEEP DIVEANY CONVERSION GOAL · READ-ONLY

What happened, why, and what’s next at every layer of the campaign.

Open any campaign mid-flight and Sentrum walks the layers the way a senior buyer does — ad sets, creatives, demographics, placements, the optimization log — answering the same three questions at each one. Built on your campaign’s actual conversion goal, whether that’s purchases, leads, installs, or ThruPlays, with every claim gated on statistical significance.

Open my first deep diveWalk a demo campaign
READ-ONLY ACCESS · 7-DAY TRIAL · EVERY LAYER, EVERY PLAN
OpenWalk the layersLeave with the move
DEEP DIVE · PROSPECTING — COREMID-FLIGHT · 14 DAYS
SPEND
$24.3K−5%
PURCHASES
812+12%
CPA
$29.93−8%
Ad setsCreativesDemographicsPlacementsOpt. log
WHAT HAPPENED
CPA improved 8% while spend eased 5% — efficiency gained, not volume lost. Purchases +12% period over period.
WHY
24% of spend shifted to the 25–34 segment at $24 cost-per-result vs $34 campaign-wide, after the Feb 1 pause removed the lowest-CVR audience.
WHAT’S NEXT
Scale the 25–34 ad set +$150/day — its saturation curve shows headroom before the next CPM step. Leave the Feb 5 budget raise in place; it added volume at flat CPA.
FUNNEL: IMPRESSIONS → CLICKS → LPV → ATC → PURCHASEBUILT FROM THIS CAMPAIGN’S OBJECTIVE
HOW THE DEEP DIVE WORKS

Mid-flight answers, not post-mortems.

STEP 01 · OPEN
Pick any live campaign. The read is already current.

One read-only connection keeps every campaign synced. When a number looks off — or a client asks — you open the Deep Dive mid-flight, not after the damage report.

READ-ONLY · NEVER-TOUCH-BUDGETS RULE
STEP 02 · WALK THE LAYERS
Ad sets, creatives, demos, placements, edits — in order.

Each layer answers the same three questions: what happened, quantified against the prior period; why, with the driver named; what’s next, tied to that cause. The diagnostic order a senior buyer runs — spend, results, efficiency, supporting metrics, mix shifts, edits.

WHAT HAPPENED · WHY · WHAT’S NEXT — PER LAYER
STEP 03 · LEAVE WITH THE MOVE
Actions that name the lever, not “test new creative.”

Every recommendation traces to a diagnosed cause and names the Ads Manager lever to pull — scale this ad set, rotate that concept, revisit Tuesday’s budget edit. If the data is too thin to call, the Deep Dive says that instead.

CAUSE-TIED ACTIONS · SIGNIFICANCE-GATED
WHAT THE DEEP DIVE ANSWERS

The questions a campaign raises while it’s still spending.

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. If the sample is too thin to call, it says so.

Funnel decomposition
STAGE-BY-STAGE
“Good days and bad days look identical — why do sales fall off a cliff?”

When CPC and CTR hold while revenue drops, the answer lives in the funnel, not the ad metrics. The Deep Dive walks every stage of your campaign’s real conversion path and points at the one that moved.

Click metrics held; mobile checkout CVR fell 19% week over week. The funnel says it’s the site, not the ads — fix the product page before touching a single budget.
Layer attribution
SIGNIFICANCE-GATED
“Which ad set is actually carrying this campaign?”

Winner and loser get called only past significance gates — real spend share, real volume. Thin samples are labeled for what they are, so you never scale a fluke or kill a prospect.

Ad set B leads at $26 CPA on 38% of spend — a proven winner. Ad set D’s $19 CPA sits on 3% of spend: promising, not proven. Give it budget before you give it credit.
Edit reconciliation
CAUSE → EFFECT
“We changed five things last week — which one worked?”

The optimization log lines every edit up against the KPI line — budget changes, new creatives, pauses — so credit lands on the change that earned it, and the next edit is informed, not superstitious.

The Feb 5 budget raise ($400 → $500/day) added volume at flat CPA. The efficiency gain traces to the Feb 1 pause — the new creative is still unproven at 4% of spend.
Objective-true funnel
ANY CONVERSION GOAL
“Does this work for lead gen — or is it built for e-comm?”

The funnel is built dynamically from your campaign’s own optimization event. Sales, leads, installs, ThruPlays — Results always mean what the objective optimizes for, and the verdict metric matches.

A lead-gen campaign reads impressions → clicks → landing-page views → leads, with CPL as the verdict metric — ROAS never gets slapped onto a lead objective.
THE DEPTH GAP

A table is not an understanding.

Your stack can already sort campaigns by ROAS. The questions that decide the next dollar — which layer moved, why, and what now — need the campaign opened, not summarized.

Typical dashboard
The Sentrum Deep Dive
Altitude
One table per campaign — sort by ROAS and squint
Layer by layer: ad sets, creatives, demographics, placements, edits — each with what happened, why, and what’s next
Funnel
A fixed e-commerce funnel, whatever your objective
Built from your campaign’s optimization event — purchases, leads, installs, and ThruPlays each get their real funnel and their real verdict metric
The why
CPA went up — here’s a chart
CPA +21% decomposed: CPM flat, mobile checkout CVR −14% — the funnel points at the landing page, not the auction
Winners
Crowns the best ROAS cell, even on 12 clicks
Winner called only past significance gates — thin samples are labeled promising, not proven
Advice
“Test new creative” on every problem
Actions tied to the diagnosed cause, naming the Ads Manager lever to pull
FIELD NOTE

The market has learned to smell AI-wrapper diagnostics: buyers describe tools that invent benchmarks with no source, treat 30 clicks like a verdict, and apply ROAS to lead-gen objectives. The Deep Dive is gated the other way — significance thresholds before any claim, the KPI your objective actually optimizes, and the evidence shown under every finding.

FIELD NOTE

Spend-tiered pricing is the category norm: caps at $500K–$1M managed, surcharges per extra $100K of spend, and credit meters on analysis — every deep question costs extra. Sentrum is flat — $19/$39/$79 — open every campaign, every layer, every day.

FIELD NOTE

Several AI ads tools now request write access at connect, and some change budgets and bids autonomously. Sentrum is read-only by architecture with a standing never-touch-budgets rule — the Deep Dive diagnoses; the account stays yours.

Dashboards summarize campaigns. The Deep Dive understands them.

SEE IT RUN

One campaign, every layer, start to move.

90-SECOND LAYER WALKTHROUGH · RECORDED ON A DEMO ACCOUNT
PRICING

Flat pricing. Every layer included.

Starter
$19/MO FLAT

1 ad account · daily diagnostics

Pro
$39/MO FLAT

3 ad accounts · full method library

Agency
$79/MO FLAT

10 ad accounts · client-ready reports

Some tools meter analysis with credits, so every deep question costs you. Sentrum is flat — open every campaign, every layer, every day, and your tool bill never moves with your ad spend.

See full pricing
FAQ

Questions you probably have.

Does the Deep Dive change anything in my account?

No. Sentrum connects with read-only permissions — read-only by architecture, not read-only until a write feature ships. It never touches campaigns, budgets, bids, or creative. There is a standing never-touch-budgets rule in the product itself. You read the layers; you make the calls.

Does it work for lead gen, SaaS, or app installs — or just e-commerce?

Any conversion goal. The funnel is built dynamically from your campaign’s own optimization event: a sales campaign gets impressions → clicks → landing-page views → add-to-cart → purchase; a lead-gen campaign gets a lead funnel with CPL as the verdict metric; video campaigns get the view funnel. Results always mean what your objective optimizes for — ROAS is never slapped onto a lead objective.

What does “significance-gated” mean in practice?

Every claim has to clear a threshold before it gets stated: roughly ten percent or more against the comparison period for spend, KPI, and efficiency moves, and a real share of spend or volume before anything is called a winner or a loser. A $19 CPA on three percent of spend gets labeled promising, not proven. If the data is too thin to support a verdict, the Deep Dive says so instead of inventing one.

What layers does it actually read?

Ad sets, creatives, demographics, placements, the optimization log, and the conversion funnel — each with the same three answers: what happened, quantified against the prior period; why, with the driver named; and what to do next, tied to that cause. The optimization log reconciles every budget change, new creative, and pause against the KPI line, so you can see which edit actually worked.

How is this different from the Daily Pulse?

The Pulse is the morning verdict and the real-time flag — it tells you something moved and why. The Deep Dive is the room you walk into when a campaign needs the full read: every layer opened, mid-flight, with what happened, why, and what’s next at each one. The Pulse links straight into it.

Can I put the output in front of a client?

The findings are written in client-ready language — quantified, sourced, no jargon detours — so the answer you read is the answer you give. For deck-grade deliverables, the Account Audit and QBR export as presentation-ready PPTX; the Deep Dive is built for the mid-flight answer, the same day the question lands.

What does it cost?

Flat $19, $39, or $79 per month depending on how many ad accounts you run — with a 7-day trial. Every plan includes the full Deep Dive on every campaign, every day. No credit meters on analysis, no spend caps, no per-$100K surcharges.

OPEN THE CAMPAIGN

Somewhere in your account, a layer moved this week. Go find out which one.

Connect read-only and open any campaign to the full layer read — what happened, why, and what’s next, with the evidence shown. If nothing moved past the gates, that’s an answer too.

Open my first deep diveSee full pricing
READ-ONLY · 7-DAY TRIAL · FLAT $19/$39/$79