Your full account, mirrored.
No watered-down data model. Sentrum preserves your real hierarchy — campaigns, ad sets, ads, creatives, placements, audiences — then lets you layer the labels, tags, and breakdowns that match how your team actually thinks about the account.
Five levels, not three.
Most tools collapse Meta into campaign + ad-set + ad. We keep the full tree so creative-level diagnostics, placement breakdowns, and audience overlap analysis all stay queryable in the same view.
- Campaign → objective + budget + structure (ASC vs Manual)
- Ad set → audience + placement + optimization goal + bid strategy
- Ad → creative attached + tracking parameters + AEM event
- Creative → asset metadata + Gemini video analysis
- Placement / breakdown → age, gender, hour, region, ASC user_segment_key
Your account taxonomy, not a generic one.
Tag campaigns by funnel stage, brand pillar, creator, promo. Labels surface in every dashboard filter and roll up into report aggregates.
Configure which breakdowns load by default per surface — age + gender on Daily Pulse, placement on Campaign Deep Dive, ASC user_segment_key on retargeting accounts.
Bundle accounts (e.g. one brand's US + EU + APAC accounts) so reports and AI insights can run across them as a single unit.
Every edit, captured automatically.
Sentrum records every budget change, audience tweak, creative add or pause, bid-strategy switch, and AEM re-prioritization. Inflection points on every metric chart are tagged with the matching log entry — so cause and effect are always one hover apart.
| Edit type | Surfaced as |
|---|---|
| Budget change | Inflection marker + delta % |
| Creative add / pause | Creative-tab change row + AI follow-up prompt |
| Audience edit | Ad-set timeline marker + learning-reset flag |
| Bid / goal change | Performance baseline reset notice |
| AEM re-priority | Account-level alert + recommendation |
40+ signals, scored against your baselines.
Sentrum builds rolling per-account baselines for every metric. Anomalies are scored relative to your normal — not a generic industry benchmark — so a CPM swing that's wild for your account but boring for someone else still gets flagged.
- Spend velocity
- Pacing vs target
- Day-of-week mix
- CPA / ROAS
- CPM / Reach
- Cost per result
- CTR / VTR
- CVR / LPV rate
- Frequency saturation
- Audience overlap
- Creative archetype mix
- Placement drift
Structure is the prep work. Analysis is the payoff.
See how the 4-pass AI engine turns the connected, structured data into specific recommendations on the Analyze Data page.