Use Cases · Business Reviews
Every weekly, every month-end, every QBR, an analyst spends two days rebuilding the same pack — the same charts, the same cuts, the same numbers as last cycle. Then in the room someone asks “why did that move?” and there's no answer until next time, when it becomes another slide. Spotonix is the AI analyst that interprets before it queries: Spotonix answers the follow-up by making its interpretation visible before SQL executes — and the full Reviews workflow (pack assembly, watched signals, and decision/outcome memory) runs today.
The alternative
The review returns on a schedule, but nothing about last cycle's prep — the definitions, the decision, what actually happened after — carries forward. So you either rebuild the pack by hand or point at a dashboard no one trusts once the questions start.
Rebuild the pack
The analyst pulls the same numbers, re-cuts the same Segments, and re-pastes the same charts into the deck — mostly reproducing last cycle's work with fresh data. The prep is the job; the thinking gets whatever time is left before the meeting.
Point at the dashboard
Someone built a standing dashboard so the review could skip prep. But when a number looks off, no one can say whether the business moved or a definition drifted — so the tiles get second-guessed and the review reverts to the hand-built pack anyway.
The follow-up dies
The one question worth the meeting — why the number moved — needs a fresh cut no one has ready. It becomes an action item, lands as a slide next cycle, and by then the moment to decide has passed. The review reports the past instead of changing the next move.
Cost of delay Every review, an analyst pays two days of prep to rebuild a pack — and the decision that mattered gets deferred a full cycle because the follow-up couldn't be answered while the people who owned it were still in the room.
The outcome
Today, use Spotonix Ask to interpret a follow-up, inspect the plan, and run the analysis. Scheduled reports and alerts are also live. Automated pack assembly, watched briefs, and decision memory remain future capabilities.
01
Schedule recurring reports and alerts for agreed review inputs. Slide assembly and a complete review-pack workflow are not currently claimed as live.
02
When someone asks “why did that move?”, Spotonix exposes the Segments, Calculations, scope, and filters selected for the follow-up before SQL executes.
03
Reviews connects the decision made in the room with the outcome observed later, so each recurring review starts from what the last one settled.
How it works
The sequence below is how a recurring review works in Spotonix today. Availability labels are part of the content, not footnotes.
Ask Live
Ask the review question in plain language and inspect the visible plan. In Copilot mode, accept that plan before query generation continues; material query bindings are checked before execution.
Insights Live
Watched signals and automatically generated briefs are part of the Insights capability, available today.
Reviews Live
Pack assembly organizes governed analytical inputs for a recurring meeting, so the pack rebuilds itself instead of an analyst rebuilding it.
Reviews Live
Decision and outcome memory connects actions to later results, so the next review starts from what the last one settled.
Where this is today Reviews is a capability group inside Spotonix Analyst, not a separate shipped product.
Representative questions
The recurring number is on the slide. These are the follow-ups asked in the room — each resolves against an accepted plan in the Context Graph, so it can be answered without waiting for next cycle's pack.
Weekly review
“Bookings are down against plan — is that one region, or across the board?”
Segment Bookings by region Calculation Actual vs. plan, WTD Pattern Drill from the accepted pack
Month-end review
“Net revenue retention slipped — which cohort moved, and by how much?”
Segment Retention by cohort Calculation Net revenue retention, MoM Pattern Re-cut on same definitions
Quarterly business review
“Last quarter we decided to lean into enterprise — did that show up in the numbers?”
Segment Enterprise cohort Calculation Growth vs. prior quarter Pattern Review follow-up with decision memory
Proof
The behavior you can watch today: ask the follow-up, inspect the visible interpretation, accept the plan in Copilot mode, and inspect the material bindings used by the generated query. Worked traces use the public TPC-DS schema and are illustrative.