Using a Hub
Read a hub end to end — the narrative, the KPI strip, the charts, and the Decisions & Actions panel — then drill into the detail. About 8 minutes.
A hub is a focused analytics workspace for one commercial or operational area. Open any hub from the sidebar — for example Price & Promo, Distribution, OSA, or Category Share. Every hub follows the same layout, so once you can read one, you can read them all. This page is about how to read and use a hub day to day — not how it’s built.
The fourteen hubs
There are fourteen hubs, organised into three groups in the sidebar. Each one answers a single, focused business question. The questions below are generic and illustrative — your tenant’s wording may differ slightly.
Commercial
| Hub | The question it helps you answer |
|---|---|
| Price & Promo | Are we priced right vs the category, and are promotions complied with and profitable? |
| Category Share | Are we winning or losing share, and against whom, where, and in which segments? |
| Margin / Value Chain | Where is margin leaking across the value chain, and which levers recover it? |
| Innovation | Are new launches landing — distribution, rate of sale, and cannibalisation? |
Supply & Operations
| Hub | The question it helps you answer |
|---|---|
| S&OP | Is demand and supply balanced, and where are the gaps in the plan? |
| Predictive Ordering | Are we ordering the right quantities at the right time to avoid stockouts and overstock? |
| Distribution | Are the right SKUs listed and present in the right stores? |
| OSA (On-Shelf Availability) | Is product actually on shelf where shoppers can buy it? |
| Waste | Where are we losing product to spoilage, shrink, or markdowns? |
| Route Optimisation | Are field and delivery routes efficient and well-prioritised? |
| Staff Allocation | Are people deployed where the workload and decision actually are? |
Intelligence (native screens, not standard hubs)
The Intelligence group looks like more hubs, but its three items are native screens rather than the standard hub layout — they link straight into the working surfaces of the app:
| Item | Where it takes you |
|---|---|
| Risks & Decisions | /app/decisions — every open card, newest first |
| Image Recognition | /app/field — field execution and shelf imagery |
| AI Decision Intelligence | /app/chat — KeyChat, the conversational assistant |
So when this page says “every hub,” it means the eleven Commercial and Supply & Operations hubs above, which share the structure described next.
How a hub is laid out
Every hub follows the same top-to-bottom structure.
1. The narrative header
At the top, an icon and title introduce the hub, alongside a group tag (Commercial or Supply & Operations) and a short description of the question it answers — for example, “Are we priced right versus the category, and are promotions complied with and profitable?” This frames everything below it, so read it first.
2. The KPI strip
A row of four headline metric cards, each with a coloured accent strip and a live figure. These are your at-a-glance health check for the area — distribution percentage, on-shelf availability, rate of sale, revenue at risk, and so on, depending on the hub. The accent colour signals status at a glance; the number tells you the size of it.
3. The charts
Below the KPIs sits the visual analysis: a full-width primary chart that carries the main story, with supporting charts in a 2-column grid beneath it. This is where you spot trends, compare segments, and trace where a headline number is coming from.
DiagramA hub's KPI strip, primary chart, and supporting charts
An honest note on the charts. A hub’s KPIs and charts are rendered natively by KeyOne against your tenant’s data. Because those visuals are generated live, we can’t show you a real screenshot of them here — so the picture above is an illustration of the layout, not a screen capture. When a metric has a mapped source, the real charts render in these positions. If a metric has no source yet, the hub will tell you so directly (see “The honest states of a hub” below).
4. Root-cause analysis
Most hubs include a root-cause section that breaks a headline number down by dimension — region, retailer, category, SKU — and compares targets against actuals, so you can see why a metric moved, not just that it did. Treat this as the bridge between “something’s off” and “here’s the specific thing to fix.”
5. The Decisions & Actions panel
This is what makes a hub more than a dashboard. At the bottom, every hub carries a Decisions & Actions panel that ties the analytics back to your work:
- A count strip of Open Risks, Decisions, and Open Tasks for this area
- The top ~5 work items relevant to the hub
- A “View all in Work Queue →” link to act on the full list
So you can go from spotting a problem in a chart to opening the exact decision that addresses it, without losing your place. Each item flows into the same process described in Working a Decision.
How to drill in
To move from the big picture to the specific action, work top to bottom:
- Read the narrative to anchor what you’re looking at and which question this hub answers.
- Scan the KPI strip for anything off-trend — a red accent, a number well below target.
- Study the primary chart, then the supporting charts, to localise the issue: which region, which retailer, which SKU is dragging the number.
- Check root-cause to confirm the driver rather than guessing.
- Jump to Decisions & Actions and open the matching work item. From there it’s the same flow as Working a Decision.
Tip: The hub already did the triage. If a KPI is healthy, you usually don’t need to study its charts — spend your attention on the cards with a warning accent.
A worked example: reading the OSA hub
The figures below are illustrative — invented to show the workflow, not real customer data, and no real retailers are named.
You open the OSA hub to check on-shelf availability.
- Narrative: “Is product actually on shelf where shoppers can buy it?”
- KPI strip: On-Shelf Availability reads 91.4% against a 96% target — its accent strip is amber. Rate of Sale and Distribution look healthy; Revenue at Risk shows ~£42k/week.
- Primary chart: A trend line shows OSA steady nationally except for a dip starting nine days ago.
- Supporting charts: A region breakdown localises the dip to a single region — call it Region North — sitting at 83% while every other region is above 95%. A store-level chart narrows it further to roughly 18 stores within that region.
- Root-cause: The breakdown attributes most of the gap to two SKUs in one category — a classic phantom-stock / replenishment pattern rather than a true out-of-stock.
- Decisions & Actions: The count strip shows 3 Open Risks for OSA, and the top items include the matching out-of-stock risks for those Region North stores. You open the most urgent one and act on it.
In under a minute you went from a single amber KPI to the specific stores, SKUs, and decisions behind it. That’s the point of a hub.
Tip: Note the gap to target, not just the headline number. A “91%” can be excellent or alarming depending on the target the card is measured against.
The honest states of a hub
A hub’s KPIs and charts are rendered live and natively by KeyOne’s own analytics engine. KeyOne is deliberate about never dressing up missing data as live data, so you’ll only ever see one of three states.
Live
The engine is provisioned and reachable: KPI cards and charts render against your tenant’s real data, and everything described above is fully interactive.
Not provisioned
If the analytics engine isn’t configured in your environment, the hub says so plainly instead of showing empty or fake charts — for example:
“This hub’s analytics are not provisioned. Configure the analytics engine to enable charts and KPIs. The Decisions & Actions panel below is available regardless.”
Crucially, the Decisions & Actions panel keeps working in this state — so you can still triage and resolve the decisions for that area even before charts are switched on.
Degraded
If the engine is configured but temporarily unreachable, you’ll see a degraded notice (“sidecar unavailable”) rather than stale numbers. This is transient: analytics restore automatically when the engine reconnects — no action needed on your part. As with “not provisioned,” the Decisions & Actions panel remains usable throughout.
Tip: Seeing “not provisioned” or “sidecar unavailable” is expected behaviour, not a bug. It’s KeyOne telling you the truth about what it can and can’t show right now.
Common pitfalls
- Reading a number without its target. A KPI’s accent colour reflects performance against target — always read the two together.
- Stopping at the national figure. The primary chart shows the trend; the supporting charts are what localise the problem to a region, retailer, or SKU. Don’t skip them.
- Treating “not provisioned” as broken. It’s an honest state. The Decisions & Actions panel still works, so you can keep triaging.
- Mistaking the illustration for live data. The picture in this doc is a layout illustration, not a screenshot — your real charts appear once the engine is provisioned.
- Ignoring the Decisions & Actions panel. The analytics tell you what; the panel is where you actually do something about it. A hub read that ends without opening a work item is usually a read left half-finished.
Where to go next
- Working a Decision — act on a work item from a hub
- Decision Inbox — see decisions land across every area
- KeyChat — ask questions in plain language
- FAQ and Glossary — terms and common questions
Next: KeyChat