An indicator you can drill into is an argument. One you cannot is a slide.
The number on the dashboard is the end of a chain. If it cannot be walked back to the records that produced it, everyone in the room is discussing a claim rather than a measurement — and the discussion goes to whoever is most senior.
Measured where the work happens.
Indicators are computed over the operational records themselves, not over a nightly extract into a separate warehouse. That removes the gap where a dashboard and the operation disagree and nobody can say which one is right, and it means an aggregate opens down to the individual cases underneath it.
Because every certified process runs on the same engine, an indicator can combine incidents, requests, changes, problems and service-level outcomes without a mapping layer between them — which is what makes a cross-process measure a configuration rather than a reconciliation project.
What it does
From a strategic objective down to the case
The measurement layer, and the structure that makes a measurement mean something.
Strategy first, then indicators →
KPIs sit under governed strategic objectives and can be related through alignment maps to goals, critical success factors and intended outcomes. Targets are set by period and deployed to accountable people or teams, so the scorecard holds the target, the owner, the period, the history and the strategic context together.
Reports, in three layers
Reports that ship with the product, reports your administrators build with the wizard including your own fields, and real-time dashboards. Worth keeping distinct when evaluating any tool: the demonstration usually shows the first, and the question is about the second. The time each case spent in each status is a report column, and reports move between installations as protected packages. The product's own reports ship organised by area — service desk, assets, platform, processes, security, surveys — read-only, so an upgrade never overwrites your version: you save your own copy. And a report's fixed conditions are enforced by the server, so whoever consults it cannot widen what the designer narrowed.
Drill-down
From an aggregate result to the governed records that produced it, subject to the same permissions as the records themselves.
Scorecards and boards →
Balanced scorecard views, and a board where milestones are cards grouped by execution status and organisational swimlane, each one opening into the underlying record.
Scheduled distribution
Reports generated and distributed to authorised audiences on defined cycles, with the underlying agreement, target and period retained so historical results stay comparable. A scheduled report can also create one activity per row it finds.
Evidence on every data point →
Every data point of an indicator takes documents and comments with their author and date — the audit trail of why a value is what it is. The improvement plan is created from that point, as an activity tied to the measurement that prompted it, and indicators captured by hand create the capture task for each owner when the period opens.
Alignment boards and KRIs →
Each indicator is assigned to who answers for it, per fiscal year, with its own target and weight and an alignment role — management, success or influence. The same element linked to a risk is a key risk indicator: the KRI is a relationship, not a checkbox.
Critical success factors →
Each fiscal year the company defines its critical success factors, each one measured by the weighted indicators linked to it, on a board that shows its compliance and the indicators that explain it.
Forecast, with a confidence level →
Know before the close whether an indicator will miss its target and how much demand each service will receive, with a confidence level: every indicator carries its estimated year-end close with a 95% prediction interval, and the dashboard flags it amber when at risk and red when a miss is likely — and indicators trigger an action when they cross their threshold.
What explains the change →
When a composite indicator moves, its record splits the change exactly across the operands of its formula, non-linear ones included, with the largest driver first — so the conversation starts from the cause, not from the total.
Measurements from outside →
Weather, traffic, epidemiological data, sensors or figures published by an outside body, loaded by integration and used as operands of the indicator formula, crossed with internal indicators.
Georeferenced views
Results presented on a map with drill-down, where location is a dimension that matters — a network, a set of branches, a territory.
Surveys
Client-defined surveys triggered from a record, at closure or on any condition, with each response linked to the case that produced it and aggregated into satisfaction indicators.
Formulas →
Return on investment, total cost of ownership, penalties, credits and performance-cost rules computed over agreement, cost, effort and outcome data, and reported by period. Every measure declares whether it is a variable that feeds formulas or an indicator with target, direction and owner; the sheet blocks contradictory combinations and the conformance verifier detects circular formulas.
Export →
Data reaches external BI tools through the API and the integration engine, for organisations that already standardised on one. Personal dashboards download to Excel or Word, with each indicator’s history.
Under the hood
How the projection is built, and why each decision is there
Projection is the feature most easily faked and most rarely inspected. Four decisions determine whether the line means anything.
| The decision | Why |
|---|---|
| The trend is regressed over percentage of target achieved, not over raw values | When the target moves between periods, raw values trend on the target rather than on performance |
| Each projected point is rebuilt against the real target for its own period | A forecast against last period’s target answers the wrong question |
| The direction of the indicator is respected | For some measures lower is better, and a projection that ignores this reports improvement as decline |
| Projections are drawn only for periods with no measurement | A measured value is never overwritten by an estimate |
Precision
The distinctions this module preserves
Four confusions that turn a measurement layer into a decoration.
On the roadmap
What arrives next, and when
Each item is available for a demonstration or a proof of concept from the month shown.
Natural-language report assistant
Users describe in their own words what they want to see, and the assistant builds the report with its charts, ready to review and save.
Root cause and recommended action
Explain why an indicator moved, which cases explain it and which action is advisable.
The proof
Measured against a target someone signed
An indicator matters when a target attached to it was agreed by a named owner for a stated period, and the outcome is compared to it afterwards. That is the last link of the chain — strategy, capabilities, services, offerings, processes, data and indicators — and the reason the chain is worth building at all.
See the traceability chain →The other capabilities of Dexon BPM
Workflow
Conditional routing, parallel tasks, dynamic approvers, deadlines on business calendars and append-only history — the same engine across every record type.
ETL & integration
REST API, scheduled and event-driven exchange with databases, files and web services, plus directory, email and discovery — with the authoritative source of every attribute on record.
Documents
Version trees, permissions, search by content, signature with time stamp — with document requirements enforced as gates the workflow will not pass without.
IT asset management
Inventory through the Dexon Agent, a CMDB that maintains itself, SNMP network discovery, licensing measured on real usage, and remote control recorded on the case.
Bring an indicator nobody trusts
The one that gets argued about every month. We will show you what it would take to make it drillable — and where the number is actually coming from today.