Comms Leader

Measurement

Part of Building a corporate communications report with its evidence states kept apart

Designing a corporate communications dashboard with four panes and a data-health rail

Build a blank, accessible communication dashboard with versioned definitions, data-quality warnings, correction controls and decision ownership.

A corporate communications reporting dashboard should expose uncertainty and action, not hide them behind coloured tiles. This blank design is for Cedar Vale Engineering Ltd's fictional England site decision. It contains no counts, thresholds or sample results.

Write the header first

Place the decision question above every chart. Add the approved communication version, reporting period, timezone, data cut-off, dashboard version, owner and next review. Show whether the pack is draft, corrected or superseded. A reader should not need a hover state to discover that evidence is late.

The Government Analysis Function's February 2026 guidance on testing dashboards for design and accessibility recommends accessibility testing and supporting data tables. It is analytical guidance, not a private-company compliance certificate. Cedar Vale needs named accessibility and equality review for the real audience and format.

Use four panes, not one blended score

The control pane contains approval, source-copy, rights, accessibility and release-gate states. The delivery pane contains precisely defined publication, dispatch, endpoint and correction events. The audience pane is empty until an authorised research method supplies evidence. The decision pane records operational outcomes, attribution status and unresolved causal questions.

Each row in any pane needs these blank fields:

measure ID | purpose | event | eligible population | numerator | denominator | unit | period | timezone | exclusions | source/query version | missingness | owner | uncertainty | response

Do not use one traffic-light colour to offset hard failures. An unapproved claim remains a stop even if a delivery measure looks favourable.

Add a data-health rail

Beside the main view, list source freshness, failed imports, unmatched records, duplicate handling, definition changes and open corrections. Preserve supplier dashboard labels as supplier definitions until the buyer reconciles them to source events. NCSC logging guidance emphasises choosing logs for the questions they must answer and configuring needed fields and time correctly. That security guidance supports auditability, not communication performance.

Every transformation should have a query or workbook version, approver and reproducible input reference. Late-arriving records enter a later version with a correction note; they do not silently rewrite a signed report.

Make the dashboard answerable

Add an action register below the visuals. Each row needs the issue, evidence ID, accountable person, decision due date, permitted response and closure proof. A chart may flag that a source is incomplete, but only the register shows who must investigate it and what closes the exception. Keep commentary attached to the relevant dashboard version so an explanation cannot survive after the underlying definition changes.

Give reviewers a plain export containing the displayed rows, definitions and revision history. They should be able to reproduce the view without relying on a presenter's memory. If the main screen shows an estimate, its publisher, method, geography and retrieval date belong beside it, not in a remote footnote.

Limit what each reader sees

An executive view can show the decision and exceptions without personal-level records. Analysts may need controlled access to reconcile underlying events. Communications staff do not automatically need employee, customer or journalist identifiers. The ICO's data-protection-by-design guidance was updated on 5 February 2026 and calls for privacy to be considered throughout a processing lifecycle. Qualified reviewers must establish Cedar Vale's purposes, roles, access and retention.

Test with a blank and a failure

Before production is enabled, test keyboard access, reading order, text alternatives, table download and zoom with synthetic records. Then introduce a missing denominator, late source and revised status definition. The dashboard passes only if it displays each problem, blocks over-interpretation and points to an owner.

Finish with a change log stating what moved, why, who approved it and which earlier reports are affected. If the accessible table, definitions or correction history cannot be recovered independently of the visual, the build stays on hold.

More in Measurement

Measurement

Building a corporate communications report with its evidence states kept apart

Build an auditable corporate communication report for an England decision, separating delivery evidence, outcomes, attribution and causal inference.

Measurement

Four attribution approaches for corporate communications, compared on one decision

Compare four communication attribution approaches on one England decision, using consistent questions, evidence fields and explicit causal limits.

Measurement

Benchmark research for corporate communications when no industry norm exists

Research comparable communication evidence for an England decision without inventing an industry norm, threshold, result or causal conclusion.

Measurement

Choosing corporate communications measures with complete definitions and blank thresholds

Choose defensible communication measures for an England decision, with complete definitions, blank thresholds and clear evidence limitations.