PR Performance · Data2026

Press Release Performance Dashboard

A static dashboard that cumulatively tracks and debriefs the "coverage status" reports generated after each press release goes out. Designed to measure quality, not just volume, through a 'PR Impact Score' that weights individual outlet scores and print placement.

Challenge

Coverage status reports for press releases were manually compiled and piled up as individual files. Each release's performance was known only by gut feel, making time-series or campaign comparison, and KPI management, impossible. A simple total-count tally showed only 'volume' — it missed 'quality' signals like major-outlet placement, business relevance, prominence, and print coverage, along with tone and how well key messages actually came through.

Approach

Extracted outlet counts (previously compiled as PDFs) and AI-analyzed the body text of clipped articles (tone, byline, key-message penetration) in-session, loading everything into a single JSON source (releases.json)

Built an individual outlet-score table (98 outlets seeded) and designed a 'PR Impact Score' formula that weights online, broadcast, and print scores

Covered the pure calculation functions (calc.js) with node:test so new releases can be added without regressions

Built and deployed a dashboard with KPI summary / time series / rankings / category ROI / tone & messaging / byline blocks, plus a single-release debrief card (card.html); results are compiled and compared the day after each release goes out

Deliberately scoped out anything the data couldn't support (e.g., tone analysis across the full online volume) — surfacing only metrics that hold up, without inflating them

Outcome

✓

Converted individually distributed press releases into a single dataset comparable across time and category

✓

Shifted the dashboard's core metric from 'count' to an 'Impact Score' reflecting major-outlet and print placement quality

✓

New releases now flow into the pipeline automatically with just one added monitoring PDF and an AI session pass — no manual re-work