What the Audience
Tells Us
01 · Engagement Over Time
The audience pulse
Open rates and click rates across all 66 editions. The spikes tell a story — certain weeks electrified the audience while others barely registered.
Open rate & click rate by edition
What the engagement data reveals
- Issue 29 (Apr 14, 2025) achieved the highest open rate at 39.7% — a week of exceptional AI news that clearly resonated with leaders.
- Issue 15 (Jan 6, 2025) hit 37.4% opens with a massive 21.8% click rate — the post-holiday return combined with major AI developments drove extraordinary engagement.
- Summer 2025 saw a sustained dip — open rates dropped to 12–14% from June through September, the classic newsletter seasonality pattern.
- The autumn recovery was strong — Issues 57 (Dec 1) and 60 (Jan 5, 2026) both exceeded 33% opens, demonstrating renewed audience interest.
- Click rates are the real signal — while opens fluctuate with seasonality, click-through spikes at 15–22% indicate genuinely compelling content that drove action.
02 · Standout Editions
What the audience loved — and ignored
The 10 highest and lowest performing editions by open rate, revealing which weeks cut through and which got lost.
Patterns in the top performers
- Post-break editions dominate — Issues 14, 15, 60 (post-Christmas) and 29 (post-Easter) consistently topped the charts, suggesting pent-up demand after gaps.
- High clicks ≠ high opens — Issue 14 had 34.2% opens but only 1.9% clicks (holiday curiosity, no depth). Issue 15 had 37.4% opens AND 21.8% clicks (genuine engagement).
- Summer is the graveyard — 7 of the bottom 10 are from June–September. The audience disengages seasonally regardless of content quality.
- Issue 20 (Feb 10, 2025) anomaly — just 10.1% opens, the all-time low outside summer. Possible delivery issue or subject line misfire worth investigating.
03 · The Four Phases
A newsletter in four acts
The data naturally divides into four distinct engagement phases, each mapping onto what was happening in the AI landscape and the newsletter’s content mix.
Phase 1: Sep–Dec 2024
The Honeymoon
Avg opens: 21.4%
Avg clicks: 4.4%
Issues: 1–14
New audience discovery. Volatile engagement as readers found their rhythm. Issues 2, 5, and 9 spiked massively on clicks — early content experiments that worked.
Phase 2: Jan–Apr 2025
Peak Engagement
Avg opens: 27.2%
Avg clicks: 8.8%
Issues: 15–30
The golden period. DeepSeek disruption, Paris AI Summit, UK AI Strategy — major events drove consistently high engagement. Contains 6 of the top 10 editions.
Phase 3: May–Sep 2025
The Summer Drift
Avg opens: 14.7%
Avg clicks: 1.6%
Issues: 31–46
Seasonal disengagement hit hard. Open rates halved from peak. Content quality held steady per the content analysis — the audience simply checked out for summer.
Phase 4: Oct 25–Feb 26
The Rebuild
Avg opens: 22.6%
Avg clicks: 4.1%
Issues: 47–66
Autumn recovery with new ad partnerships (Scrum.org, AIWeek, Zoocha, Sopra Steria). Open rates recover but click rates remain below Phase 2 peaks.
Open rate by edition — coloured by phase
04 · Audience Evolution
Growing, shrinking, or shifting?
Tracking total audience size, unsubscribes, and bounces to understand audience health over 17 months.
Audience health assessment
- Gradual attrition — the core audience declined from ~73,500 (Issue 1) to ~64,600 (Issue 66), a net loss of ~8,900 or roughly 12% over 17 months.
- Issue 40 anomaly — sent to 128,623, nearly double the usual. Labelled “AI Pulse Light” — an expanded-list experiment that produced the same 13.8% open rate.
- Low unsubscribe rates throughout — averaging 0.13%, well below the B2B benchmark of 0.2–0.5%. The audience that stays, stays.
- Partnership-driven growth — Issues 49–50 (Scrum.org) saw 746 and 328 new subscribers. This is the proven model for offsetting natural attrition.
- Most attrition is passive — email decay (job changes, stale addresses) rather than active unsubscription. The low unsub rate confirms this.
05 · Opens vs Clicks
Reach versus resonance
Each dot is an edition, coloured by phase. High opens + low clicks = curiosity without substance. The top-right quadrant is the sweet spot.
Open rate vs click rate — each dot is one edition
The four quadrants
- Top right: reach + depth — Issues 2, 5, 9, 15, 16, 29, 57, 60, 62. Both high opens and high clicks. These should be studied for what made them work.
- Bottom right: curiosity gap — Issues 14, 17, 18, 30. Strong subject lines or timing drove opens, but content didn’t compel clicks.
- Bottom left: summer cluster — Phase 3 editions cluster here tightly. Seasonality overwhelmed content quality — the content analysis confirmed no decline in curation during this period.
- Phase 2 (green) dominates the upper half — Jan–Apr 2025 was when the AI landscape and audience engagement were both at peak intensity.
06 · Metric Shifts
How the numbers moved phase to phase
Comparing key metrics between the peak period (Phase 2) and the current period (Phase 4) — what recovered, what didn’t.
07 · Monetisation Signal
The advertising story
From Issue 48, advertising click data and named sponsors appear. This tracks the commercial evolution of AI Pulse.
Advertising clicks by edition (from Issue 48)
Commercial observations
- Issue 57 (Dec 1, AIExpert + DL LinkedIn) generated 9,694 ad clicks — extraordinary, driven by the 33.9% open rate and strategic ad placement.
- Issue 60 (Jan 5, 2026) — 10,652 ad clicks on 35.2% opens. Post-Christmas return driving peak commercial performance too.
- Partner diversity is growing — Scrum.org, AIWeek, Zoocha, AIExpert, DPA, CIOMirror, Sopra Steria. Eight partners in 19 editions.
- Ad clicks correlate with open rates — but not perfectly. Issue 62 (DPA, 32.8% opens) drove 8,872 ad clicks while Issue 55 (Zoocha, 19.9% opens) drove only 81.
08 · Content × Engagement Synthesis
When content meets audience
Drawing on the content analysis of 72 editions and 2,026 curated items alongside this engagement data — the strategic picture.
What the correlation reveals
- Major AI events drive massive engagement — the highest-performing editions (15, 16, 29, 57, 60) coincide with landmark AI news weeks (DeepSeek launch, Paris Summit, UK strategy, year-start roundups). The content analysis showed higher item counts in these weeks too.
- Innovation and Policy content drives clicks — editions heavy on Productivity & Efficiency (the dominant category) maintain steady opens, but the click spikes come when Innovation & Collaboration or major Regulation stories dominate.
- The volume–engagement paradox — the content analysis showed editions growing from ~22 to 30–40 items, but click rates didn’t grow proportionally. More items may mean more diluted attention.
- Post-break demand is the strongest signal — Christmas, Easter, and bank holiday gaps consistently produce the highest-opening subsequent editions. The audience accumulates demand during breaks.
- Summer quality doesn’t matter — even well-curated summer editions (content analysis confirmed no quality decline) couldn’t overcome seasonal disengagement. This is structural, not editorial.
- Sustainability coverage remains under-engaged — the content analysis showed it stayed flat throughout, and click data suggests it’s not what drives action. The importance-engagement gap here is notable.
Strategic recommendations
- Lean into the “return from break” effect — consider a “What you missed” format for post-holiday editions to capitalise on proven pent-up demand.
- Accept the summer trough — rather than fighting it, consider reduced frequency or a lighter format. Preserve audience goodwill rather than training people to ignore you.
- Click rate as the quality metric — open rates are increasingly unreliable (Apple Mail Privacy, preloading). Click rate better measures genuine engagement. Target 5%+ consistently.
- Fewer, stronger items may outperform volume — the paradox between growing item counts and flat click rates suggests curation intensity matters more than breadth.
- The advertiser value proposition is proven — 65k+ audience with 20%+ average opens, named engaged senior leaders, and ad click rates that spike to 10,000+ in peak weeks.
- List growth via partnerships works — the Scrum.org subscriber influx shows the model. More co-branded content partnerships could offset the natural ~7% annual attrition.
For “Making AI Work for Britain”
- This is primary research — 65,000 UK digital leaders’ engagement with curated AI news over 17 months is a dataset few organisations possess. The patterns reveal what practitioners actually care about versus what the media covers.
- The engagement peaks map to policy moments — DeepSeek, Paris Summit, UK AI strategy announcements drove the biggest audience responses. When policy is clear and consequential, leaders pay attention.
- The Innovation > Productivity click pattern tells a story — leaders read about productivity gains (duty) but click on innovation breakthroughs (interest). The book should speak to both motivations.