How SEO, AI search, conversion and content reshaped acquisition in the first half of 2026, measured against the same period in 2025.
Total new contacts held roughly flat year on year (about 16,900 both halves). What changed is where they came from. Paid was cut hard and organic plus earned channels grew to fill the gap. The acquisition base is now cheaper and compounds over time.
In the first half of 2025, organic traffic declined month on month. In 2026 it climbed every month, and the year-on-year gap widened as the period went on. Three independent sources agree on the direction.
| Month | 2025 | 2026 | Change |
|---|---|---|---|
| January | 19,804 | 20,493 | +3.5% |
| February | 19,339 | 21,466 | +11.0% |
| March | 19,433 | 21,841 | +12.4% |
| April | 19,217 | 22,155 | +15.3% |
| May | 18,233 | 22,171 | +21.6% |
Authority improved over the same window: SEMrush domain rank moved from about 26,475 to 23,391 (lower is better).
| Metric | 1 Jan to 15 Jun 2026 |
|---|---|
| Clicks | 46,300 |
| Impressions | 3.88m |
| Average CTR | 1.2% |
| Average position | 22.8 |
| Branded share of clicks | ~54% |
Property: www.joblogic.com. A clean year-on-year view is not possible in Search Console because data retention is limited to 16 months, so January 2025 is out of range. SEMrush and HubSpot carry the year-on-year comparison instead.
AI search visibility is a 2026 capability, so there is no fair 2025 comparison. AirOps tracking began on 9 February 2026. What we can show is the trajectory since, plus the first traffic and the first contacts arriving from AI assistants.
Generic category prompts, measured against eight tracked competitors across 345 prompts. June is a partial month to the 17th, which lowers the bar.
ChatGPT key events rose from 2 to 17 over the same window. Gemini and Copilot are present but small, and Gemini can misattribute to gstatic.com, so they are excluded from this view.
Conversions grew modestly in absolute terms. Sitewide conversion rate is not a reliable headline this year for two specific reasons, both noted below, so this section leads on absolute key events and HubSpot contacts.
| Segment | 2025 | 2026 | Change |
|---|---|---|---|
| All key events | 18,545 | 19,410 | +4.7% |
| Organic search key events | 6,663 | 6,970 | +4.6% |
| Engagement rate | 84.7% | 87.9% | +3.8% |
| Avg engagement time / session | 4m 46s | 4m 59s | +4.6% |
Direct traffic inflation. Direct is 83% of GA4 sessions (5.87m of 7.05m), well above any believable human level for a UK B2B site. That bot and dark traffic sits in the denominator and makes any sitewide rate meaningless.
Demo page tracking gap. The new demo booking pages record close to zero conversions in GA4 while HubSpot sees roughly 15. Until that gap is fixed, a sitewide rate understates real performance.
Whether you look at the first touch or the last, the pattern is the same. Organic and earned channels gained; paid search fell by the same 38% on both models. HubSpot, contacts created in the period. The chart shows the marketing-influenced channels only.
| Group | 2025 | 2026 | Change |
|---|---|---|---|
| Earned (organic, social, referral, AI) | 495 | 824 | +66% |
| Paid (search and social) | 2,012 | 1,266 | -37% |
Offline and Direct are excluded from this view. Offline is inflated by the historic Clik and Protean CSV imports, and Direct last-touch is a known measurement artefact, so neither is a reliable performance signal. Paid search is the cleanest attribution signal in the portal.
Every figure here is traceable to a named source and an explicit date window. The caveats are stated up front so nothing unravels in the room.