The “That’ll Do” Line
We audited 391 UK & US recruitment agencies of 50 people or fewer against 34 marketing signals read straight from their live websites. What we found was not a spread of strong and weak marketers. It was a wall — and it stands in the same place for almost everyone.
AGENCIES
Most agencies are missing the marketing basics, well before anything advanced comes into it.
A score of 27.9 out of 100 reads like a sector failing at sophisticated marketing. Mostly it stopped earlier, at “the website exists”. Only three of the 34 signals we measured clear 75% adoption: a live site (99%), a modern responsive build (81%) and an XML sitemap (76%). Every other signal, including free ones that take an afternoon, is a minority.
Nearly half of these agencies cannot say how many people visited their site last month, because only 54% run GA4. The most basic free analytics install there is. 56% have a meta description. Half have a genuine enquiry form rather than a job-search box. 44% have a blog, and 22% have published anything on it since 2025. These are the floor, and much of the sector sits below it.
The “that’ll do” line is real, and it sits lower than the tools conversation assumes. For most agencies the next move is finishing the foundation, analytics, a real enquiry form, a meta description, long before automation or answer-engine optimisation.
Almost the whole sector settles for the same thing.
The average agency scores 27.9 / 100.
99% sit at Tier 2 or below; just 3 of 391 reach Tier 3, and none reach Tier 4.
The plateau lands in the same place for almost everyone.
A website, analytics and a contact form, then the spending stops. 54% sit exactly there.
Adoption collapses the moment marketing needs intent.
Visitor identification 1%, marketing automation 7%, gated content 2%.
Being citable by AI answer engines is open ground.
AEO readiness sits at 7% overall and in low single digits on the signals that matter. Only one sector reaches double figures.
The tech recruiters come last.
Engineering/Construction (32.4) and Finance/Accounting (32.3) lead; IT/Tech (22.9) trails, and even the strongest sector averages a Tier 1 score.
A wall at Tier 1.
Where 391 agencies land on the 0–100 scale. Scroll to walk the tiers.
BROCHUREWARE
FOUNDATIONS
ACTIVE MARKETER
DEMAND-GEN
GROWTH ENGINE
Analytics, a contact form, basic SEO. Functional but passive. Waiting for inbound it never sees.
The scores cluster into a wall, which reads as a shared habit rather than a range of budgets.
If maturity tracked budget or headcount, the 391 scores would spread into a long tail. Instead they bunch: 82% inside a 40-point band, 54% inside a 20-point one. The tight clustering is the signature of consensus, an industry-wide agreement about where marketing stops.
It stops early. The consensus position is the handful of things a founder commissions once and never revisits: a site, a tracking snippet, a contact form. Climbing past Tier 1 needs someone to own marketing as a job, and in a sub-50-person firm billing on placements, no one holds it.
So the ceiling is organisational more than technical. The plateau is the height a recruitment business reaches when marketing is a one-off project, and it lands in the same place for almost everyone.
Half can watch traffic arrive.
Almost none find out who it was.
The widest gap in the dataset
Half of all agencies can watch traffic arrive. Almost none find out who it was.
The journey rarely reaches conversion
For most agencies a contact form catches whatever demand it can, if anything does.
Each stage is matched to its nearest directly-observable website signal. Measured from rendered site code, not self-reported.
The funnel leaks at the top as much as the bottom.
The obvious reading of the seven stages is well-funded presence and ignored conversion. The measurement layer is thinner than that: only 53% run GA4, so nearly half of these agencies cannot yet count a visit.
That reorders the headline gap. 53% can watch traffic and 1% can name the company behind it, a real drop. But naming a visitor you are not yet counting is impossible, so for most sites the first missing stage is analytics, then a genuine enquiry form at 50%.
Visitor identification and answer-engine work only pay off once a site already measures its traffic and captures enquiries. For much of this sector that groundwork is still missing, which is where the blindness actually begins.
What 391 agencies actually have
The foundations are widespread. Everything that turns a website into a measurable, demand-capturing channel is rare — automation, visitor identification and AEO all sit in single digits.
Analytics is the one habit that stuck.
Behind every maturity score is a stack of specific tools. Beyond measurement it thins out fast: the tools that turn anonymous traffic into named pipeline appear on just 4 sites in the whole sample of 391.
Most of these sites should finish the top of this stack before buying the bottom of it.
The stack thins steadily: analytics 53%, tag management 23%, paid pixels 17%, automation 7%, visitor identification 1%, four sites in the whole census. The gap is genuine, and the first mover in each niche takes it uncontested.
The sequencing advice is the boring one. A site with no meta description, no analytics and no genuine enquiry form has nothing for visitor-ID or answer-engine optimisation to build on. Those tools multiply a foundation; with no foundation they multiply zero.
AEO runs the same way. 98% leave AI crawlers unblocked and 7% publish anything worth citing. The opening is wide, and it is worth taking once the basics are in place.
Who builds these sites
WordPress runs over half of all sites. Just under a quarter are bespoke or can’t be fingerprinted from rendered markup; only 8% run a purpose-built recruitment platform.
Outsourced by default
38% sit behind a managed host or CDN (26% Cloudflare), yet 49% set no security headers at all and 46% have no cookie-consent layer. Where an applicant-tracking system was visible, Bullhorn and JobDiva led — for most of these firms the ATS is the only serious software layer.
8% already run a named recruitment website platform. The rest are the open market.
The plateau is universal.
Only the starting line moves.
IT/Tech — the agencies placing the very people who build this technology — sit at the bottom of the table, while the hands-on sectors of Engineering and Finance lead. Every sector average still lands inside Tier 1.
| Sector | N | Avg | Web | Blog | GA4 | Paid | Vis-ID | Lead | AEO |
|---|---|---|---|---|---|---|---|---|---|
| Engineering / Construction | 39 | 32.4 | 92% | 49% | 46% | 21% | 0% | 69% | 5% |
| Finance / Accounting | 41 | 32.3 | 95% | 54% | 76% | 22% | 2% | 46% | 2% |
| Sales / Marketing | 83 | 30.0 | 72% | 52% | 63% | 16% | 1% | 54% | 7% |
| Legal | 44 | 28.9 | 89% | 48% | 61% | 5% | 0% | 52% | 9% |
| Industrial / Logistics | 26 | 27.4 | 69% | 27% | 62% | 23% | 0% | 54% | 0% |
| Healthcare / Life Sciences | 44 | 26.2 | 80% | 36% | 43% | 20% | 0% | 45% | 9% |
| Exec / Generalist | 30 | 25.3 | 70% | 50% | 43% | 13% | 0% | 37% | 17% |
| IT / Tech | 84 | 22.9 | 80% | 36% | 37% | 17% | 2% | 43% | 5% |
Web = HTTPS + mobile responsive · Paid = any advertising/retargeting pixel · Vis-ID = any visitor-identification tool · Lead = genuine lead-capture form · AEO = flagged AEO-optimised.
The sector that places technology talent markets itself the worst of the eight.
IT/Tech recruiters score 22.9, last of eight and nine points behind Engineering/Construction (32.4). They place the engineers who build attribution models and answer engines, and their own sites are the least instrumented in the study, with GA4 on 37%, the lowest of any sector.
The likeliest cause is market structure. IT/Tech recruitment is the most crowded, most contingency-driven, most outbound-dependent niche there is. When LinkedIn and a phone reliably produce placements, the website never has to earn its keep, so it never does.
Sector only moves the starting line. Every one of the eight averages lands inside Tier 1 Foundations, so no specialism has made good marketing table stakes, which keeps it a differentiator everywhere.
Five conclusions we’d act on
Interpretation, not data. These are the judgements we draw from the figures above — the places where the numbers point clearly enough that we would change something.
Benchmark honestly.
The realistic peer set for a small agency is other small agencies. Against that peer set a Tier 1 site is normal, and there is no shame in it. The real question is whether to stay there.
Most of the gain is basic.
Half the sector is missing free, one-time installs. Analytics (54% have it), a meta description (56%), a genuine enquiry form (50%). These come before automation or AEO, and they are where the average agency’s cheapest points sit.
Above the line, the differentiators are few and consistent.
For agencies that already have the basics, the same four recur: measuring demand, capturing it actively, nurturing it, and being findable by AI. Four things done well. But only once the foundation is laid.
The most valuable missing capability is knowing who is on the site.
53% can watch traffic arrive; 1% can name the company behind it. It is the widest gap at the top of the funnel. For the sites that have finished the bottom of it.
AEO is the clearest first-mover position, in sequence.
With near-zero adoption sector-wide, the first agencies in each niche to structure content for AI answer engines face an open field. A strong second move once the basics are done, rarely the first.
How does your site compare?
Tick what your site has today. The figure beside each signal is the share of the 391 audited agencies that have it — an empty box next to a low number is an easy win few of your peers have made. Your score and tier update live on the same model the study uses.
Website & SEO25 pts
Lead-capture forms15 pts
Paid & martech25 pts
Visitor-ID / lead-gen15 pts
AEO readiness20 pts
Same 34-signal codebook and weighting as the study. This is an honest self-assessment — you tick what your own site actually has.
Every figure is reproducible.
This is first-party research. Nothing here is survey-reported or bought in. Every figure traces back to a signal read from a named agency’s live website, following a documented rule.
Sample
391 recruitment agencies, UK and US, all ≤50 employees, across eight specialisms. Firmographics from Apollo.io (NAICS 5613, Employment Services) using sector keyword tags and the 1–10 and 11–50 employee bands, capped at ~50 per specialism. Global staffing groups deliberately excluded. US 203 · UK 158 · country not returned for 30.
Detection
Each site assessed on 34 binary signals read from its rendered homepage and key inner pages (contact, blog, FAQ). Path files (robots.txt, sitemap.xml, llms.txt) validated by content. AEO readiness covers FAQ/Article schema, author / E-E-A-T markup, llms.txt and question-structured content.
Sector verification
Each agency’s specialism was confirmed from its own website during the audit rather than trusting the firmographic tag alone. Sector counts reflect post-verification classification.
Scoring model
Each group score is (signals present ÷ signals in group) × group weight. Deterministic: re-applying the codebook to the same sites reproduces each score within rounding.
Tier bands
Limitations
Tracking tags are read at homepage load, so consent-gated tags may be undercounted; site-wide tags (analytics, pixels, visitor-ID) are representative. Builder-default path files were excluded from AEO credit. Percentages are rounded. Provider detection is category-level. 367 of 391 sites responded; 24 were unreachable at audit time.
The signals in full
All 34 signals, their weight and the exact detection signature used, so anyone can collect them the same way. This is the instrument behind every number on this page.
Cite this research
Figures may be reproduced with attribution to Redsun. Every section has a stable anchor link, and the full PDF carries the same numbering.
Redsun (2026). The State of Recruitment Marketing 2026: The “That’ll Do” Line. First-party audit of 391 UK & US recruitment agencies of ≤50 employees, June 2026.
Redsun audited 391 small UK and US recruitment agencies against 34 website marketing signals. The average scored 27.9/100. 54% stop at a website, analytics and a contact form; 1% can identify visiting companies; 7% are AI-citable. Source: Redsun, The State of Recruitment Marketing 2026.
The complete 17-page PDF carries every figure on this page plus the full codebook and self-audit scorecard.
Download PDFFor most agencies the opportunity in this data is finishing the basics.
Read one way, the dataset is an indictment: an industry that sells the value of finding the right person has largely declined to apply the same rigour to finding the right client. Read another, it is useful news. The bar among realistic peers is 27.9 out of 100, and it has held there.
Nothing at the top of this sample is exotic. The highest scorers measured demand, captured it deliberately, nurtured it and stayed legible to AI, and those capabilities sit on a foundation most of the field has yet to lay. For the average agency the first wins are duller: analytics, a real enquiry form, a meta description, a blog that is genuinely kept up.
Our own bias is worth stating plainly: we build recruitment websites, and we ran this audit because we suspected the gap was wide. It turned out wider. The figures stand on their own, and the codebook is published so anyone can check them.
More research like this
We re-run this audit every year and publish new recruitment-marketing data between waves. Get the next edition and the raw signal counts when they land — no newsletter, just the research.
The few that cross it do a short, consistent set of things well.
Measuring demand, capturing it, nurturing it, and being findable by AI. Redsun builds all four into a recruitment website from day one — which is why we ran the audit in the first place.