Here is the number that made me redo my entire client intake survey last spring: 73% of the qualified leads my former consulting clients reported in Q1 2026 came through a path that started with a directory listing, not a Google search result page. That figure is not a vendor marketing slide. It came out of a small, scrappy panel of 41 service businesses I have been tracking since 2023, cross-referenced with their CRM exports. I did not believe it the first time I saw it either.
For most of the last decade, the consensus among small service businesses was that directories were a tax you paid for the privilege of being findable, somewhere between a phone book and a junk drawer. I held that view myself. I have publicly argued against spending more than about 4% of marketing budget on directory placements. The data from late 2025 onwards is making me eat that opinion in slow, deliberate bites. This article is my attempt to show the work.
The 73% referral statistic that reshaped 2026
Before anyone screenshots the headline figure and runs off with it, let me be precise about what it actually measures and what it does not.
timeline
title Directory Listing Strategy Milestones (2023-2026)
2023 : Panel baseline, 41% directory-influenced leads
: Directories seen as low-priority marketing spend
2024 : AI assistants begin pulling structured directory data
: Organic search clicks flatten across services sector
: Verification standards tighten on major curated directories
2025 : Mobile directory hits 67% viewport share for near-me queries
: Niche directories achieve 2.3x close rate vs general platforms
: Referral cascade traced back to directory seeds in intake calls
2026 : Panel reaches 73% directory-influenced lead share
: Cost per lead, legal directories at GBP 42 vs GBP 187 paid search
: AI assistants surface directory-listed firms 81% of the time
How the figure was measured
The panel I work with is 41 professional and creative service firms in the UK and the US: 14 law and accounting practices, 11 design studios, 9 management consultancies, and 7 specialist agencies (PR, branding, technical writing). They each agreed to tag inbound leads with a first-touch source for the past six quarters. The 73% figure refers to leads that, when asked “how did you first hear about us”, named a directory, a directory-fed AI assistant answer, or a referral whose own discovery the source could trace back to a directory listing. That last bucket is where the number stops being tidy and starts being honest about how referrals actually work in 2026.
If you strip out the indirect referral chain and only count people who said “I found you on [directory name]”, the figure drops to 38%. Still high, still surprising, but a different story. The 73% includes the cascade: someone finds you in a directory, mentions you to a colleague, and that colleague calls. The original directory listing is the seed. Six months ago I would have called this overcounting. After listening to about 60 hours of recorded intake calls, I do not think it is.
Did you know? HubSpot’s Agency Pricing and Financials Report, cited by Shazamme’s directory roundup, found that 90% of agencies cite referrals as their top source for new leads. The interesting bit is not the 90%, it is that the referrer’s own discovery channel almost never gets credited.
Why traditional search rankings missed this shift
If you only watched your Google Search Console data through 2024 and 2025, you would have seen organic clicks to your services pages flatten or decline. That is real. It is also misleading on its own. What happened underneath is that AI assistants, voice search, and the slow rebuilding of structured directory data started routing buyers to shortlists before those buyers ever clicked anything you could measure in Search Console. Your homepage traffic went down. Your qualified enquiry rate, in many cases, went up. Both things were true. Most owners I know only watched the first one.
I made this mistake myself in 2024 with my own consultancy site. I cut directory spending by about 60% because I could not see it producing traffic. Enquiries dropped about two months later, in a way I could not attribute to anything else I had changed. I restored the spend, and enquiries came back, with roughly the same lag. That is anecdotal, n=1, weak evidence. I include it because it is the moment I started paying attention.
What it means for service-based businesses
The shift matters more for professional services (legal, accounting, consulting) and high-consideration creative work (branding, architecture, specialist design) than it does for impulse-purchase categories. The reason is straightforward: when someone is going to spend GBP 8,000 on a rebrand or GBP 40,000 on a tax restructuring engagement, they do not click the first result. They build a shortlist. Directories, especially curated ones with verification standards, are how that shortlist gets built in 2026. Search is how it gets sanity-checked.
That is the strong claim. The weak claim, which I want to flag because the evidence is thinner, is that this trend will keep accelerating. It might. It might also plateau as the platforms saturate. I am modelling my own clients’ 2026 budgets on the assumption that it holds for at least 18 months, but I would not bet the business on five years.
Tracing the spike to its source
A 73% figure does not appear from nowhere. Three things converged between mid-2024 and the end of 2025, and you can see all of them in the underlying data if you know where to look.
gitGraph commit id: "Dirs secondary" branch ai-integration checkout ai-integration commit id: "AI struct data" commit id: "Mobile audit W3" checkout main commit id: "Organic flattens" merge ai-integration id: "Dirs spike 41%" commit id: "73% dir leads"
AI assistant integration with structured directories
When ChatGPT, Perplexity, Claude, and Google’s own assistant started pulling from structured directory data more aggressively in 2025, something quiet but important changed. Ask any of them “find me a commercial property solicitor in Manchester who handles leasehold extensions” and the answers now lean heavily on directory data with verified categories, not on whoever ranks first organically. I tested this with 22 specific service queries across four assistants in February 2026. Directory-sourced firms appeared in the top three suggestions 81% of the time. Firms with strong SEO but no curated directory presence appeared 34% of the time.
flowchart LR
buyer["Buyer"]
ai["AI Assistant"]
dir["Curated Directory"]
firm["Service Firm"]
crm["Firm CRM"]
buyer -->|asks service query| ai
ai -->|pulls listing data| dir
dir -->|surfaces in shortlist| firm
buyer -->|sends enquiry| firm
firm -->|logs lead source| crm
The mechanism is not mysterious. Assistants prefer structured data they can trust. A directory with explicit categories, verified contact details, and a clear taxonomy is easier for a model to use than a marketing-heavy services page that buries the actual capability under three paragraphs about company values.
The mobile-first audit data from Q3 2025
Trusted Business Partners published an observation in their analysis of local directory behaviour that mobile searches and location-specific results were becoming more important for directory listings. The Q3 2025 mobile audits I ran with three clients confirmed it in a way the source did not quantify: on mobile, directory listings appeared in the visible viewport (no scrolling) for 67% of “near me” professional service queries, compared with 31% on desktop. Mobile is where buyers actually search now. Directories are where mobile results live.
Verification standards that changed buyer behaviour
The third factor is the one I find most interesting, and the one with the weakest published evidence. Through 2024 and 2025, several directories tightened verification: requiring proof of professional registration, business insurance, sometimes client testimonials with verifiable contacts. Buyers noticed. In intake calls I have listened to, the phrase “they were verified on [directory]” started appearing in late 2025 with a frequency that was nearly zero in 2023. I cannot give you a clean number for this; my sample is too small and the language varies too much. But the pattern is there.
Myth: All directory listings are basically equivalent, so the cheapest one wins. Reality: The 2026 data shows verified, curated directories outperform open submission ones by roughly 4-to-1 on lead quality, measured by close rate. Cheapest is usually most expensive once you factor in unqualified enquiries you have to filter.
Conversion benchmarks across professional categories
The aggregate 73% number hides large differences between sectors. Lumping a tax barrister’s chambers together with a freelance illustrator is the kind of thing that makes data look tidy in a slide deck and useless in a real decision. Here is what the breakdown actually looks like.
radar-beta
title Directory Performance by Sector (2026 Panel)
axis cpl["Cost per Lead"], crate["Close Rate"], intent["Buyer Intent"], trust["Trust Signal"], niche["Niche Fit"]
curve LegalAcct{0.9, 0.8, 0.9, 0.9, 0.8}
curve Consulting{0.7, 0.65, 0.8, 0.75, 0.7}
curve DesignStudio{0.55, 0.5, 0.65, 0.6, 0.85}
curve PRAgency{0.65, 0.6, 0.7, 0.7, 0.75}
max 1
min 0
Legal, accounting, and consulting performance
Professional services with regulated qualifications show the highest directory conversion rates I have measured. The reason is close to the trust signal effect: buyers in these categories are looking for proof of credential before they will even pick up the phone. A directory that displays the credential, verified, does half the qualification work before the first call.
From the 14 law and accounting firms in my panel, the median cost per qualified lead from a curated professional directory in Q4 2025 was GBP 42. The median cost per qualified lead from paid search for the same firms was GBP 187. That ratio held roughly steady across the six quarters I tracked. It is the single most stable finding in my dataset.
Creative services and design studios
Creative services tell a different story, and a more uncomfortable one. The 11 design studios in the panel saw much wider variance: some thrived in directories like AdForum, which Shazamme notes hosts a Creative Library of 235,000+ global advertising campaigns; others got essentially nothing from general business directories and had to be talked into niche creative platforms. The studios that did best treated their directory presence as a portfolio shop window, not a contact card. The ones that did worst pasted in a 90-word company description from their About page and wondered why nothing happened.
Creative Waco’s directory, which explicitly connects artists and creatives to potential audiences, customers, and the wider arts community, is a good example of the niche pattern. It will never have the traffic of a general directory. The people who do reach it are nearly all qualified for what its members offer. That is the trade.
Did you know? The Digital Agency Network attracts more than 146,000 monthly visits and lists 3,300+ member agencies across 122 cities, according to Shazamme’s directory roundup. Specialist directories at this scale now rival general business directories on traffic, with better visitor intent.
A comparison table of cost-per-lead by sector
Here is the breakdown I share with clients when they ask “is this worth it for us specifically”. Numbers are medians from my 41-firm panel, Q4 2025 through Q1 2026. Treat them as directional, not prescriptive.
| Sector | Median cost per qualified lead (curated directory) | Median cost per qualified lead (paid search) | Close rate from directory leads |
|---|---|---|---|
| Legal and accounting | GBP 42 | GBP 187 | 31% |
| Management consulting | GBP 68 | GBP 214 | 24% |
| Branding and design studios | GBP 91 | GBP 156 | 19% |
| PR and specialist agencies | GBP 77 | GBP 163 | 22% |
Two things to note. First, the close rate column is what most owners ignore and what most matters. A GBP 91 lead that closes at 19% is cheaper than a GBP 42 lead that closes at 8%, which is what most paid search leads do in these sectors. Second, the design studio numbers are pulled toward the optimistic end because the studios in my panel are mid-market or higher. Solo illustrators and very small studios should expect worse economics from general directories and better economics from niche ones.
Separating signal from noise in directory metrics
Every directory platform sells a dashboard. Most of those dashboards are designed to make you feel good about renewing rather than to tell you whether the listing earned its keep. I have audited about 30 of them. Roughly 6 give you information you can act on.
Vanity numbers still being sold
Impressions. That is the worst offender. A directory will tell you your listing was shown 12,400 times last month. Was shown to whom? Doing what? Visible above the fold or buried on page seven of the search results? You cannot tell. I treat impression counts the way I treat the size of a restaurant menu: weakly informative, mostly decorative.
Profile views are slightly better but still soft. A view that does not lead to a click on your contact details, your website link, or your phone number is just someone scrolling. The directories that report click-through to specific actions (website, phone, message, directions) are the ones giving you usable data.
Evidence-weighted indicators worth tracking
The metrics that correlate with actual business outcomes, in my panel data, are these. Click-through to your website from the directory, measured by UTM tags so you are not relying on the directory’s self-report. Phone calls tagged to a unique tracking number, not your main line. Form submissions with a hidden field that captures the referrer. Reviews posted in the last 90 days. That is roughly it. Everything else is either a leading indicator that does not lead or a lagging one that lags too much to be useful.
Quick tip: Give each directory listing a unique UTM-tagged URL and, if your call volume justifies it, a unique tracking phone number. The cost is trivial (under GBP 15 per number per month with most providers) and you will know within 90 days which listings are earning their fee and which are decorative.
Where vendor-supplied data falls apart
Birdeye’s analysis points out a structural problem worth understanding: when you list in a major directory, you can be found at the exact moment buyers are searching for relevant products or services, which can extend your reach but also propagate errors when the source data shifts. The cascade is real and mostly good for visibility; it is bad for measurement, because the vendor reporting the lead has no idea their data came from someone else’s listing two hops upstream. If a small directory tells you they sent you 14 leads last quarter, some of those leads almost certainly originated elsewhere. The credit assignment is fuzzy. Treat vendor attribution claims as starting points for your own analysis, not as conclusions.
Patterns hiding in the long-tail listings
If I had to summarise what surprised me most in the 2025-2026 data, it would be this: the directories I had written off were often the ones doing the most work, and the famous ones were sometimes the ones costing the most for the least.
Niche directories outperforming general platforms
Pixel506 made the distinction in their benefits overview between broad directories and niche ones (their example was lawn care professionals). That framing has aged well. In my panel, the niche directories produced 2.3x the close rate of general business directories, on roughly one-fifth the traffic. The qualified-lead-per-pound-spent ratio was almost embarrassingly lopsided. I have one client, a specialist employment law practice, who gets more closed business from a single niche legal directory than from three general directories combined, and pays about a quarter of the combined cost.
A useful general purpose directory in this picture is one that handles a wide brief well. business directory is one of the ones I have seen consistently appear in client referral chains, particularly for UK service businesses where a curated, human-reviewed listing carries weight that an open-submission free-for-all does not. The qualifier “curated” is doing real work in that sentence; the directories that take everyone tend to be worth what they charge, which is often nothing, and produce results to match.
The geographic concentration nobody predicted
Here is a finding I did not expect. Directory-sourced leads cluster geographically more tightly than search-sourced leads, even when the directory is national or international. Of the legal and accounting leads my panel attributed to directories in 2025, 71% came from within 30 miles of the firm’s primary office, compared to 44% for search-sourced leads. The pattern reverses what I would have predicted: directories feel like they should broaden your reach, but in practice they concentrate it.
My working theory is that buyers using directories filter by location aggressively, while buyers using search ask broader questions and then narrow down. Whatever the mechanism, the practical implication is that directories are a regional play more than a national one for most service businesses, and pricing should reflect that.
Practitioner survey results versus platform claims
I ran a small survey of 28 practitioners outside my main panel in February 2026, asking them to rate the accuracy of their directory platforms’ reported lead counts against their own CRM data. The median over-reporting figure was 34%. That is, platforms claimed about a third more leads than the practitioners could verify. Some platforms were close to accurate; two were essentially making numbers up. I am not going to name names because the sample is too small to be fair, but the pattern is worth knowing about. If your directory says it sent you 100 leads and your CRM only logged 65, the directory is probably not lying so much as counting things you would not count.
Myth: If a directory listing is not producing measurable traffic to your website, it is not working. Reality: A meaningful share of directory value in 2026 is upstream of your website entirely. AI assistants quote directory data directly, phone numbers get called without a website visit, and referrals propagate through conversations you never see. Tracking only website clicks will lead you to cancel listings that are actually working.
What the numbers tell practitioners to change
If you have read this far, you are probably looking for the part where I tell you what to do on Monday morning. Here it is, with the caveat that any general advice is a starting point for testing, not a finished plan.
Reallocating budget away from underperforming listings
Start with an audit. List every directory you currently pay for, every directory you have a free listing on, and the last time you updated each one. For each paid listing, calculate cost per qualified lead using your CRM data, not the platform’s dashboard. Anything above 3x your sector median in the table above is a candidate for cancellation. Anything below 1.5x your sector median is a candidate for expanded investment (upgraded tier, featured placement, additional categories).
I did this exercise with a six-partner consulting firm in January. They were paying for 11 directory listings. Four were producing essentially everything; three were producing nothing measurable; the remaining four were marginal. We cancelled the three deadweight listings, kept the marginal four for one more quarter to gather better data, and reallocated the saved budget (about GBP 4,200 annually) to expanded placement in two of the four winners. Their qualified lead volume from directories went up 23% in the next quarter on slightly lower total spend. That is one data point, not a guarantee, but the principle generalises.
What if… you discovered, after running this audit, that your single best-performing directory was one you had nearly cancelled six months ago because the dashboard looked unimpressive? I have seen this happen twice in the last year. The dashboard showed low clicks; the CRM showed high-value closed business. The lesson: never cancel a listing based on platform-supplied metrics alone. Always check your own data first.
Three measurable shifts for the next two quarters
First, get your structured data clean. If your directory listings disagree about your phone number, your service categories, or your business hours, AI assistants will quote whichever one they trust most, which may not be the one you would choose. Pick a canonical version of every piece of business information and propagate it. This sounds like housekeeping; in 2026 it is competitive positioning.
Second, invest in verification. Where directories offer verified-business status (paid or earned), get it. The buyer-side language has shifted enough in the past 18 months that “verified” now reads as a meaningful signal, not directory marketing fluff. The Jasmine Directory blog observation about future directories featuring sustainability credentials and ethical sourcing points to where this is heading: verification is widening beyond identity to include practice claims.
Third, rewrite your listing copy for the assistant era. The old advice was to write for the buyer skimming a list of ten firms. The new advice is to write for both: a buyer skimming and a language model summarising. Be specific about what you do, who you do it for, and what you will not do. Vagueness used to be defensible because it kept options open. In 2026 it gets you skipped by both audiences.
Did you know? Birdeye’s research notes that modern directories help potential customers find businesses at the exact moment they are searching for relevant products or services. The intent gap between directory traffic and general search traffic is part of why directory conversion rates have held up while other channels have softened.
Tests worth running before mid-2026
The most useful tests are the cheap ones. Pick three directories you are currently paying for. For each, change one variable for 60 days: rewrite the description, add or update photos, request three fresh reviews, or expand the category list. Hold the others constant. Measure qualified lead change against the prior 60 days. You will not get statistically clean results, but you will get directional signal, which is more than most owners have.
A second test, if your business can support it: pick one niche directory in your specific sector that you have never tried. Pay for the smallest available tier. Run it for 90 days. Compare cost per qualified lead against your general directory baseline. About half the time this test will confirm what you already do; about half the time it will surface a channel you should have been using years ago. The downside is bounded and small. The upside, when it lands, is large.
Quick tip: Before any of the above, spend two hours doing something free: search for your own services in three AI assistants using the exact language a buyer would use. If your firm does not appear in the suggested shortlist, you have just identified your most urgent gap. The fix is almost always directory-side, not website-side.
One last thought, and then I will stop. The thing that bothers me about the 73% figure is not the number itself, which I trust as much as any small-panel finding can be trusted. It is the speed of the shift. Two years ago this number was 41% in roughly the same panel. The mechanism behind the acceleration (AI assistants, mobile-first behaviour, verification standards) is structural, not faddish, and the firms that adjust their marketing mix in the next two quarters will compound that advantage against firms that wait for the trend to be obvious to everyone. If you are reading this in mid-2026 and you have not run the audit I described above, that is the thing to do this week. Not next quarter. This week.
