Here is the uncomfortable truth I learned running a service business for eight years: most of the directory advice you read online is written by people who have never paid for a listing out of their own bank account. They rank directories by Domain Authority, then publish a top-50 list, then move on. That worked in 2018. It does not work now.
What follows is a framework I have built and refined across two businesses and roughly 60 directory audits — RICE-D. It is not elegant. It does not promise overnight rankings. What it does do is stop you wasting Saturday mornings filling in listing forms for directories that will never send you a single qualified lead.
Why the old “DA + traffic” model fails in 2026
Did the backlink model ever really work? What changed between 2018 and now? And why do so many businesses still chase Domain Authority as if it were a ranking signal?
The old model — pick the directory with the highest Domain Authority, file your NAP, move on — assumed that backlinks from high-DA domains pulled rankings up like a tide. For a few years they did. Then Google stopped treating most directory links as endorsements — the AI search layer arrived on top of the SERP — and the directories themselves either professionalised or collapsed into spam. The DA number on a Moz dashboard kept rising. The phone kept not ringing.
The vanity metric trap
Domain Authority is a third-party estimate of link equity. It is not a measure of buyer intent, lead quality, or whether the directory’s audience overlaps with yours at all. Apple Maps Connect carries a DA of 100 according to W3Era — which sounds like a slam-dunk reason to claim your listing. For an iOS-heavy local restaurant it probably is. For a B2B compliance consultancy selling to procurement teams, a DA-100 map listing is decorative.
I made this mistake myself in 2019. I chased DA, ignored fit — and ended up on twelve directories that drove three calls in a year, two of them wrong-number.
What Google’s 2025 spam updates broke
The directories that survived recent spam updates share a pattern: human curation, real editorial standards, and verifiable business records. The ones that lost visibility were the auto-scraped citation farms that recycled the same business data across a hundred subdomains. If your strategy in 2024 was “submit to 200 directories with a script” — a chunk of your citation graph has quietly vaporised. Audit it.
Hidden costs of low-intent directories
A free listing is not free. Each one demands a writeup, a category choice, a verification call, and ongoing maintenance when your address or hours change. Premium upsells, support delays, and inconsistent moderation all add real cost on top of the headline price. Multiply that by 40 directories and you have built yourself a part-time job with no salary.
Myth: More directory listings always mean more SEO benefit. Reality: After roughly the top 15-20 relevant citations, additional listings deliver diminishing returns and can introduce NAP inconsistencies that actively hurt you.
Introducing the RICE-D evaluation framework
How do you move past Domain Authority? What single framework can replace the old checklist? And can you really score a directory in four minutes?
RICE-D scores each candidate directory on five factors — Relevance, Intent, Citation weight, Exposure, Durability — and produces a number out of 100. You can score a directory in about four minutes once you have done a few. The point is not the precision of the score; the point is forcing yourself to look past DA before you commit time.
Relevance, intent, citation weight, exposure, durability
Relevance asks whether the directory’s audience overlaps with your buyer. Intent asks where in the funnel the visitor sits when they arrive. Citation weight is the SEO contribution — the part the old model fixated on. Exposure is the raw reachable audience size. Durability — and this is the factor most people skip — asks whether the directory will still matter in 18 months.
Durability is the hidden multiplier. A directory with strong relevance and weak durability is a sandcastle.
How the five factors interact
The factors are not additive in the simple sense. Low Relevance caps the upside of everything else. A perfectly durable, high-citation directory aimed at the wrong audience still sends you nothing. I weight Relevance and Intent at 25 points each, Citation weight at 20, Exposure at 15, Durability at 15. Anything scoring under 50 gets rejected. Anything 50-70 goes on a maybe list. Above 70 gets a listing within the month (see Figure 1).
stateDiagram-v2 [*] --> Candidate Candidate --> Scoring : add to shortlist Scoring --> Rejected : score below 50 Scoring --> Watchlist : score 50 to 70 Scoring --> Active : score above 70 Watchlist --> Scoring : quarterly review Active --> Scoring : quarterly review Active --> Dropped : durability collapse Rejected --> [*] Dropped --> [*]
Scoring directories on a 100-point scale
Here is the rubric I use. Keep it on one page; if your scoring guide needs a wiki, you will stop using it.
| Factor | Weight | Score 1-5 means | Typical evidence | Common mistake |
|---|---|---|---|---|
| Relevance | 25 | Audience overlap with your ICP | Category depth, peer brands listed | Confusing reach with relevance |
| Intent | 25 | Funnel stage of arriving visitor | “Compare” pages, request-quote CTAs | Ignoring browse vs. Buy distinction |
| Citation weight | 20 | SEO contribution, NAP propagation | DA, indexation, schema markup | Treating DA as the whole picture |
| Exposure | 15 | Reachable audience size | Traffic estimates, app installs | Trusting self-reported numbers |
| Durability | 15 | 18-month survival probability | Funding, editorial standards, age | Skipping it entirely |
| Pass threshold | 50/100 | Minimum to justify time | Composite weighted score | Adding without weighting |
Did you know? OnToplist reports that 80% of consumers search online for local businesses at least once per week, and 31% of top-10 organic results for average local searches are business directory pages — meaning your competitors are already occupying the SERP slots you want.
Applying relevance and intent filters
What does relevance look like when you have to score it in four minutes? What separates a high-intent listing visit from a low-intent one? And how do you stop yourself rationalising every directory upward because the form felt productive?
Vertical match versus horizontal reach
Horizontal directories — Google Business Profile, Yelp, Bing Places — give you reach but treat you as one of millions. Vertical directories give you fewer eyeballs but a far higher proportion who care about your specific service. Industry-specific directories consistently produce higher-quality leads than generalist ones — particularly in services with comparison-heavy buying behaviour.
The rule I use: every business needs the top three horizontal citations. Beyond that, vertical wins.
Buyer-stage signals in listing traffic
Read the directory’s category pages like a buyer. Does the page have a “Request quotes from 3 providers” button? That is a bottom-funnel visit. Does it have a “Top 10 in [city]” editorial article? That is mid-funnel. Does it have a glossary and a beginner’s guide? That is top-funnel — fine for awareness, terrible for short-term leads. Score Intent accordingly.
Worked example: a local HVAC business
Take a hypothetical Birmingham HVAC firm, eight engineers, residential and light commercial, average ticket around 2,400 GBP. Their first instinct is to chase national directories with DA above 85. Wrong call. Their buyers either ask Google “boiler repair near me” at 7am on a cold morning, or they ask a neighbour. The candidate set narrows fast.
RICE-D scores for this firm: Google Business Profile lands at 92 (relevance 25, intent 24, citation 18, exposure 15, durability 10, the durability is capped because Google moves the goalposts every year). Checkatrade scores 84, vertical match, transactional intent, established brand. Yelp UK scores 51, barely passes because British consumers under-use it. A generic “UK Business Directory” with DA 78 but no editorial curation scores 34. Reject.
Quick tip: If a directory does not let you see a real competitor’s listing before you sign up, score Exposure no higher than 2. Hidden audiences are usually small audiences.
Measuring citation weight and exposure
Citation weight in 2026 is not the same beast it was in 2020. The link itself matters less. What matters more, and this is the shift most owners have missed, is whether your business data propagates cleanly to the systems that feed AI search overviews.
NAP consistency in the AI-search era
NAP, Name, Address, Phone, consistency was always important for ranking. It is now important for whether ChatGPT and Google’s AI overviews quote your business correctly when someone asks “who does emergency plumbing in Leeds”. The large language models trained on the open web learned your business from these directories. If half of them list your old address, the AI averages the wrong answer and confidently recommends a building you moved out of in 2022.
In practice, I have audited maybe 30 small businesses in the past year. Roughly 80% had at least one stale citation. Two had eleven.
LLM training data inclusion as a signal
Directories whose pages are crawled, indexed, and ingested into LLM training sets carry citation weight the old DA model cannot see. The proxy is rough but workable: is the directory’s structured data clean, is robots.txt permissive, do its listings appear in Google’s AI overviews when you test buyer-intent queries (see Figure 2)? If yes, score citation weight high regardless of DA.
requirementDiagram
requirement nap_consistency {
id: 1
text: business NAP must match across all active citations
risk: high
verifymethod: inspection
}
requirement schema_present {
id: 2
text: directory listing must expose LocalBusiness schema
risk: medium
verifymethod: test
}
requirement durable_domain {
id: 3
text: directory must show 18 month survival signals
risk: medium
verifymethod: analysis
}
element rice_d_audit {
type: quarterly review
}
rice_d_audit - satisfies -> nap_consistency
rice_d_audit - satisfies -> schema_present
rice_d_audit - satisfies -> durable_domain
Worked example: a B2B SaaS startup
Consider a Series-A SaaS company selling contract-review software to mid-market legal teams. Their buyer never types “contract software near me”. They search G2, Capterra, Software Advice, and increasingly they ask an LLM directly. Google Business Profile? Useful for brand search and the recruiter side, almost irrelevant for pipeline. The RICE-D scores skew the opposite way from the HVAC firm: G2 hits 88, Capterra 81, Product Hunt 62 (high exposure, low intent), Google Business Profile a surprising 58, high citation weight, low relevance to their actual buyer.
Myth: SaaS businesses do not need traditional directory presence. Reality: They need a different directory mix, review platforms, comparison sites, and category leaders, but the underlying RICE-D logic is identical.
Full walkthrough: a regional law firm in 2026
How do you build a candidate list from scratch? What does the scoring process actually look like in practice? And how many directories should a professional services firm realistically maintain?
Let us run the framework from start to finish. Imagine a six-partner law firm in Manchester, mixed practice, family, commercial property, employment. Existing online presence is thin: a claimed Google Business Profile that has not been updated in two years, a Law Society listing, and a Yelp page nobody manages.
Building the candidate list of 40 directories
The candidate set comes from four sources. First, the obvious horizontal trio, Google, Bing Places, Apple Maps. Second, the vertical legal directories: Law Society Find a Solicitor, Chambers, Legal 500, ReviewSolicitors, The Law Superstore. Third, regional and city directories: Manchester Evening News business listings, local chambers of commerce, BNI chapters that publish member directories. Fourth, the edited general directories, including Business Web Directory, and a couple of niche UK lists.
You will not list on 40. You will score 40 to find your 12.
Running each through RICE-D scoring
The scoring shakes out roughly like this. Law Society Find a Solicitor scores 91, perfect relevance, transactional intent, durable by statute. Chambers and Legal 500 score in the high 80s but require editorial relationships you cannot buy. ReviewSolicitors scores 79. Google Business Profile scores 81, high citation weight saves it despite mixed intent. Bing Places scores 64 (Microsoft’s Copilot integration boosts durability). Three of the regional directories score above 65. About 22 candidates score below 50 and get cut (see Figure 3).
block-beta columns 3 A["Horizontal trio"] B["Vertical legal"] C["Regional citations"] D["RICE-D scoring sheet"]:3 E["Active 12"] F["Watchlist 6"] G["Rejected 22"]
Final picks and projected 18-month return
Final active set: 12 directories. Three horizontal, four vertical legal, three regional, two edited general. Projected effort: roughly 14 hours to set up cleanly, then 90 minutes a month for maintenance and review responses.
Projected return is harder to pin down. What I have seen in practice: directory-focused firms see meaningfully higher inquiry volume. For this law firm, an honest projection over 18 months is 25-40% lift in inbound inquiries, concentrated in family and employment work where buyer search behaviour is most directory-influenced. Commercial property will move less because that work travels by referral.
Did you know? Central Media notes that Google holds over 90% of search market share, which is why Google Business Profile is the only directory that, even when it scores poorly on relevance for your niche, still earns a listing on citation weight alone.
What if… your top-scoring directory gets acquired by a competitor’s parent company mid-year? It happens, Yelp has reshuffled partnerships, smaller vertical directories have been rolled up into media groups. Durability is not just survival, it is independence. Drop the score by 4 points if a directory’s owner has a conflict of interest with your category.
Edge cases the framework handles poorly
RICE-D is not magic. There are situations where I have watched it produce misleading scores, and you should know about them before you trust it with a budget decision.
Hyperlocal directories with no public metrics
A village business association website with 400 monthly visitors and no DA score will read as worthless under RICE-D. Sometimes it is. Sometimes it is the single most important citation in your portfolio because every local buyer in a 5-mile radius checks it before hiring. The framework cannot see this. You have to ask your customers where they found you and weight accordingly.
Paid placements that distort scoring
Some directories sell top-of-category placement. The placement boosts your Exposure score artificially, you are not seeing organic exposure, you are renting it. When you cancel the spend, the score drops. Always score the unpaid version of your listing. If the directory only works when you pay, treat it as advertising, not citation.
Industries with no dominant vertical directory
Some niches, bespoke industrial services, certain B2B consultancies, anything where the buyer pool is under 5,000 people globally, have no meaningful vertical directory because the audience is too small to sustain one. RICE-D will keep returning low scores and you will conclude directories do not matter for you. That conclusion is half right. Horizontal citations still matter for the trust signal in AI search; the framework just stops being useful past that minimum.
Myth: If RICE-D rejects most directories in your industry, the framework is broken. Reality: It is telling you something true, your channel mix should weight elsewhere, probably toward conferences, referrals, and content. Listen to it.
Putting RICE-D into your monthly workflow
A framework that lives in a Google Doc gets used once and forgotten. The version that survives is the one that integrates into work you already do, invoicing day, monthly accounts close, the Friday afternoon when nothing else is moving.
Quarterly re-scoring cadence
Score every active directory once a quarter. Not every directory in existence, just the ones you currently hold a listing on, plus any new candidates someone has flagged. The re-score takes 30 minutes for a 12-directory portfolio. What you are looking for is movement: a directory that dropped from 78 to 61 is signalling something. A directory that climbed from 60 to 75 deserves more attention in your content updates.
Directory (Latin directorium, a guide) was never a static list, it was an instrument that pointed somewhere, and instruments need recalibrating.
When to drop a previously strong directory
Drop signals I trust: two consecutive quarters of score decline, ownership change to a competitor, removal of review functionality, sudden category page redesigns that bury small businesses below paid placements. Do not drop on a single bad quarter; directories have noisy traffic patterns. Do drop when the durability factor collapses, because once durability goes, the rest follows.
What you watch for is the order of collapse. Trust goes first, then attention, then revenue.
Tracking sheet template and inputs
Your tracking sheet needs nine columns: directory name, URL, claim date, R score, I score, C score, E score, D score, composite, last-reviewed date, status (active/watchlist/dropped), inbound leads attributed last quarter. That last column is the honesty mechanism, it forces the framework to confront actual results, not just scoring elegance.
I keep mine in a single Google Sheet, one tab per business if you run several, and I attribute leads through a unique tracking phone number per high-scoring directory (see Figure 4). Yes, tracking numbers cost roughly 3 GBP each per month. Yes, it is worth it, you cannot defend a directory’s score with anecdote when the attributed-leads column says zero.
kanban column-0-header Score this quarter column-1-header In progress column-2-header Done this quarter Audit Google Business Profile,High Re-score top 5 vertical,High Update NAP after office move,owner Refresh 3 listing descriptions,marketing Dropped 2 stale citations,Q1-014 Added regional chamber listing,Q1-022
The framework is not the point. The discipline is. RICE-D works because it forces a thirty-minute pause before you spend an hour filling out a form for a directory that will never pay you back. If you remember nothing else, remember this: score before you submit. Reject before you regret.
For the rest of 2026, I would put my money on three trends. The directories that survive will be those whose data feeds AI search overviews cleanly. The directories that grow will be the edited vertical ones, because LLMs build on human editorial judgement they cannot replicate. And the businesses that win will not be the ones with the most listings; they will be the ones whose twelve listings all point to the same address, the same phone number, and the same story. Pick your twelve. Score them honestly. Then go answer the phone.
