Quality over quantity: how many directories to submit to

A founder rang me last spring, frustrated. His agency had built him “a strong backlink foundation” with 47 directory listings. Traffic was flat. Leads were flat. The invoice was not flat. It had grown every quarter for two years.

I told him I would audit before adding anything. He thought I was stalling. I was not. I was about to find that six of his 47 listings were doing almost all the work, and the other 41 were, at best, decorative.

What follows is the walkthrough of that engagement, the numbers, the calls I got wrong, and the framework I now use on every directory audit. If you are about to commission a fresh batch of submissions because someone told you “more is better”, read this first.

The client situation: a B2B SaaS with 47 directory listings

The client sells a mid-market workflow tool. Annual contract values sit between GBP 8,000 and GBP 35,000. Sales cycle is 60 to 90 days. Marketing is one in-house generalist plus me on a fractional retainer. Pipeline target last year was roughly GBP 1.2m in new ARR.

Nothing exotic. The kind of business where every channel needs to justify itself with a number, not a vibe.

Inherited backlink profile from previous agency

The previous agency had run what I would politely call a volume play. Between January 2022 and December 2023 they submitted the company to 47 directories. The mix was roughly half general business directories, a quarter “tech” or “SaaS” listings of varying credibility, and a quarter local or regional sites that made no sense for a company selling internationally.

Cost over those two years, including listing fees, premium placements, and the agency’s submission labour, was just over GBP 18,400. I have the spreadsheet. It is bleak reading.

Traffic plateau despite “good” directory count

Organic referral traffic from directories had flatlined at around 380 sessions per month for nine consecutive months. Of those, the booked-demo conversion rate from directory referrals sat at 0.6%, so roughly two to three demos per month from the entire directory portfolio.

The agency’s monthly report said things like “domain authority of submitted directories increased by 4 points” and “completed five new high-DR submissions this quarter.” Nobody on the client side could tell me what those sentences meant for revenue. Neither could the agency, when I asked.

Why we audited before adding anything

My rule, learned the hard way on a fintech project in 2019: if you cannot explain why the existing assets are not working, do not add new ones. You will just produce more of the same failure at scale.

There is a useful parallel in academic research, where the same temptation exists. A piece in The Conversation describes how metric-led publishing pushes researchers into “salami slicing”, spreading thin findings across many papers to inflate counts. Directory submission run badly does the same thing to a brand: lots of mentions, no substance, declining trust signals.

Sorting signal from noise across the existing 47

Before deciding what to add, I needed to know what was actually earning its place. I gave myself two weeks and three data sources: Google Analytics 4, the CRM (HubSpot in this case), and the directories’ own dashboards where they offered any reporting.

erDiagram
  DIRECTORY ||--o{ LISTING : "has"
  LISTING ||--o{ REFERRAL_SESSION : "generates"
  REFERRAL_SESSION }o--o{ OPPORTUNITY : "attributed to"
  OPPORTUNITY ||--o| CONVERSION : "becomes"
  DIRECTORY {
    string name
    int domain_rating
    float annual_cost_gbp
    bool human_curated
    bool category_page_ranks
  }
  LISTING {
    string url
    date last_refreshed
    string description
    bool has_screenshots
    bool has_reviews
  }
  REFERRAL_SESSION {
    date session_date
    string source_page
    int session_count
  }
  OPPORTUNITY {
    float contract_value_gbp
    int days_to_close
    string status
  }
  CONVERSION {
    float closed_won_gbp
    date close_date
  }

The referral traffic filter I applied first

First pass was brutal and quick. Any directory that had sent fewer than 5 sessions in the previous 12 months went into a “probable cut” pile. That removed 19 listings immediately. A few of those were on directories that had genuinely died; the sites were still up but had clearly lost their own audience.

sankey-beta
  47 Directories,Zero Traffic (27),27
  47 Directories,Low Traffic (8),8
  47 Directories,Active (12),12
  47 Directories,Top Six,6
  Zero Traffic (27),No Leads,27
  Low Traffic (8),No Leads,5
  Low Traffic (8),Qualified Leads,3
  Active (12),Qualified Leads,6
  Active (12),No Leads,6
  Top Six,Qualified Leads,32
  Qualified Leads,Opportunities (41 total),41
  No Leads,Dropped,38

Another 8 listings had sent between 5 and 30 sessions in 12 months. I parked these as “maybe”, because session count alone is a thin signal. A directory that sends 20 sessions a year but two of them become GBP 20,000 contracts is not a cut candidate.

That left 20 directories sending meaningful traffic. Now the real work started.

Domain authority versus actual conversion data

Here is where I disagree with most directory advice you will read. Domain authority and similar third-party scores are useful context, not selection criteria. I have watched DR 70 directories send zero qualified leads in a year while a DR 28 niche site sent eleven.

For each of the 20 candidates I pulled, in order: referral sessions, demo bookings attributed in HubSpot, opportunities created, and closed-won revenue. I went back 18 months because the sales cycle is long and I did not want to penalise a directory that had sent a great lead in month 14 of the lookback window.

Six listings driving 80% of qualified leads

The result, when I sorted by opportunities created: six directories were responsible for 32 of the 41 qualified opportunities the channel had produced in 18 months. That is 78%. Pareto, predictable, still a bit shocking when you see it on your own client’s data.

The six were a mix I would not have predicted before running the numbers. Two were vertical SaaS review sites where buyers actively compare options. One was a regional business chamber listing, of all things, that happened to sit on page one for a specific phrase the client’s buyers used. One was a curated general directory with editorial review. Two were category-specific directories I had never heard of until this project.

The other 14 of the 20 candidates? Between them, 9 opportunities in 18 months. The 27 we had already filtered out: 0 opportunities. So 47 listings, six doing the work.

The shortlist that replaced volume thinking

Once you know which six listings are pulling weight, you stop asking “how do we get to 70 directories?” and start asking “what do these six have in common, and where else does that pattern exist?”

flowchart LR
  buyer["Target Buyer"]
  verticalDir["Vertical Review Sites"]
  generalDir["Curated General Dir"]
  saas["SaaS Product"]
  crm["HubSpot CRM"]

  buyer -->|compares tools| verticalDir
  buyer -->|researches category| generalDir
  verticalDir -->|refers traffic| saas
  generalDir -->|refers traffic| saas
  saas -->|logs demo bookings| crm
  crm -->|attribution report| saas

Building criteria from the winning six

I sat with the six winners and looked for shared attributes. Four stood out:

• Each had a search-visible category page that ranked on page one for at least one buying-intent query in the client’s space.
• Each had recent editorial activity; someone was actually curating, removing dead listings, adding new ones.
• Each let the listing include enough detail (proper description, screenshots, pricing context, integrations) for a buyer to self-qualify before clicking through.
• Each had either a review mechanism with real reviews, or a strict editorial process. Pure pay-to-list with no curation never appeared in the winners.

Those four attributes became my screening criteria. I then went hunting for other directories that matched.

Industry-specific versus general directories

The instinct, when you find that vertical directories outperform, is to abandon general ones entirely. I almost did. Then I looked again at one of the winners, a general directory with serious editorial standards, and noticed something. The leads it sent were larger on average than the vertical-directory leads. Lower volume, higher contract value.

My working theory: the buyers using vertical comparison sites are already deep in evaluation, often comparing three to five tools. The buyers landing on the curated general directory were earlier in their thinking, often researching the category itself. They took longer to close but bought bigger packages because they had room to scope properly.

So the shortlist needed both. I added two new vertical directories the client was not on, and one carefully chosen general directory in the same editorial mould as the winner. For the general slot I picked Web Directory because it screens submissions, organises by category in a way that mirrors how buyers search, and does not bury paid placements as if they were editorial picks. That matters for the kind of buyer who notices.

Why I rejected 23 “high DR” options

I had a list of 31 potential new directories the previous agency had recommended but not yet submitted to. After applying the four criteria above, 23 were out. The rejections fell into roughly four patterns:

1. Directories with high DR but no organic search traffic to their category pages, so the link sits on a page nobody visits.
2. Directories with no curation, where the client’s listing would sit next to obvious spam and adult content; this is a reputational risk, not just an SEO one.
3. Directories with a “premium” tier costing GBP 400-GBP 900 per year that, on inspection, sent traffic comparable to free listings on sites I already valued more.
4. Directories whose audience was geographically irrelevant.

Submission economics over a 90-day window

Once the shortlist existed (8 directories: the 6 keepers plus 3 new candidates, minus one I would replace) I built the economics. Spreadsheet, three tabs, nothing fancy.

Cost per qualified lead by directory tier

Below is the comparison table I built for the client, rebuilt here with slightly rounded numbers for confidentiality. “Tier” is my shorthand, not a formal industry term.

The bottom row is the one to internalise. The “free” submission to an uncurated directory costs roughly 15 to 25 minutes of someone’s time. If it produces zero leads forever, the cost per lead is undefined. Worse, the listing may appear next to content that damages how an editorial search ranker views your brand association.

Time investment versus paid placement math

A proper directory submission, done well, takes me about 35 to 50 minutes the first time. That includes writing a tailored description (not the same boilerplate copied 47 times, which is what most agencies do), choosing appropriate categories, uploading suitable images at correct dimensions, and verifying the listing once it goes live.

At GBP 75 per hour blended rate, that is roughly GBP 45 to GBP 65 of labour per submission. So a “free” listing on a directory that sends one qualified lead a year and produces two demos that never close has an effective acquisition cost of, well, more than the lead is worth.

Paid placements need to clear a higher bar because the money is visible. Free placements get a free pass that they often do not deserve.

The breakeven I model for each candidate

For each candidate directory I calculate a simple breakeven: how many qualified leads per year do I need for this listing to clear its cost (paid placement plus labour)? Then I ask whether that number is plausible given the directory’s traffic and the client’s conversion rates elsewhere.

The arithmetic is unromantic. If a premium listing costs GBP 2,400 and labour is GBP 150 in year one, and the client’s qualified-lead-to-closed-won rate is 11% with an average contract value of GBP 14,000, then a single closed deal pays for the listing five times over. So breakeven is about 1 qualified lead per year if the close rate holds. That sounds easy until you remember that “qualified” has to mean qualified, not “filled out a form”.

What changed at day 120

We cut the portfolio from 47 listings to 9 over a six-week period. The cut was not just removing the listing; in several cases I asked the directory operator to delete the page entirely, not redirect it, because some of those pages were drawing low-quality traffic that polluted the analytics.

Lead volume up 34% with fewer listings

By day 120, directory-sourced qualified leads were up 34% against the same period the previous year. Demo bookings from directory referrals went from roughly 2.5 per month to 4.1 per month. The portfolio was now 9 listings instead of 47.

Two things were happening. First, the surviving listings (the original six plus three new ones I had carefully chosen) were each getting more attention from the in-house marketer; descriptions were updated, screenshots refreshed, integrations list maintained. Second, the new vertical directory placements were producing leads faster than I had expected, partly because the client was new to those audiences.

Sales team feedback on lead quality shift

Numbers are one thing. The sales team feedback was more telling. Within 90 days the BDR lead told me, without prompting, that directory leads had become “easier to qualify”. Buyers were arriving with more context, sometimes citing the specific comparison page they had found the client on.

This matters because lead quality compounds. A BDR who trusts a channel works the leads harder. A BDR who has been burned by 30 junk leads from a low-grade directory stops calling promptly when the next one arrives. I have watched this dynamic kill otherwise sound channels.

Two directories I almost cut but kept

Two of the original 20 candidates were borderline cases I kept after wobbling. I want to flag them because they show the danger of being too purist.

The first was a regional chamber listing that had sent only 14 sessions in 12 months but two of those had become opportunities worth a combined GBP 41,000. Tiny traffic, enormous yield. I kept it and refreshed the description.

The second was a niche directory with terrible UX, an outdated design, and a category page that looked abandoned. By every quality heuristic I would normally apply, it should have gone. But it ranked on page one for a long-tail phrase the buyers actually typed, and the listing was producing four demos a year quietly. I kept it on the principle that you do not interrupt something that is working, even when it offends your aesthetics.

Adjusting the playbook for different constraints

This walkthrough described a B2B SaaS with budget. The framework changes when constraints do, sometimes quite a lot. Below are the three variants I get asked about most.

Running this on a $500 monthly budget

If your total monthly marketing spend is $500 and directories need to come out of that, you cannot afford premium placements on the top vertical review sites. You also cannot afford to waste 40 minutes per submission on bulk directories that return nothing.

My approach here is aggressive concentration. Pick three to five directories: one curated general, one or two category-specific that match your industry, and one regional if you have a geographic edge. Spend the budget on getting each listing to look professional. Photography, a properly written description, every field filled, claim and verify ownership where possible.

The total directory count for a small business with this budget should rarely exceed 10. I would rather see 6 excellent listings than 30 mediocre ones; the time savings alone funds a better activity elsewhere.

Local service businesses versus SaaS

If you are a plumber, dentist, accountant, or any service business that serves a defined geography, the calculus inverts. Local directories, Google Business Profile equivalents, and chamber listings move from being a small slice to being the core. Vertical comparison sites become much less useful unless your industry has a strong national review platform (Checkatrade, Trustpilot for some categories, etc.).

For local service businesses, the number of directories that genuinely matter in any given market is usually somewhere between 8 and 20. Beyond that you are scraping diminishing returns hard. NAP consistency (name, address, phone) matters more than for SaaS because local search algorithms specifically check for it.

I would still audit before adding, but the audit will be quicker; there are fewer credible local directories than there are general business or SaaS directories, so the field self-narrows.

When a six-week timeline forces shortcuts

Occasionally a client comes in with a hard deadline: a funding round, a product launch, an event. They want directory presence in six weeks. Full audit and curate cycles need eight to ten weeks to do properly, so something has to give.

The shortcut I use: skip the full historical audit and run on a pure shortlist. Identify the top 5 to 8 directories in the client’s category by a quick combination of search visibility (does the category page rank?) and category fit. Submit only to those. Accept that you will revisit in 90 days and may find some were not worth the time.

This is worse than the proper process, but it is not catastrophic, provided you pick on category fit and editorial standards rather than on directory count.

Principles I now apply to every directory audit

After running this process on roughly 30 client engagements over the last four years, a few principles have proved themselves repeatedly. They are unglamorous. That is the point.

stateDiagram-v2
  [*] --> Screened : candidate identified
  Screened --> Rejected : fails 2+ of 3 criteria
  Screened --> Submitted : passes screening
  Submitted --> Active : listing goes live
  Active --> QuarterlyReview : 90 days elapsed
  QuarterlyReview --> HighPerformer : leads up >30%
  QuarterlyReview --> Borderline : leads flat or low
  QuarterlyReview --> Underperformer : zero leads or leads down >30%
  HighPerformer --> Active : refresh & keep
  Borderline --> Active : qualitative review, keep
  Borderline --> Rejected : qualitative review, cut
  Underperformer --> Rejected : cut & request deletion
  Rejected --> [*]

The three questions before any submission

Before I add any directory to a portfolio, I answer these:

1. Does this directory’s category page rank, organically, for a query that my buyer actually types? If not, the only way the listing gets seen is through internal directory navigation, which is rare.
2. Is the directory curated by a human, even loosely? Pure pay-and-list directories accumulate spam and get devalued by search engines. Human editorial standards are a leading indicator of durability.
3. If I were the ideal buyer for this product, would I find the listing useful? Not “would I tolerate it”, but useful. Could I compare options, check pricing context, see screenshots, read reviews?

Two out of three is sometimes acceptable. One out of three almost never is.

Reviewing the portfolio quarterly, not annually

Directory ecosystems change faster than people expect. A site that was authoritative two years ago may have been acquired, redesigned badly, or quietly stopped curating. A new entrant may have captured your category’s organic ranking. Annual reviews miss this; quarterly reviews catch it while it is still cheap to react.

The quarterly review I run is light: 90 minutes, pull updated traffic and lead numbers, flag any directory whose performance has moved by more than 30% in either direction, decide whether to investigate. It does not require a full re-audit each time.

When more directories actually does make sense

I have spent this article arguing for restraint. Let me close with the honest caveat: there are situations where higher directory counts are correct.

Multi-location businesses (a chain of clinics, a franchise) genuinely need broader directory presence because local search algorithms reward NAP consistency across many credible sources. A SaaS with multiple distinct product lines may need separate listings per product on category-specific sites; the count grows naturally with product surface area. Brands entering new geographies sometimes need to rebuild a regional directory presence from scratch.

In those cases the count goes up, but the principles do not change. Each individual listing still needs to clear the three questions. The growth is in the number of relevant categories you legitimately occupy, not in volume for its own sake (see Figure 1 for the triage logic that still applies).

If you take one thing from this walkthrough, take this: open your CRM tomorrow, pull directory-attributed opportunities for the last 18 months, sort descending, and look at the bottom of the list. The directories that produced zero are not neutral. They are costing you the labour to maintain them and, in some cases, the brand association of sitting next to whatever else they accept. Cut them this week. Then go and do the harder work of finding three more directories that match the shape of your six winners. That is the whole job.