City directories: source-listed, not individually verified

A city directory contains up to 200 businesses grouped by category, plus a private reserve of replacement records that is not included in the advertised count. Where fewer businesses in a city pass the checks, its directory is smaller, never padded with unchecked rows, and priced in proportion. Rows retain the business name, category, address, phone as listed, available website or social link, and the source and original snapshot date. A preparation date never replaces the source date.

Before a directory goes on sale, every delivered field must have compatible source permissions. We check required fields, phone format, duplicates and current exclusions. An independent review covers at least 24 records per city, including the reserve. At least 95 percent of the sample must be confirmed by a second public source, with zero severe mismatches. Sample review does not mean every business was individually rechecked.

In UK cities, a row that lists a website is included only if that website still showed the listed postcode, and no notice of closure, when we opened it, and the row carries the date of that check. In Nigerian cities websites rarely print a postcode, so a row that lists a website is included only if that website showed the listed phone number and named the city, with no notice of closure. Where a Nigerian city has too few such businesses, the rest are supplied as listed, and a sample row also passes review when the business's own page is live under the same name with a post in the last 12 months. The 95 percent pass mark and the zero severe mismatches rule are the same in both countries, and each city page states which of its sample rows passed in which way.

Directories do not promise current phone reachability, an active business today, a missing website, any other business problem, or buyer intent. Review the dated source and current business details before using a record. A city directory is offered for sale only after these checks pass for that city, and its page shows the review result.

Every row comes from OpenStreetMap, under the Open Database License, or from the Overture Maps Foundation Places theme, under CDLA Permissive 2.0, Apache 2.0 or CC0 1.0 depending on the upstream dataset. Records that came from a map search, a third-party directory or unattributed manual research are never included. Rows outside the city's postal area, rows with a phone number from another country, public bodies and places of worship are removed before review. Because part of each directory derives from OpenStreetMap, the file is supplied under the Open Database License: buyers may use and adapt it commercially, and must keep the attribution notice that comes with it. A business can ask to be removed at any time through the privacy page, and a removed business is never delivered again.

Researched problem packs

The research stages and product-specific standards below apply to researched problem packs, not city directories. Their existing evidence, freshness, independent audit and release requirements are unchanged.

Admission standard

A released record must identify a current local business, include a usable public contact route, keep its public source licence, and pass the method for its named product. Every row stores a check date, methodology version, confidence score, and verification evidence.

Research stages

Candidate discovery. We collect public business candidates from attributed map and web sources.
Identity match. We compare name, location, category, address, phone, and public profiles to avoid mixing businesses.
Website audit. We inspect source website fields, current domain signals, and exact name, address, and phone searches.
Contact check. We retain only records with a usable public phone or social route.
Exclusions. We reject matching official websites, inactive businesses, obvious duplicates, chains excluded by the release rules, and unverifiable identities.

Verification labels

Human verified means an operator performed the exact business web and contact review. Research verified means the record passed the documented current-source, domain, and exact-search audit used for the controlled city release.

No Website Prospects

The business must have a usable public contact route and no matching working standalone website found during the recorded audit. Social profiles and marketplace pages do not count as standalone sites. Search indexes and public sources can be incomplete, so this is a dated finding rather than proof that a site cannot exist.

Website Upgrade Opportunities

A qualifying business already has a website, and the pack records at least one dated finding about that website. A released pack is a frozen, ranked list for one city. It is built in four steps.

The business. Rows come from the Overture Maps Foundation Places theme. A business is kept only if its own website still showed its listed postcode (United Kingdom) or its listed phone number and named the city (Nigeria), with no closure notice. Chains, schools, charities, banks and public bodies are left out.

The findings. The homepage is read twice by two different clients: once as plain HTML with scripts switched off, and once in a real browser with scripts running. A finding is recorded only when both readings agree. The findings are: the HTTPS address shows a certificate error or does not answer; the HTTP address loads without sending visitors to HTTPS; no mobile viewport setting; a homepage link that leads to a missing page on the same site, confirmed on two requests; no meta description; no main heading (H1); and no LocalBusiness structured data. A page that blocks the check, redirects to another site, shows a placeholder or cannot be read clearly yields no finding, and that business is left out. Each site's robots.txt is honoured. Nothing is submitted and no page content is kept, only counts, types, status codes and the address of a broken link.

The review. Before release, a sample of 24 rows is checked two ways: a reviewer looks each business up against a second public source or its own website, and a separate checker re-tests each row's main finding with a different tool. A pack is released only when every sample row is checked, at least 95 percent pass both checks, and no severe mismatch is found. The review is agent-assisted and covers a sample, not every business.

Shelf life and remedy. Websites change, so a release is sold for 30 days from its checks and then closes by itself. Each row carries its check date. A finding is a dated observation about a public web page, not proof that a business wants a service, and a missing heading or description is not described as proof that a rebuild is required. If you cannot reproduce a finding, or a number is dead or a business has closed, tell us within 14 days and we replace the row from the pack's private reserve while reserve rows remain.

Local Growth Opportunities

A qualifying business has a website and public contact route, plus at least two distinct findings with confidence of 80 or more. One must be a primary signal. Version 1 signals cover acquisition-page indexing blocks, material business-information conflicts, mobile information loss, broken or missing conversion routes, sitewide crawl blocks, absent local relevance, generic home titles, and structured business data gaps.

Paid Ads Opportunities

A qualifying business has a matching active website, a public contact route, and at least two distinct campaign-readiness findings. One must be primary. Version 1 checks supported analytics measurement markers, public conversion routes, controlled mobile page performance, and documented title or primary-heading gaps. The result does not claim that the business currently buys advertising.

Email and CRM Opportunities

A qualifying business has a repeat-customer category or a public booking, order, quote, or manual enquiry route, plus at least one additional public signal. Version 1 checks newsletter capture, public business email, conversion routes, and whether the licensed category supports repeat visits or purchases. It does not claim access to the business's private customer systems.

Recruitment Opportunities

A qualifying business must have at least one current public job opening plus an active public application route or a dedicated public recruitment contact. Version 1 can also record multiple open roles and a current careers page. General employer branding or old careers copy without a named current role does not qualify.

IT Modernisation Opportunities

A qualifying business has at least two documented public configuration signals, with one primary. Version 1 uses passive public DNS checks for published DMARC, SPF, and MTA-STS records and documented mail policy conflicts. These are public configuration observations, not vulnerability, breach, compromise, or private-system claims.

E-commerce Growth Opportunities

A qualifying business must have an active public storefront plus at least one additional storefront finding. Version 1 checks supported product structured data, controlled mobile performance, public checkout entry, and product discovery routes. No purchase is attempted and no private checkout or customer data is accessed.

Hospitality Target Accounts

A qualifying business has a licensed public hospitality category and at least one additional public signal. Version 1 distinguishes food and beverage, accommodation, and event or group services, then checks for a public location and contact route or a booking, order, or enquiry route. The list identifies a documented target-account market, not current procurement intent.

Freshness and city certification

Premium opportunity findings use a 30-day freshness window. Before publication, one methodology version must cover the release, at least 10 percent and never fewer than 24 records receive an independent recheck, agreement must reach 95 percent, and severe mismatches must be zero. Manual certification can approve, reject, retry, or quarantine a candidate. A public sample never identifies the audited business.

Each city release is standardised at 200 licensable organisations. Additional current records are retained as a replacement reserve and are not included in the advertised release size.

Passive checks and evidence disputes

Research uses passive checks of public pages and public records. It does not log in, submit forms, bypass access controls, scan private systems, or interact with payment actions. A buyer or listed business may raise an evidence dispute. We preserve the order-time finding, review current evidence separately, and correct or suppress current inventory where appropriate.

Important limitation

An opportunity is not buyer intent, consent, or proof of budget, urgency, ranking impact, traffic value, revenue, or conversion value. Businesses and websites can change after review. The source, evidence, and checked-at fields let a buyer re-check before outreach.

Sources and attribution

Research may include OpenStreetMap contributors under ODbL and Overture Maps source data with its required source-specific attribution. Public search evidence is used for verification, not as a claim of source endorsement.

Corrections

A listed business or representative can request access, correction, removal, or objection through the form on the privacy page. We review identity and evidence before changing or suppressing a record.