Measurement and advertising traffic

Updated 6 October 2026. Contact support@techlatch.co.uk about this information or your data.

Optional website measurement

If you accept measurement, we keep a random browser identifier and first/latest campaign details in your browser for up to 30 days. A session changes after 30 minutes without activity. We remember your measurement choice for up to six months. We record pages viewed, time thresholds, scrolling, contact-link clicks and enquiry flow references. If delivery fails, up to 50 pending events are kept in this browser tab for up to 24 hours and retried while measurement remains accepted. Rejecting measurement clears that queue. We keep campaign identifiers, referrer hostname and an approximate device category. We do not put your name, email address, enquiry message or photos into the visit report.

When you enquire with measurement enabled, campaign details and a reference accompany the enquiry sent through our existing form service. The private report can receive your selected service and postcode district, such as KA7, to help us understand whether enquiries are relevant. We manually confirm receipt and record outcomes. A click or form-service return is not proof of a completed enquiry or sale.

Google Ads measurement, when accepted, records phone, email and WhatsApp link clicks and returns from the enquiry service using Google cookies. A random event reference helps prevent duplicate counting. WhatsApp clicks and form returns are observation-only actions, not confirmed messages, received enquiries or qualified leads. We do not currently upload qualified enquiries or job values to Google from this report. See Google's privacy information and FormSubmit's privacy information.

OpenAI advertising measurement

With your permission, a return from the enquiry service sends a measurement event to OpenAI. This includes a random event reference and, when available, the OpenAI ad attribution identifier from your arrival. We keep that attribution identifier in your browser for up to 30 days after an ad arrival. The browser request also exposes normal connection information to OpenAI. We do not use automatic customer-information matching or send your name, email address, telephone number, message or photos in this event. A form-service return is an observation, not confirmation that an enquiry reached our inbox. Test visits are excluded. See OpenAI's privacy information.

Your choice

You can reject optional measurement and still browse, call or enquire. Use Measurement preferences at the bottom of a website page to change your choice. Withdrawal stops new optional events and clears our attribution identifiers; it cannot undo data already processed by another service. Please contact us about existing records. Do not send us passwords or payment-card information.

Basic site traffic counts

Our server counts successful requests for public website pages, whether or not optional browser measurement is accepted. We keep hourly totals by page, broad referral category (including the tagged advertising platform), device category and likely-bot/test label for 30 days. These counters contain no IP address, browser identifier, full user-agent, query string or referrer URL. Repeated page requests are not deduplicated into people. Assets, private report activity and missing pages are excluded. Detailed browser journeys remain optional.

Separate ad-traffic review

Our server separately records successful landing requests containing a Google click ID to investigate repeated advertising traffic. This does not require browser tracking cookies. Records include request time, landing page, permitted campaign parameters, coarse browser/device information and a protected network code. A recent connection IP is encrypted and available only in our private report. It can represent a shared connection, proxy or Google service rather than an individual person.

Repeated clicks prompt manual review; they do not automatically block a visitor or identify a competitor. We do not perform automatic third-party IP location lookups. Network research is recorded separately with its source and limitations.

Server security and reliability diagnostics

We review existing HTTPS hosting access logs to investigate repeated requests, automated scanning and website faults. Our private report groups requests by a protected network reference, with daily counts, first and last request times, busiest-minute counts, recognised page paths, bot hints, error counts and whether an ad-click tag was present. Network references change weekly and can represent shared connections or proxies, not individual people. We do not copy raw IP addresses, full browser descriptions, referrer URLs or query-string values into this diagnostic dataset, or connect it to optional browser journeys. Other request paths are grouped as assets or other paths. No automatic blocking or advertising conversion upload results from this view.

The report also counts successful Google- and OpenAI-tagged page requests by day, page and device within those logs. These are aggregate counts, not identified journeys or proof of charged clicks. The diagnostic window is seven days. Opening or refreshing the private report reads the latest available hosting logs immediately. Background refreshes triggered by public website activity run no more than hourly. Hosting log-writing delays may still apply. Expired rows are omitted from the report and removed from the saved snapshot on the next successful refresh; stored data can remain while the system is idle or refreshes fail. Original hosting logs have separate retention arrangements.

Storage and retention

Detailed website events, network research notes and click records are retained for a 30-day reporting window. IP reveal expires seven days after the latest relevant arrival. Minimal enquiry outcomes are kept for 90 days; their detailed click/session attribution is removed after 30 days. Imported daily campaign totals and change notes are kept for up to one year; imported search/location breakdowns and report audit actions for 90 days. Data is kept outside the public website behind private sign-in.

Deletion runs when ad capture, event capture or the report is used. There is no independent daily cleanup schedule yet, so expired data may remain on disk while the system is idle; expired IPs cannot be revealed through the report. Hosting logs, enquiry emails, manually downloaded reports and backups have separate retention arrangements. Ask us about access, correction or deletion of information relating to you.