Search, AI visibility and enquiry systems, UK-wide

Get found. Then stop losing the enquiries you get.

We build search and AI visibility systems for B2B and premium e-commerce. And the system that catches what comes in.

Free, no account access needed, and yours to keep either way.

Illustrative
"Who are the best suppliers for this in the UK?"

Based on the sources I can see, these are worth looking at:

  • 1 A competitor
  • 2 A competitor
  • 3 A competitor
  • ? You, if the sources say so

An illustration of the pattern, not a real answer about a real company.

01 / What changed

No page one inside an answer

Search gave your buyer ten links. An assistant gives them a paragraph naming two or three companies. Everyone else is absent.

The shortlist forms without you

That conversation decides who gets contacted. You never see the ones you lost.

Crawlers may not read your site

The crawlers behind ChatGPT run no JavaScript. Firewall bot rules block them more often than robots.txt.

One assistant tells you nothing

ChatGPT leans on Bing. Perplexity fetches live. Being in AI Overviews says little about the rest.

02 / Enquiry systems

Getting found is only half of it

An enquiry that lands at eleven at night and sits in an inbox is a lost customer. The system here runs on this page.

Catch the enquiry

One capture endpoint behind the form. No plugin, no forwarding rule, nobody watching an inbox.

  • Captured server-side
  • Saved once a valid email is entered
  • Abandoned enquiries kept

Filter the junk

Bots are scored and quarantined, not deleted, because a wrongly flagged enquiry is a lost customer too.

  • Time on form, gibberish and links
  • Quarantined, never silently binned
  • Every flag shows its reason

Know immediately

An email the moment a real enquiry lands, reply-to already set to the sender.

  • Instant alert on real enquiries
  • Silent on spam and duplicates

Own the data

Your own database and dashboard. Status pipeline, notes, export, and deletion that actually deletes.

  • A database in your name
  • CSV export any time
  • UK GDPR retention built in
03 / How an answer is built

Where you are either included or absent

One question becomes several. Each is answered from sources the engine already trusts. You are either in that set or you are not.

BUYER ASKS "who should we buy from?" who supplies this? who is credible? how do they compare? SOURCES USED THE ANSWER A competitor A competitor You?

Simplified. The retrieval step is where our work lands, because that is the only part you can influence.

04 / What we build

Seven parts, access first

Most of this is fundamental SEO done properly. The different part is the plumbing and the third party sources.

Access and rendering

If crawlers cannot read the page, nothing else matters. Fastest to fix, easiest to prove.

  • Robots and firewall bot rules
  • Rendered on the server
  • Indexed in Bing as well as Google

Entity foundation

A machine works out who you are before it names you. Ambiguity gets you swapped for a clearer competitor.

  • Schema that describes the business
  • Links to LinkedIn and Companies House
  • One name and category everywhere

Answer-ready content

Content gets lifted out of context and quoted. Direct answer first, evidence underneath.

  • Sections that stand alone
  • Comparison and pricing pages
  • Claims backed by dated sources

Source footprint

Much of what an assistant says about you comes from somewhere else.

  • Review platforms, trade press, directories
  • Genuine participation, never coordinated posting
  • Video, where the transcript is the work

Product data for e-commerce

An assistant reads your feed long before it reads your art direction.

  • Every attribute filled in
  • Product schema matching the feed
  • Reviews structured to be quoted

Measurement you can check

Trends across repeated runs, never one screenshot.

  • Your buyers' real questions, on a schedule
  • Who got cited instead of you
  • AI crawler activity from your logs

Enquiry capture

The system that catches what the visibility work brings in.

  • Capture endpoint behind the form
  • Spam scored and quarantined
  • Your own database and dashboard
05 / Real work

Systems we have shipped

Two consumer businesses. The counts are form throughput, not business outcomes. We have not measured revenue or return on spend, and claim none.

Hospitality, Newcastle

Jungli

wearejungli.co.uk ↗

Indian restaurant, cafe, bar and events venue in Newcastle city centre. We built the site and the forms behind it.

  • 202 email signups collected
  • 4 contact enquiries received
  • Honeypot spam trap on both forms
Read the case study →
Clinic, London

Amor Therapies

amortherapies.co.uk ↗

Recovery clinic inside State of Change Clinic on Bethnal Green Road. We built the site and every route into a booking.

  • Booking requests with date and time
  • Booksy, WhatsApp and profile links
  • Honeypot spam trap on both forms
Read the case study →
06 / The method

Published, not proprietary

You can repeat every check, and so can your developer. Everything is in your name from day one.

01

Baseline

We date where you stand before touching anything. You get the raw captures, not a summary.

02

Fix the plumbing

Access, rendering, speed, schema, feed. Cheapest first, because an unreadable site will not be cited.

03

Build the evidence

Answer-first content, the commercial pages you have avoided, then the off-site sources. The slow part.

04

Re-run and report

Same prompts, same schedule, read as a trend. One page a month: what moved, what did not.

07 / Fit

Who this is for

Founder-led and deliberately small. We would rather turn work away than take money for something that will not pay you back.

A good fit

B2B with a considered sale, or premium e-commerce with a high order value. Someone has already asked whether an assistant names you.

A poor fit

If you need leads this month, wrong spend. If the problem is pricing or follow-up, visibility will not fix it.

What you need in place

Access to your site and analytics, someone who can approve copy, and willingness to publish comparison and pricing pages.

08 / Straight talk

What we will not promise

This category sells certainty that does not exist. Here is the honest edge, before you ask.

No rankings, because there are none

There is no position one in an answer. You are named or you are not. Anyone selling rank one is lying.

No clean attribution

Search Console reports AI impressions, not clicks. A mention with no click can still win the deal and leaves no trace.

No repeatable single answer

Same question, different day, different sources. We report a distribution and treat any single screenshot as an anecdote.

No promises about training data

Nobody outside the labs can put you in a model's training data or prove you are in it. We set the policy, not the outcome.

09 / How we work together

Quoted per project. No lock-in.

No rate card, because no two of these jobs need the same work. Quoted and agreed in writing before anything starts.

Visibility audit

The growth report with the evidence attached. Prompt baseline, server logs, render and access testing, entity and feed review.

Build or rebuild

The site, or the broken parts of it. Fast, server-rendered, structured to be extracted. Fixed scope, fixed quote.

Enquiry system build

Capture behind your forms, spam quarantine, instant notification, and a dashboard on a database in your name.

Ongoing visibility

Continuous work across plumbing, content and sources, with one honest report a month. Cancel on 30 days and keep everything.

What you own

Accounts, feeds, schema, content and documentation, in your name from day one. If you leave, nothing switches off.

Free, and yours to keep

Find out whether the answer names you

We audit your presence the way an answer engine sees it, grade five areas out of ten, and walk you through it on a free 20 minute call. Yours to keep either way.

  • AI answer presence. Are you named, and who is named instead.
  • Machine readability. Can the crawlers reach and parse your site.
  • Entity clarity. Can a machine tell exactly who you are.
  • Source footprint. The places engines already cite in your category.
  • Answer-ready content. Do the pages buyers ask for exist, extractably.

Get your free growth report

Tell us where to look. We reply within a couple of days.

So a genuine enquiry isn't lost, we save what you've typed once you've entered your email address, even if you don't press the button. We use it only to follow up on this request, never for a mailing list, and abandoned enquiries are deleted after 90 days. Privacy policy.

Something went wrong sending your details. Please try again in a minute, or email nyle@kgsdigitalltd.co.uk.

Got it. Your report's on its way.

We'll email you within a couple of days to book your walkthrough.

Questions we get asked a lot

A grade out of ten across five areas with the evidence attached, a prioritised fix list, and a 20 minute call. No account access needed.

Most of it is good fundamental SEO, and any agency that will not say so is selling snake oil. What differs: ChatGPT's crawlers run no JavaScript, and it leans on Bing.

Not an off-the-shelf one. We build the capture layer and a dashboard on a database in your name. If you already run a CRM we scope an integration into it. The full system runs on this page.

No, and nobody can. There is no ranking to buy. We commit to what is verifiable: crawlers can read your site, your entity is unambiguous, your facts are right in the sources engines cite.

Fair question. Those are consumer businesses, here as evidence of build quality, not sector credentials. Take the free report, hand it to your technical person, and judge it before spending anything.

Access and rendering fixes are verifiable the day they ship. Engines that fetch live can reflect changes within weeks. Entity and source work compounds over months.

Quoted individually once we have looked at your site, with scope written down first. Monthly work runs on 30 days notice with no minimum term.