Skip to content

AI consultancy for small UK businesses

An AI consultant that builds what it recommends.

Most AI projects never cross the threshold: they demo well and change nothing. We find where AI would pay back first in your business, then build it: assistants and automations that answer enquiries, chase follow-ups and write the reports.

Enquiry assistantExample flow
  1. 09:14 · Enquiry in

    “We’re a 30-person firm. Quotes take us two days and we lose jobs to faster replies.”

  2. 09:14 · Assistant replies

    Thanks, Sam. Two quick questions so the right person calls you: how many quotes a week, and where does your pricing live?

  3. 09:16 · Qualified

    • 30 staff
    • ~40 quotes a week
    • Start this quarter
  4. 09:17 · Booked and logged

    Call booked for Thursday 10:30. CRM updated, owner notified.

Task platforms, bots and reporting automations in production for a media intelligence company
DailyTask platforms, bots and reporting automations in production for a media intelligence company
In marketing, in-house and independent
10 yrsIn marketing, in-house and independent
Brand website catalogue across the Prohibition Partners group
PortfolioBrand website catalogue across the Prohibition Partners group
From audit to build to reporting. No account managers, no handoffs
1 personFrom audit to build to reporting. No account managers, no handoffs

00The threshold

The threshold is the point where effort starts to change the result. Below it, you can work hard every day and stay exactly where you are.

A chatbot nobody uses. A pilot that never ships. Ad spend nobody has honestly measured. It all looks like progress and the numbers stay flat. We build for the point where they start to respond, and we measure whether they did.

01What we build

Systems that take work off your team.

Each one does a specific job, reliably, inside the tools you already use.

AI assistants and agents

Assistants that draft, answer and triage inside your email, chat and CRM. Built narrow on purpose: one job done reliably beats a chatbot that does everything badly.

Try a working example →

Lead handling and follow-up

Every enquiry answered in minutes, qualified against your criteria, booked in and chased until there’s an answer.

The AI receptionist →

Workflow automation

The jobs that quietly eat the week, rebuilt to run without anyone: chasing updates, moving information between tools, keeping trackers current.

Reporting pipelines

Analytics, spreadsheets and channel data pulled into a report that writes itself on schedule and gets signed off by a human.

Custom internal tools

Dashboards and utilities for the workflows off-the-shelf software doesn’t fit.

02Around the system

A system needs something to work on.

An assistant can only answer the enquiries that arrive. So we also run the marketing that brings them in and build the website that turns visitors into conversations, measured from day one.

Marketing

Search and content
Built around what your customers actually search for, including how they ask AI assistants.
Paid media
Google and Meta, with tracking set up properly before a pound is spent.
Email and retention
Usually the cheapest revenue a business is ignoring.
Measurement
A monthly report you can read in five minutes, assembled by pipelines and signed off by a human.

Web and brand

Design and build
Fast, distinctive sites designed around the enquiry. No templates, no page builders.
Brand and positioning
A point of view, a voice and a look that doesn't blend in. Small, sharp systems.
Conversion
Landing pages, forms and booking flows tested against real behaviour.

03How it works

Audit first. Then build the thing that pays back.

  1. 01

    Audit

    Before anything gets built, we measure where you are: where the team's hours go, how enquiries get handled, what your marketing spend returns. You get a ranked list of what would pay back first. £1,500, credited in full against the build.

  2. 02

    Build

    We build the one system at the top of that list, inside the tools you already use, and put it into real use within weeks. Narrow first: it has to earn its place before it grows.

  3. 03

    Run and improve

    We watch the numbers, report in plain English and adjust. What works gets more. What stays below threshold gets cut, even if we built it.

Free · two minutes

What are missed enquiries costing you?

Your own numbers in, a plain answer out: the work at risk each year, what an assistant would cost, and whether it would pay for itself.

Run the calculator

04Selected work

The record, with the role stated plainly.

Health media & events

Marketing a science-led national title

Two stints growing the UK's national medical cannabis title — search, site, analytics and CRM — in a sector where the big paid channels are mostly closed to you. The constraint is the story.

Cannabis Health · In-house marketing

In progress

B2B services

A B2B engagement, reported live

A live engagement, reported as it happens: audit first, then the work, then honest numbers. The unglamorous early stage included.

Facilities management client · Above Threshold client

Clinical exercise physiology

The founder's own lab

Above Threshold's method, tested on our own business first: a clinical exercise physiology practice built from zero — positioning, brand, website and content, with every claim on the site cited to evidence.

RAD Exercise Physiology · Founder — brand, site & marketing

Free · two minutes

Where does your business sit against the threshold?

A handful of numbers in, a plain readout out: what a customer costs you, how fast they pay back, and what manual admin is costing.

Run the scorecard

05Questions

Asked on most first calls.

What does an AI system actually look like?

Usually something unglamorous. An assistant that answers and qualifies enquiries. A follow-up sequence that never forgets. A report that assembles itself. It lives inside the tools your team already uses, not in a new app they have to learn.

What happens to our data?

Data access is scoped in the audit, before anything is built: what the system can read, where it runs and who sees the output. You approve that in writing before go-live, and nothing gets access it doesn't need.

Who owns what you build?

You do, outright. There is no per-seat pricing and no licence back to us. The running costs of the underlying services are billed to you directly, so you can see exactly what the system costs to run. The exception is an enquiry assistant on the yearly price, where running costs are included up to an allowance agreed at the audit.

What does it cost?

The audit is £1,500, credited in full against the build. An assistant is priced at 10% of the yearly salary of the person who handles your enquiries, or would. That is per year, with the build, running and support included. The minimum is £2,400 a year. Other work runs as a fixed project or a monthly programme, scoped from the audit rather than from a rate card.

Is our business big enough for this?

There needs to be enough repeated work for a system to pay back: a team, a steady flow of enquiries, hours going on admin every week. If it wouldn't pay back for you, we'll say so on the first call.

Do you only do AI?

No. A system can only work on the enquiries that reach it, so we also run the marketing that brings them in and build the website that converts them. Plenty of clients start there.

Next step

Find out what's worth automating first.

A 30-minute call. We look at where your team’s hours go and what your marketing returns, then tell you honestly whether an AI system would pay back. No deck, no pressure.