Four colleagues gathered at one desk in an open-plan office, watching a seated colleague work through a spreadsheet with an AI assistant panel open beside it

AI adoption consultants for UK businesses

AI adoption consulting is the work of getting staff to use AI assistants in their real jobs, rather than holding licences that nobody opens after the first week. It sits between buying the tools and building systems with them. The AI Consultancy delivers it for UK SMEs, mid-market firms and enterprise teams, and it is vendor-neutral: we roll out Claude, ChatGPT and Microsoft Copilot on their merits for each team, and we take no referral, reseller or deployment income from any of them. Where the answer is a Claude workspace for a firm of up to about a hundred staff, the fixed-price route is our Claude Enablement install, and we will say so.

What AI adoption consulting covers

AI adoption consulting covers four stages: readiness, rollout, change management and usage measurement. Readiness decides which teams, tasks, data and policies are in place before anything is bought. Rollout puts the tool on a company-owned tenancy with the right access, connectors and role prompts, and launches it in stages. Change management assigns the roles, writes the policy and runs the first 30 days. Usage measurement reads what the plan reports and turns it into a keep, fix or retire list.

What it is not: it is not a software build, which is our AI implementation service; it is not licence resale, because we do not resell; and it is not a training course, which is Learn AI's work. The output is a working rollout that people use, with the evidence to show it.

The word enablement is often used for the same work, and we use it too: enablement is what the organisation puts in place, adoption is the usage that results. We plan for one and measure the other.

Why AI rollouts stall, and what the adoption gap costs

AI rollouts stall, in our experience, for reasons that have nothing to do with the model. The licences were bought before anyone chose the tasks; the tool could not see the company's data, so every use meant copying and pasting; there was no written policy, so the most careful staff stayed away; and nobody owned the rollout after the launch email. The result is a familiar pattern: a burst of use in week one, a handful of enthusiasts by week four, and a renewal decision made on anecdote.

The cost is easy to state. A seat that is paid for and not used is a direct loss; a seat that is used badly, on tasks the tool is poor at or with data it should not see, is a risk. Both are avoidable with the same preparation. Our guides to why employees resist new AI tools and to the first decisions in a ChatGPT or Claude rollout set out the pattern in detail, and the Office for National Statistics' business survey data on UK business AI adoption gives the latest figures on scale and depth of use.

Readiness: what we check before any rollout

AI readiness, for adoption purposes, is the answer to four questions: which teams and tasks will the tool be used for, where does the data for those tasks live and who may see it, which policies and approvals have to exist before launch, and who will own the rollout once we leave. We answer them in a short assessment before any licences are bought, because the answers change which tool, which plan and how many seats you need.

  • Task map. Three to five workflows per team where an assistant is likely to save time or improve quality, with a baseline measured before launch.
  • Data and access. Which systems hold the data (Microsoft 365, Google Workspace, the CRM, the finance system, document management), what a connector would expose, and an exclusions list.
  • Policy and approvals. Whether an acceptable use policy exists, whether the rollout needs a DPIA screening, and which plan tier the governance requirement points to.
  • Ownership. A named owner, a champion per team and an escalation route for the questions that arrive in week two.

The formal route for this stage is our AI Readiness Assessment, from GBP 1,000, which ends in a written report you keep whatever you decide next. Our screening guide on whether a Claude or Copilot rollout needs a DPIA covers the data protection question.

Rolling out Claude, ChatGPT and Microsoft Copilot

A rollout puts the chosen tool on a company-owned tenancy, connects it to the data each team needs on the narrowest useful scopes, gives each role a starting configuration, and launches in stages rather than to everyone at once. The sequence is the same for Claude, ChatGPT and Copilot; the plan tiers, admin controls and connector options differ, and the choice between them follows the task map from the readiness stage. Many organisations end up running more than one.

Claude

Claude Team or Enterprise tenancy, Claude Cowork as the default surface for non-technical seats, connectors on least-privilege scopes, shared skills and role prompts. For firms of roughly five to one hundred staff this is the fixed-price Claude Enablement install; larger rollouts are scoped under this service.

ChatGPT

ChatGPT Business or Enterprise workspace, admin and data controls set against the tier you hold, shared workspace configuration per team, and a role-based first-week plan. Our ChatGPT implementation service covers the API and integration side where a build is needed.

Microsoft 365 Copilot

Tenant settings, SharePoint and OneDrive permission hygiene before launch (Copilot works within the permissions each user already holds), connector governance, and task-based training in Word, Outlook, Excel and Teams. See our note on the Copilot tenant setting to check.

Licences are bought by you from the vendor and are not part of our fee. We do not resell, and we take no referral, reseller or deployment income from Anthropic, OpenAI, Microsoft or anyone else; the detail is on our independence and vendor relationships page.

Change management for AI: roles, policy and the first 30 days

Change management for an AI rollout means three things: named roles (an owner, a champion per team, and someone who answers questions in the first month), a short written acceptable use policy that staff have read, and a 30-day plan that moves each role from a first guided task to unaided weekly use. Without these three, usage fades after launch; with them, it holds and grows.

The 30 days are structured. Week one is guided: each person completes two or three of their own tasks with the tool alongside a champion. Weeks two and three widen the task list and collect the questions and failures. Week four is the review: which tasks stuck, which were dropped and why, which connectors or prompts need changing, and what the usage data says. Our 30-day Claude adoption plan is the written version of this for Claude; the shape is the same for ChatGPT and Copilot.

The policy work draws on our acceptable use policy template for UK SMEs, and the training is pitched against our five levels of AI literacy so a finance team and a sales team are not given the same session.

Measuring adoption and usage: what your plan actually reports

Adoption measurement depends on the plan you hold, and we will not promise a metric your tier does not provide. On Claude Enterprise, Owners have audit logs and the Compliance API. On Claude Team, Owners and Primary Owners get the admin usage analytics dashboard (weekly active members, adoption by product, connector read and write counts, skills usage and a spend export), plus the option to stream Claude Cowork events to a SIEM through OpenTelemetry. On the individual Pro and Max plans there is nothing organisation-wide, which is one reason we install on a company-owned tenancy rather than on personal accounts.

For ChatGPT and Copilot we read the admin reporting your tier includes, confirm at scoping what it does and does not show, and put that in writing before launch. Usage dashboards report activity; they do not say whether a task got done. So we add two measures the dashboards cannot give: a weekly sample of real tasks completed with the tool, reviewed by the champion, and the before-and-after time on the three workflows chosen at readiness.

The output at day 30 is a keep, fix or retire list per team and a decision on seats before the renewal date, not after it. Our guide to measuring AI ROI in the first 90 days sets out the longer measurement plan.

Training: what we do in a rollout, and where Learn AI fits

Training inside an adoption engagement is task-based enablement: a role prompt, two or three worked examples from each person's real work, and a first-week plan. It is inside the engagement and it is not a course. Structured programmes (executive briefings, team workshops and multi-week enablement for professional and regulated teams) are delivered by Learn AI, our training brand, and we refer you there rather than sell a course from this page.

Where a ChatGPT rollout needs a half-day or full-day in-house workshop with a UK GDPR module, that is available through our ChatGPT training for teams service, from GBP 1,500.

AI adoption, Claude Enablement or AI for SMEs: which page do you need?

Three of our services touch adoption, and they are deliberately different in scope. This page is the vendor-neutral adoption and enablement service for any size of organisation. Claude Enablement is the fixed-price, on-site Claude workspace install for firms of roughly five to one hundred staff, and the right choice when you already know the answer is Claude. AI for SMEs is the budget-conscious entry point for smaller businesses that want quick wins across rollouts, training and automations in one engagement.

If adoption needs a senior owner inside the business for six months or more, rather than a project, that is the Fractional Chief AI Officer retainer, from GBP 3,000 a month at the Advisor tier.

What does AI adoption consulting cost?

There is no separate adoption price list. The work is sold through routes that are already priced and published, so you can compare them before a call:

RouteWhen it fitsPublished price (ex VAT)
AI Readiness AssessmentYou do not yet know which teams, tools or plan to start with.From GBP 1,000; typically GBP 3,000 to GBP 8,000 for a full assessment.
Claude EnablementThe answer is a Claude workspace for a firm of up to about a hundred staff.Team Install from GBP 4,500; Business Install from GBP 7,500; 30-day adoption window included.
Fractional Chief AI OfficerAdoption needs a senior owner inside the business for six months or more.From GBP 3,000 a month at the Advisor tier.

Multi-tool or multi-site rollouts outside these packages are scoped in a written proposal after a free 20-minute call. Licences are bought by you from the vendor and are not part of our fee.

Related services and guides

Related services

Guides

Frequently asked questions

What is an AI adoption consultant?+
An AI adoption consultant helps an organisation move from owning AI tools to using them in everyday work. The job sits between buying licences and building systems: deciding which teams and tasks to start with, configuring the workspace and its connections to company data, setting the policy and roles that make usage safe, training people in their own tasks, and reading the usage data afterwards to see what stuck. The AI Consultancy delivers this for UK SMEs, mid-market firms and enterprise teams across Claude, ChatGPT and Microsoft Copilot, and does not resell any of them.
What is the difference between AI adoption and AI enablement?+
In practice the two terms describe the same work from two ends. Enablement is what the organisation does to make use possible: the tenancy, the access controls, the connectors, the role prompts, the policy and the training. Adoption is the result: the share of licensed staff who use the tool in real tasks each week, and the time or quality that changes because they do. We plan for enablement and measure adoption. Our Claude-specific version of this work is the fixed-price Claude Enablement install.
How is this different from Claude Enablement?+
Claude Enablement is a fixed-price, on-site install of a Claude workspace for firms of roughly five to one hundred staff, with two packages (Team Install from GBP 4,500 and Business Install from GBP 7,500, ex VAT) and a 30-day adoption window inside the price. AI adoption consulting is the vendor-neutral service around it: it covers firms choosing between Claude, ChatGPT and Copilot, firms that already hold licences for more than one, and rollouts larger than the install packages. Where the answer is Claude for a firm of that size, we recommend the Claude Enablement package because it is fixed-price and already includes the 30-day adoption window.
Which should we roll out: Claude, ChatGPT or Microsoft Copilot?+
It depends on the work each team does, the systems the data lives in and the plan tier you can justify, and we do not decide it in advance. The readiness stage maps each team's tasks to the tools and plans available, tests the shortlist on real tasks, and writes the recommendation down before any licences are bought. Many organisations end up with more than one. We take no referral, reseller or deployment income from any vendor, so the recommendation is written against your tasks, not a margin.
How do you measure AI adoption?+
With whatever your plan actually provides, confirmed at scoping rather than promised. On Claude Enterprise, Owners have audit logs and the Compliance API. On Claude Team, Owners and Primary Owners get the admin usage analytics dashboard (weekly active members, adoption by product, connector read and write counts, skills usage and a spend export) plus the option to stream Claude Cowork events to a SIEM through OpenTelemetry. On the individual Pro and Max plans there is nothing organisation-wide, which is one reason we install on a company-owned tenancy. For ChatGPT and Copilot we read the admin reporting your tier includes and say in writing what it does and does not show. Alongside the vendor data we track two things the dashboards cannot: a weekly sample of real tasks completed with the tool, and the before-and-after time on the three workflows chosen at readiness.
Do you provide the staff training?+
Structured training programmes, executive briefings and multi-week courses are delivered by Learn AI, our training brand for professional and regulated teams, and we refer you there rather than selling a course from this page. Inside a rollout we do the task-based enablement that any install includes: a role prompt, worked examples from each person's real work and a first-week plan per role. A half-day or full-day in-house workshop as part of a ChatGPT rollout is available through our existing ChatGPT training for teams service, from GBP 1,500.
What does AI adoption consulting cost?+
There is no separate adoption price list. The work is sold through routes that are already priced: an AI Readiness Assessment from GBP 1,000 (typically GBP 3,000 to GBP 8,000 for a full assessment) when the question is where to start; a Claude Enablement install from GBP 4,500 or GBP 7,500 when the answer is a Claude workspace for a firm of up to about a hundred staff; and a Fractional Chief AI Officer retainer from GBP 3,000 a month when adoption needs an owner inside the business for six months or more. All figures exclude VAT. Licences are bought by you from the vendor and are not part of our fee.
Why do AI rollouts stall after the first month?+
Four reasons come up repeatedly in the rollouts we see. Nobody chose the tasks the tool was for, so people tried it once on something it was bad at and stopped. The tool could not see the company's data, so every use meant copying and pasting. There was no policy, so cautious staff, often the ones with the most valuable work, did not use it at all. And nobody owned the rollout after launch, so the questions people had in week two went unanswered. The readiness stage is designed to remove the first three before launch and the change-management stage to assign the fourth.
Engagement models and pricing

Readiness Sprint from GBP 3,500 · Claude Enablement from GBP 4,500 · Discovery and Pilot from GBP 15,000 · Build and Embed from GBP 40,000 · Day rate GBP 950 to GBP 1,500. All prices exclude VAT.

Tell us which tools you hold and who is not using them

A free 20-minute call: which teams, which licences, what the data looks like and who owns the rollout. You get a written recommendation on the route (readiness, a Claude install, or a scoped rollout) and nothing is sold on the call.

Or call +44 2033 550 558