BCG’s AI at Work study confirms what aibl’s research already found
I told you so. It's petty I know, but it feels good when a monster like BCG reinforces a message from your own...
Read moreBy Richard Breeden, Founder & CEO, aibl Media
Most organisations building an AI enablement programme hit the same wall, usually somewhere between the pilot and the board meeting where ROI gets discussed. The tools are in. People are using them. And yet the performance improvement that was supposed to be obvious by now is either not showing up in the numbers or is showing up in ways nobody can attribute to anything specific.
AI enablement is the work that closes that gap.
It is not a product, a software platform, or a specific methodology. AI enablement is the discipline of building the internal capability: the skills, the processes, the governance, and the measurement infrastructure, that turns AI deployment into provable, sustainable business performance.
Deploying AI is the easy part. 85 per cent of UK mid-market organisations already have AI embedded in at least one operational area, according to aibl’s State of UK AI Adoption Survey 2026 (n=755, January–March 2026). The harder part, which only 49.5 per cent have managed, is being able to show the board what it’s producing.
AI enablement is how the other half gets there.
“AI enablement” is a term that the market is starting to use, but nobody has defined it with much precision yet. That matters because vague terms invite vague programmes, and vague programmes produce exactly the kind of AI activity that looks busy and measures badly.
At aibl, we define AI enablement as covering eight distinct service categories:
These eight categories are how the AI Enablement Directory , aibl’s curated list of vetted UK implementation partners, is organised. They cover the full range of support a UK mid-market organisation might need to move from AI activity to AI outcomes.
The categories are not interchangeable. An organisation that needs help defining its AI strategy has a different problem to one that has a strategy but cannot operationalise governance. An organisation buying AI training for its workforce has a different requirement to one that needs to build agentic workflows into its operations. Getting the category right matters because buying the wrong kind of support is one of the most consistent ways organisations stall.
It is not the same as AI adoption. Adoption refers to the deployment of AI tools, getting them into the organisation and in front of the people who will use them. Adoption is relatively straightforward and, as the survey data shows, largely done: most UK mid-market organisations have adopted AI in some form.
It is not AI transformation either, though the two are related. AI transformation describes the organisation-wide change in how a business operates when AI is embedded at scale. That is the destination. AI enablement is the capability-building work that makes it possible to get there and stay there.
And it is not AI training alone, though training is one of the eight categories. AI enablement that focuses only on skills development without addressing governance, measurement, and implementation support tends to produce upskilled individuals inside unchanged processes. The performance improvement does not follow.
The most direct evidence on what produces measurable ROI from AI comes from comparing capability-building approaches. The aibl survey asked organisations how they build AI capability, and the difference in outcomes depending on approach is large enough to be the single most useful piece of information for any senior leader deciding where to spend an AI capability budget.
The training-led mixed approach, combining structured training for existing employees with targeted strategic hiring, delivers 72.7 per cent measurable AI ROI in our dataset. The Workforce and HR function data is in the HR AI Benchmark 2026, covering capability approach, ROI outcomes, and the 91% deployment rate. Relying primarily on external consultants delivers 13.6 per cent. That is a 59-point gap, from roughly the same starting point.
This does not mean external consultants produce no value. It means that consultants who design and implement without building internal capability alongside them consistently produce lower ROI than those who do. The organisations that hold their ROI are the ones whose own teams can run, maintain, and evolve what gets built, not the ones that remain dependent on external partners to keep the lights on.
The implication is specific: when an organisation is choosing between AI enablement approaches, the question to ask is not “can these people build what we need?” but “will we be able to operate it without them after they leave?”
No piece of writing about AI enablement that draws on our research can avoid the governance finding, because it is the largest single predictor of measurable AI ROI in the dataset , more predictive than which tools are deployed, which function is leading, or how much has been invested.
The rate of measurable AI ROI rises from 4.5 per cent among organisations with no AI governance to 85.2 per cent among those with mature, embedded governance (L5 in our five-level model). The 80-point spread is the strongest structural finding in the survey.
Sixty-two per cent of UK mid-market organisations in our dataset sit at governance levels L1 to L3, below the threshold where ROI at scale is consistently achievable. Most of them believe their governance is adequate, because they have documentation. What they lack is consistent application, measurement, and board-level accountability.
AI enablement that does not address governance is capability-building on an unstable foundation. Training people to use AI tools in an organisation where those tools are not properly inventoried, where nobody can audit what data they touch, and where there is no owner accountable for the programme’s performance will produce individual wins that cannot be aggregated into a board-level ROI story.
The governance work is not glamorous. It is the part most organisations want to skip, usually because it is organisational change work rather than technology work. Our data suggests that skipping it costs roughly 55 percentage points of measurable ROI. For the full governance data across all C-suite functions, see the C-Suite AI Benchmark 2026.
AI enablement services come from several types of organisation, and matching the right type to the right need is where most procurement decisions go wrong.
Strategy and readiness consultancies help organisations assess where they are on the AI maturity curve, define the business case for AI investment, and set the governance architecture before deployment begins. They are most valuable at the start of a programme or when an existing programme has stalled and nobody is quite sure why.
Implementation and deployment partners translate strategy into working systems. They design and build the AI workflows, integrations, and automations that produce actual operational change. The distinction between a good implementation partner and a weak one is whether they build the internal handoff as seriously as they build the system.
Process and document automation specialists operate in a specific part of the AI value chain: the high-volume, repetitive workflows where AI can produce fast, measurable cost reduction. Often the fastest route to a provable ROI story for an Operations or Finance function.
Generative AI applications providers build AI products and applications on top of foundation models. Relevant when an organisation’s need goes beyond off-the-shelf tools into custom capability.
AI agents and agentic workflow specialists work on the more autonomous end of the AI spectrum: workflows where AI systems take sequences of actions with limited human input at each step. A growing category as agentic capability matures.
Training and capability building organisations build AI literacy and applied AI skills across workforces. The training-led mixed approach in our data outperforms other approaches by a wide margin, which makes this category particularly important to get right. Look for providers who can evidence measurable change in how people work, not just completion of training modules.
Data and analytics partners address the infrastructure that makes AI work well, data quality, pipelines, and the measurement systems that connect AI activity to business outcomes. Often underweighted in AI budgets and overrepresented in the reasons programmes stall.
AI governance and risk specialists design and operationalise the accountability frameworks, audit processes, and risk controls that turn an AI policy on the intranet into a governance practice the board can rely on.
The AI Enablement Directory lists vetted AI implementation partners across all eight categories, built specifically for UK mid-market organisations. Providers can join as AI transformation consulting partners.
Every partner is independently reviewed before listing. Use the category filters to identify the type of support your organisation needs at this stage of its programme. A strategy and readiness consultancy is the right starting point if your AI programme lacks a clear business case or governance architecture. A training partner is the right starting point if the tools are in place and the capability gap is the bottleneck. A governance specialist is the right starting point if your AI programme generates activity but struggles to produce board-level reporting.
The diagnostic question is: where, specifically, is the gap between what your AI programme produces now and what you need it to produce? The answer tells you which of the eight categories to start with.
One finding from the aibl survey that does not fit neatly into a capability category but belongs in any honest discussion of AI enablement: management alignment. The broader organisational context is in aibl’s guide to AI transformation for UK mid-market businesses.
Only 24 per cent of UK mid-market organisations in our dataset report full management alignment on AI priorities, meaning all five of the senior leaders we asked in each organisation agreed on what the AI programme was for and how it was being run. The ROI difference between full alignment and partial alignment is 41 percentage points. Organisations where all five senior leaders agree report 78.6 per cent measurable AI ROI. Where three out of five agree, the figure is 37.2 per cent.
AI enablement services can build capability into teams. They cannot fix a situation where the CEO, COO, and CHRO each have a different view of what the AI programme is trying to accomplish. The enablement work lands on whatever organisational foundation exists. If the foundation is fragmented, the capability does not compound.
The alignment question to ask before commissioning any AI enablement engagement: do the senior leaders who own the functions this programme will touch agree on what success looks like in 12 months? If the answers diverge significantly, that conversation needs to happen before any external partner is brought in.
What is the difference between AI enablement and AI adoption?
AI adoption refers to the deployment of AI tools, getting them into the organisation and in active use. Adoption is relatively mature across UK mid-market: 85 per cent of organisations have AI embedded in at least one operational area. AI enablement is the work that comes after deployment, building the skills, governance, measurement infrastructure, and organisational capability that turn AI usage into provable business performance. Most organisations have adopted AI. Far fewer have enabled it.
What is the difference between AI enablement and AI transformation?
AI transformation is the organisation-wide change in how a business operates when AI is deeply embedded across functions. It is the destination. AI enablement is the capability-building work, covering training, governance, implementation support, and measurement, that makes transformation achievable and sustainable. Transformation without enablement tends to produce activity at scale without outcomes at scale.
What does an AI enablement partner do?
It depends on the category. Strategy and readiness consultancies help organisations define their AI business case and governance architecture. Implementation partners build the AI workflows and integrations that produce operational change. Training providers build applied AI capability across workforces. Governance specialists design and operationalise AI accountability frameworks. Data and analytics partners address the measurement infrastructure that connects AI activity to business outcomes. The AI Enablement Directory covers all eight categories with vetted UK partners in each.
How do you measure the success of AI enablement?
By measurable changes in the business metrics the AI programme was intended to move, process cost, time-to-completion, revenue-adjacent metrics, error rates, or whatever the specific business case specified. The aibl survey asked organisations about measurable AI ROI, defined as reported measurable return on AI investment. Across 755 respondents, 49.5 per cent of organisations can demonstrate this today. The gap between that 49.5 per cent and the organisations that cannot is almost entirely explained by governance maturity and capability-building approach, the two things AI enablement directly addresses.
Why do AI enablement programmes fail?
The most common failure modes in our data are governance that exists on paper but is not applied consistently, capability-building that develops individual skills without changing organisational processes, and management misalignment on what the programme is for. External consultants who implement without building internal capability are also overrepresented in underperforming programmes. The training-led mixed approach, which combines structured training for existing teams with targeted strategic hiring, produces measurably better outcomes than consultant-dependent models across our full survey.
Where can I find vetted AI enablement partners in the UK?
The AI Enablement Directory lists independently reviewed AI implementation partners across eight service categories, built for UK mid-market organisations. Categories include AI Strategy & Readiness, AI Implementation & Deployment, Process & Document Automation, Generative AI Applications, AI Agents & Agentic Workflows, AI Training & Capability Building, Data & Analytics, and AI Governance & Risk.
Source: aibl State of UK AI Adoption Survey 2026. n=755 UK mid-market business leaders, January–March 2026, in partnership with Executive Summary (summary.global).
The AI Enablement Directory lists vetted UK implementation partners across eight service categories. All listings are independently reviewed by aibl.
I told you so. It's petty I know, but it feels good when a monster like BCG reinforces a message from your own...
Read more
Rupert Kemp, director of strategy and special projects in the office of the CEO at Google DeepMind, advises...
Read more
A new Harvard Business School working paper by Hyunjin Kim (INSEAD) and Rembrand Koning finds that AI-native...
Read moreGet ahead with the most actionable insights, playbooks and real-world AI use cases you can adopt right now, in your inbox every week