From Experimentation to Impact: Leveraging AI for Business Results

On Thursday, 4 June 2026, Venari Partners convened an executive roundtable at the Pan Pacific hotel in London to explore how organisations can move beyond AI experimentation and begin delivering measurable business value. Led byBayo Ibirogba,  whose experience spans strategy consulting and founding an AI led technology business, the discussion examined a question that is becoming increasingly relevant for leadership teams: as AI capability continues to accelerate, how do organisations create lasting value from it?

One message emerged consistently throughout the session. While model capability continues to evolve at extraordinary pace, the principal challenge facing most businesses is no longer technological capability but organisational adoption. Success increasingly depends on identifying where AI creates value, embedding it into established ways of working and ensuring that productivity gains translate into measurable commercial outcomes.

Here are the key takeaways from the roundtable:

Across many organisations, AI use is still being driven byindividuals and teamsrather than through cohesive enterprise rollout. Employees are already using AI to research, analyse, draft, code and automate routine work, while leadership teams continue to develop governance, operating models and investment priorities.

Three interconnected barriers emerged repeatedly during the conversation. The first is knowledge: understanding what AI is capable of today rather than what it could do only a short time ago. The second is the business case: identifying where AI creates measurable commercial value rather than isolated productivity gains. The third is organisational will. Harnessing dividends from AI often requires redesigning established processes and operating models, and these changes typically progress more slowly than the technology itself.

Collectively, these barriers help explain why individual experimentation has outpaced organisation-wide implementation. The opportunity for leadership teams is therefore not simply to encourage AI use, but to create the conditions in which successful individual adoption can be translated into repeatable organisational capability.

Foundation models continue to improve rapidly. The principal bottleneck has shifted from what AI can do to how effectively organisations absorb those capabilities into day-to-day work. Drawing on research referenced during the discussion, Bayo referred to  the'red-blue gap': an expression taken from research by Anthropic to describe the difference between what AI could theoretically achieve and what has actually been embedded into existing workflows. The opportunity, he argued, lies in bridging this divide rather than simply waiting for the next generation of models.

This represents an important shift in emphasis. Competitive advantage is becoming dependent less on access to the latest technology than on an organisation's ability to integrate existing capabilities into workflows where they improve productivity, quality and decision making.

A significant part of the discussion focused on AI’s economic value and where it is currently accruing. Many workplaces are paying for AI tools, yet much of the immediate productivity benefit remains with individual employees. Unless workflows, expectations and operating models evolve, those gains are unlikely to be captured by the enterprise. As things currently stand, the company pays for the tools while the employee captures the saved time.

AI’s emerging cost paradox was another topic of interest. Although token costs continue to fall, overall AI expenditure is increasing as organisations increase use. The challenge, therefore, lies not simply in reducing the cost of AI, but in deploying increasingly capable models that create meaningful value, rather than using it predominantly for simple tasks (likened during the talk to taking a 'Concorde to go to the supermarket'). The productivity dividend is personal today, but must become organisational if AI investment is to deliver sustainable commercial returns. This will require organisations to rethink workflows, redeploy capacity and decide how the additional value created by AI should be realised.

Despite the many prevailing assumptions surrounding AI and employment, this technology is progressively automating individual tasks rather than replacing professions overnight. Work that is computer-based, repetitive, highly structured, requires limited judgement and involves low stakeholder complexity represents the clearest opportunity for automation.

As such, organisations are likely to redesign roles before fundamentally reducing headcount. Software engineering as an early example, with changing demand for junior developers potentially providing an indication of how other knowledge-based professions may evolve. For leadership teams, the more useful question is not which jobs disappear, but the activities within each role that are most likely to be augmented or automated.

One of the most thought-provoking discussions centred on the subject of ‘AI slop'. AI has raised the baseline quality of written output, but it has also increased the volume of polished material that lacks depth, originality or critical thinking. The recurring question was simple: 'Did you do the work and AI refined it, or did AI do the work and you refined it?' The concern is not that AI produces poor output, but that convincing presentations can disguise weak reasoning or discourage appropriate challenge.

Against this backdrop, the premium placed on human judgement is likely to increase rather than diminish. The ability to question assumptions, apply context and exercise professional judgement becomes more valuable as AI becomes more capable. AI should therefore be treated as a capable junior colleague: it requires context, guardrails, quality assurance and, above all, human supervision.

As access to advanced AI models becomes increasingly widespread, sustainable competitive advantage is unlikely to come from the technology itself. Instead, differentiation will depend on assets that competitors cannot easily replicate, including proprietary data, customer relationships, trusted brands, operational excellence and deep sector expertise.

The key strategic question is therefore not simply how to use AI, but how AI can amplify what already makes the organisation distinctive. If everyone has access to the same models, what remains uniquely yours? The businesses best placed to create lasting value will be those that use AI to reinforce existing strengths rather than viewing the technology as a source of competitive advantage in its own right.

Three practical lenses emerged for assessing AI opportunities:

Knowledge

• Where is the gap between what AI could achieve and what is currently embedded?

• Are we benchmarking ourselves against competitors or against what is genuinely possible?

Business case

• Where does the productivity dividend currently go?

• If capacity were no longer constrained, how would work be redesigned?

Will

• Which activities should AI assist, automate or reinvent?

• What organisational barriers are preventing faster adoption?

These observations point to a broader shift in how organisations should think about AI. The discussion was less about the technology itself than the organisational capabilities required to translate technological progress into measurable business value.

For many organisations, the question is no longer whether AI is capable of transforming work. The challenge is embedding it thoughtfully, redesigning workflows where appropriate and ensuring that the benefits are realised at an organisational rather than purely individual level. Ultimately, the organisations most likely to create lasting value will be those that bridge the gap between technological possibility and organisational execution, embedding AI for measurable commercial impact while strengthening the capabilities that already set their business apart.

Tags: