Start With Value, Not Hype: How to Find Your AI Starting Point

The question most organisations are not asking

Boards are asking whether they have an AI strategy. CEOs are asking whether their competitors are ahead of them on AI. Technology teams are asking which platforms and vendors to evaluate. Almost nobody is asking the question that actually determines whether AI investment generates returns:

In this specific organisation, given our specific operations, our specific market position, and our specific competitive dynamics — where will AI create the most measurable value, fastest, with the resources we are actually prepared to commit?

This question is harder to ask than it sounds, because it requires honesty about three things most organisations find uncomfortable to examine directly: what their operations actually look like at a granular level, where their competitive position is genuinely strong or weak, and what their organisation is actually capable of executing.

The organisations that skip this question and begin AI implementation by following what their industry peers are doing, or by adopting whatever the most prominent vendors are currently promoting, spend significant money and time to discover that the AI they built solves a problem they do not have, or a problem they have but that is not their most expensive one.

Following what others are doing with AI is a strategy for not falling too far behind. It is not a strategy for gaining competitive advantage. Advantage comes from finding, first, the place where AI creates disproportionate value in your specific context — and then moving faster than your competitors to capture it.

The honest organisation assessment

Before a meaningful AI strategy can be built, three assessments need to be completed honestly — not optimistically, not strategically, but accurately.

Where does time actually go?

In most organisations, there is a significant gap between where leadership believes skilled labour time is spent and where it is actually spent. AI creates the most value by automating or augmenting the activities that consume the most time from the most expensive people — but only if those activities are correctly identified.

This requires going below the level of job titles and departmental functions to the level of individual tasks. How many hours per week does a compliance team member spend reading and summarising regulatory documents? How many hours does a finance analyst spend reformatting data from one system to another? How many hours does a customer service team spend on queries that have identical answers? These are the numbers that reveal where AI ROI is actually waiting.

The activities with the highest AI ROI are typically high-volume, information-intensive tasks that require judgement but follow recognisable patterns. They are rarely the tasks that appear most valuable on an organisation chart.

Where are the competitive pressure points?

AI strategy in isolation from competitive context is incomplete. The relevant question is not only where AI can create value internally, but where competitive dynamics mean that AI-enabled performance improvement will translate into durable market advantage.

In markets where speed of response to customer queries is a differentiating factor, AI-enabled customer operations can move the needle on a metric that directly affects revenue retention. In markets where pricing agility matters, AI-enabled market analysis and automated pricing can create an advantage that compounds over time. In markets where regulatory compliance is a barrier to entry or a source of client confidence, AI-enabled compliance operations can strengthen a position that competitors cannot easily replicate.

The starting point for AI should be where the combination of internal impact and competitive leverage is highest — not where the technology is most impressive or most frequently discussed.

What can this organisation actually execute?

The most honest and most frequently avoided assessment is of organisational execution capability. AI implementation requires data that is accessible and reasonably clean, technology infrastructure that can support integration, leadership with both the authority to make decisions quickly and the commitment to see implementation through past the difficult middle, and frontline teams that are prepared to change their workflows.

None of these conditions is guaranteed to be present. All of them can be created — but creating them takes time and investment that must be factored into the strategy. An AI initiative that assumes data is clean when it is not, or that assumes organisational alignment exists when it does not, will fail regardless of how good the underlying technology is.

The honest capability assessment identifies which AI opportunities are genuinely executable in the near term with current conditions, which are executable with specific pre-conditions that can be established in a defined timeframe, and which require fundamental changes in data, systems, or culture that make them medium-to-long-term propositions regardless of how attractive the ROI case looks.

The market and competitive assessment

Understanding where competitors are investing in AI is useful input. It is not a strategy. The mistake is treating competitive parity — not falling behind — as the goal, when the actual goal should be competitive advantage.

Parity-seeking AI investment has a structural problem: if everyone in an industry implements the same AI capabilities at roughly the same time, the productivity gains are competed away in pricing, and the industry ends up with higher costs of AI infrastructure and no durable margin improvement. This is the pattern that tends to emerge when AI adoption is driven by fear of being left behind rather than by analysis of where differentiated value can be created.

The more productive framing: where in our specific competitive context is AI capability currently underexploited — either because competitors have not yet seen the opportunity, or because executing on it requires capabilities we have and they do not?

This analysis requires genuine knowledge of competitive operations, not just competitive products. How do competitors actually handle customer queries? How quickly do they update pricing? How comprehensively do they monitor compliance? How efficiently do they process the same administrative tasks you do? These operational differences are where AI creates asymmetric advantage.

How to prioritise: the value-speed-feasibility framework

Given honest assessments of internal operations, competitive context, and organisational capability, prioritising AI initiatives is a matter of applying three criteria:

• Value. What is the quantified business impact — in revenue, cost, time, or risk reduction — if this AI initiative performs as designed? Not a theoretical maximum, but a conservative estimate based on real baseline data.
• Speed to value. How quickly can a working version of this AI be in production generating real returns? This is determined by data availability, technical complexity, integration requirements, and organisational readiness — all of which must be assessed honestly.
• Feasibility given current conditions. Can this organisation actually execute this initiative, with its current data, systems, team capability, and leadership commitment? Not could it in principle — can it now, with what it actually has?

The highest-priority AI initiative is the one that scores well on all three — not necessarily the highest on any single dimension. A high-value initiative that will take eighteen months to be feasible is less valuable as a starting point than a moderate-value initiative that can be in production in eight weeks, because the latter generates returns that can fund and inform the former.

The organisations that make the best AI investment decisions are those that resist the temptation to start with the most ambitious initiative and instead build a track record of successful, measured deployments that create compounding capability and confidence.

What this means in practice

The outcome of this analysis should be a short list — not a long one — of AI initiatives ranked by their combination of value, speed, and feasibility. The top priority should be executable within a defined timeframe, should have a clear and measurable success criterion, and should create genuine value that the organisation can point to as evidence of AI's business impact.

That first successful deployment is not just valuable in itself. It builds the organisational experience, the data infrastructure, the cross-functional relationships, and the leadership confidence that make subsequent AI deployments faster, cheaper, and more effective. The organisations that have been doing this systematically for two years are not just ahead in AI — they are ahead in the organisational capability to keep improving at AI. That gap is very difficult to close quickly.

The question to ask before any AI investment is not "what are others doing?" It is "what is the specific problem in our specific organisation where AI will create the most value, that we are genuinely positioned to solve right now?" Answer that honestly, and the strategy follows naturally.