The mandate arrives looking like an opportunity. By the time you’ve read it properly, you realise the board has set a timeline that doesn’t match the work. They want AI running through the whole operation, results visible by the next quarterly update. You know that’s not realistic. The question isn’t whether to say something; it’s how to say it without the board reading “we’re behind” when what you mean is “the original timeline was never achievable.”
There is a skill to this. And it has nothing to do with managing expectations downward.
What’s actually driving the board’s quarterly AI expectations?
Boards have been absorbing AI headlines since late 2022. By the time a mandate reaches someone inside the business, the board has already formed a view of what AI does and how fast it arrives. That view is shaped by vendor demos, competitor announcements, and research briefings that emphasise output without explaining the underlying delivery sequence. The expectation gap comes from genuine ignorance of that sequence, not from cynicism or bad faith.
McKinsey’s research on generative AI in the workplace describes productivity uplift that is real but distributed across a long implementation curve. BCG’s 2025 analysis found AI usage rising while measurable business impact lagged behind adoption rates. This is the pattern boards consistently miss: deployment activity becomes visible almost immediately; the compounding of that activity into P&L impact takes considerably longer.
Meaningful ROI from enterprise AI commonly runs to 12 to 24 months from the start of serious implementation work, and some analyses place it at two to four years when the business first needs to address data readiness, process redesign, and staff enablement. The board’s quarterly expectation reflects a different kind of project. The work you’ve been handed reflects another kind entirely.
What goes wrong when you accept the board’s timeline without comment?
When you accept a timeline without comment, you own it. Every quarter that passes without visible results becomes evidence of slow execution in the room where it matters, not evidence of inflated expectations set before you arrived. The accountability trap closes at the first briefing. Staying quiet in that initial conversation means carrying the gap every quarter thereafter, with the original timeline as the only reference point the board has.
EY’s research on board AI governance found that boards are increasing their scrutiny of AI programmes and expecting clearer ROI framing from the executives responsible. The delegate sits at the centre of that scrutiny. If the board was told results would arrive in one quarter and they arrive in six, the credibility question lands on the person who managed the programme, not on whoever shaped the original expectation.
There is a practical problem too. Unrealistic timelines drive poor sequencing decisions. Boards apply pressure; delegates respond by cutting corners on data readiness and governance to show visible activity quickly. PwC’s AI predictions research consistently identifies inadequate preparation, rather than technical capability, as the leading cause of AI initiative failure. A timeline you have not challenged becomes a constraint you have agreed to deliver against.
Where does the expectation gap tend to surface?
The gap between board expectation and operational reality usually surfaces at the quarterly update, not at project inception. The initial mandate meeting has momentum and goodwill; the board is engaged. Three months later, the work is still in readiness assessment, no P&L results are visible, and the board is comparing progress against the vendor demonstration they saw six months ago. That is the hardest conversation to walk into unprepared.
Korn Ferry’s research on AI leadership readiness describes a recurring pattern in which strong operators are handed AI mandates without the supporting structure to set expectations accurately at inception. Their framing is the “AI readiness paradox”; the people assigned to lead these programmes often lack AI-specific competencies, and the boards overseeing them often lack operational context. Both sides are working from incomplete information, and the quarterly update is where that shows.
The readiness phase doesn’t look like progress from the outside. Data audits, process mapping, staff capability assessments, and governance frameworks produce no visible output the board can point to in a slide. The work is genuine. The artefacts are not the kind that land well in a quarterly update.
How do you reset the timeline without losing authority?
The framing matters as much as the substance. Presenting a revised timeline as “we need more time” loses authority. Presenting it as “here is the phased delivery sequence and the value checkpoint at each stage” is what a competent leader does when they understand the programme. The difference is the presence or absence of a structure the board can evaluate.
A three-phase shape works well in practice. The first six months focus on foundation work, readiness assessment and early pilots rather than P&L impact. Months six to eighteen shift into an expansion phase, where quick-win results accumulate and the board begins receiving real data. From month eighteen onwards, compounding returns start showing in the numbers.
The compounding argument is the one that lands with boards. A business that gets its data right in months one to three, and its processes right in months four to six, reaches genuine ROI faster than one that skips those stages and deploys tools onto unprepared ground. Slower in the first quarter, faster overall. Boards tend to accept this when it is presented as an engineering constraint, not a scheduling preference.
Scaled Agile’s board communication research identifies five questions every board should be able to answer about an AI programme. How does it align with the core business model? How will ROI be measured? What risks are being managed? What capabilities are being built? What competitive advantage is being created? A phased plan with named value checkpoints answers all five in one document, which is why it lands better than a revised end date presented in isolation.
What sits alongside timeline management?
Timeline resetting is one part of a broader board communication challenge for the delegate. Once the board accepts the phased sequence, the next pressure point is reporting. How do you demonstrate progress when financial results have not arrived yet? The dual-ROI approach, tracking leading indicators alongside eventual financial outcomes, addresses that specific problem and connects directly to how the phased timeline is framed.
Propeller’s research on AI ROI measurement describes this as tracking “trending ROI”, meaning early activity metrics, adoption rates, time saved per task, and error reduction rates, in parallel with the “realised ROI” that follows months later. When you present the phased timeline to the board, framing it around both types of measurement gives them something concrete to follow during the foundation and expansion phases. The board’s confidence holds when it can see indicators moving in the right direction, even before the financial outcome is confirmed.
The related questions of how to report AI progress before results arrive, and how to build a business case for a sceptical board from the outset, are each addressed elsewhere in this series. Timeline resetting is the entry point; the reporting discipline is what holds the board relationship across the 12 to 24 months that follow.
If your board is already applying pressure on a timeline that was never realistic, the starting point is usually working out what a phased plan actually looks like in your specific business. Book a conversation if that’s where you are.
