Day one. You sit down and the inbox is already full. Vendor outreach, demo requests, “quick chats”, and somewhere in there a board member asking when they will see something. Everyone wants a meeting. The instinct is to start saying yes, to book the demos, to look like a person who is moving. The better first move is quieter and harder to hold to. Talk to your own people before you talk to a single supplier.
This is the second post in a six-part run on being handed the AI mandate. The first looked at why a problem-first stance beats a tool-first one. This one turns that stance into a day-by-day first month, because the first thirty days decide whether the rest of the mandate works.
Why does the first month decide the whole mandate?
The first thirty days set the trajectory because the choice you make in them is binary. You either spend the month assessing or you spend it deploying, and the version that assesses is the version that tends to work. Deploying early reads like progress to a board, but it commits you to tools before you understand the problems they are meant to solve.
The pull to show movement is strong, and it is exactly what tips initiatives into the failure column. MIT’s 2025 research found that the large majority of generative AI pilots produce no measurable impact on the P&L, and the failures sit upstream of the technology, in unclear problems and unready data. The pressure is sharper still if you were handed this because you make things happen, not because you know AI. Korn Ferry calls that the AI readiness paradox, strong operators given AI leadership without the AI-specific footing the task needs. A demo you said yes to in week one is a commitment made before you had the information to make it well. A month of honest groundwork costs you the appearance of speed and buys you the substance of it.
What does the listening tour in weeks one and two look like?
Weeks one and two are a listening tour. You sit with the people who do the work and you learn how it actually flows, where the friction is, what gets redone, what waits on someone else. You are not pitching AI. You are building the map that tells you where AI would add value rather than automate a process that is already broken.
There is a second reason to start with people rather than suppliers. Employee fear, about job loss, about being watched, about a tool that gets things wrong in their name, hardens into passive resistance the moment it is ignored. Open with listening and you defuse some of that before it sets. Carry one test into every conversation, the one Addepar puts to executives weighing an AI purchase. Would this initiative still matter if it did not use AI. If the honest answer is no, you have found a demo dressed as a problem, and you can let it go. The reframe that helps here is to describe the work as building AI readiness, not implementing AI, and to talk about freeing people from low-value work rather than replacing them. The language is not spin. It changes what you go looking for.
How do you read the current state across five readiness dimensions?
You read the current state across five dimensions, and you write each one down honestly. The five are data maturity, team enablement, technology infrastructure, strategic alignment, and governance and risk. Data is one of the five, not the whole job. The output is a single page that says, plainly, where the business stands on each, with no effort to make it look further along than it is.
Data tends to be the constraint that bites first. Poor data quality is named by 77 per cent of firms, on Gartner’s figures, as the biggest barrier to using AI responsibly, so a candid read of how clean and accessible your information is matters more than any tool comparison. Strategic alignment is the one delegates underweight. If you cannot say in a sentence how AI connects to what the business is actually trying to do over the next year, that gap will surface at the board table whether you name it now or not. Read all five. Resist the urge to start fixing any of them yet.
What is shadow AI and why map it first?
Shadow AI is the AI your people are already using off the books, the personal accounts and free tools they have brought in without anyone sanctioning it. You map it in the first month because it is both your real starting line and your first risk surface. The sanctioned picture understates the actual footprint. A meaningful share of employees are already using personal AI tools at work, often before any formal programme exists.
Read the map as diagnostic information rather than a hunt for wrongdoing. Where people have reached for AI on their own, they have told you where the friction is real enough to be worth solving, which is the most honest demand signal you will get. The risk side matters too, because client data pasted into an unsanctioned tool is an exposure you now know about and can manage. Read shadow AI as a free piece of research and a first governance task at the same time. Both readings are correct.
What do you commit to at day 30?
At day thirty you give the board an honest current-state read and a plan to plan, not a tool you have already bought. You show them where the business stands across the five dimensions, what AI your people are already using, and where the genuine value looks likely to sit. Then you commit to bringing them a ranked shortlist of opportunities in the next thirty days. You do not deploy anything yet.
Where you point them matters as much as what you show. The instinct is to promise a customer-facing showpiece, because that is what looks impressive. The evidence points the other way. MIT found that back-office automation produces the highest returns while sales and marketing pilots show the lowest, despite attracting the most funding. So aim the early attention at the unglamorous work, document processing, internal admin, the repetitive tasks that eat hours without anyone counting them. The sixtieth-day post in this series turns that shortlist into a ranked plan, and the board update post handles the harder conversation about resetting expectations on timelines. For now, the move is to assess well, say so honestly, and resist the demo. If you want a second pair of eyes on your first thirty days, book a conversation.
