Data Booster
5 min read
September 4, 2026
Everything we build is customised. A programme runs on the client's own tools, their own data and the use cases their people argue about in the corridor, and that is the reason it lands rather than being politely completed. We are not about to change that.
But customisation has an awkward property: it asks the buyer to decide before there is anything to look at. A deck describes an approach. A demo shows a platform. Neither answers the question that actually decides it, which is what one of your colleagues will be doing for an hour on a Tuesday, and whether they will still be doing it on the Thursday.
So we have started putting courses on a shelf. Not a sample and not a trailer: the real thing, in the same platform a full programme runs in. Four can be started today. They come from the six flagship courses we have the most evidence for, which is why these and not six others.
An hour settles three things that months of scoping usually does not.
It settles the level. Every organisation we meet is convinced its people are further along than they are, or further behind. Put a group through AI Fundamentals, 1,445 learners so far, and you find out within the hour whether they can tell a real use case from a demo. That is the single most useful thing to know before designing anything.
It settles the tooling argument. A Microsoft organisation that has paid for Copilot and is not seeing it used usually has a habit problem rather than a licence problem. Introduction to Microsoft Copilot puts it to work across Outlook, Teams, Word, PowerPoint and Excel, and an hour of that tends to end the debate about whether the tool was the issue.
It settles how deliberate people already are. Introduction to Prompt Engineering, rated 4.7, goes past writing single prompts to designing, refining and reusing them. Groups that thought they were fluent discover they were improvising, which is a much better thing to find out in an hour than in month three of a programme.
If you want to pick one, pick by the problem rather than the topic.
A group that has dashboards but does not trust them should start with Data Literacy Fundamentals, 1,248 learners, which is about asking the right question and knowing which analysis answers it. A leadership team that keeps approving AI pilots without a way to judge them starts with AI Fundamentals. People who already prompt daily go to Introduction to Prompt Engineering. And an organisation with Microsoft 365 everywhere and Copilot nowhere starts with the Copilot course.
Those four are the shelf today. None of them needs a conversation with us first.
Two of the six are not on the shelf yet, and they are the ones with the most behind them. Effective Prompting for Everyone is the course the most people have been through, 1,902 of them, and it almost always resolves the same debate: the problem was never which assistant the company bought, it was that nobody had been taught to ask it properly. Responsible AI, rated 4.2 across 389 ratings, is the one that reliably produces the awkward questions about bias, accountability and who carries the judgement call. If those come up in an hour, they were going to come up in the programme, and it is considerably cheaper to hear them now.
The shelf keeps moving with the work, and right now that means agents: what makes a system agentic, what changes in your own role once an agent writes the first draft, and how to build one that takes a real task off your hands. Next to those sit courses on the assistants teams have actually standardised on, Claude and Gemini alongside Copilot.
None of this replaces the design conversation. Which courses go into a programme, in what order, and what gets built from scratch around them still depends on who you are training and what they need to do afterwards. The point is only that the conversation is better when the people in it have already done an hour.
All six are on our courses page. To do that hour first, request test access. To go straight to the design conversation, book a discovery call.
Member of
Nominated
