Close the skills gap in three moves. Fund enablement as core spend with a ring-fenced line, not an afterthought. Mandate AI and data fluency as a baseline skill every role holds. Then build training, peer learning, and reference material around the specific stack you run. Capability you cannot operate is capability you do not have.
This is the operating sequence behind why your team cannot run the tools you bought. The article named the mechanism: recruiting AI skills, at 49%, and training staff, at 46%, are the top barriers leaders cite to capturing AI value (Wharton, 2025), and the same shortage leaves half your martech dark (2X). This playbook is how you put the capability back.
Step 1: Fund enablement as core spend, not overhead
Start with the budget, because that is where enablement dies. The training line is the first thing cut when the quarter tightens, which is why the last three platforms never got adopted.
Fix it with a rule: no tool budget is approved without an enablement line attached. Ring-fence a set share of every licence for the people who run it, and protect that share the way you protect the licence itself. The point is not the exact percentage. It is that usage stops being a hope and becomes a funded thing with an owner.
Give the line one owner, sitting across the budget, not inside a single channel team. They are accountable for the money that turns paid capability into operated capability. Buying capability and assuming usage follows is the reflex you are replacing, so make the enablement spend a condition of the buy, not a nice-to-have after it.
What to measure at this step: an enablement line ring-fenced against every major tool, with a named owner and a figure that survives the next budget cut. That is the whole deliverable of step one, and it costs a policy, not a platform.
Step 2: Mandate AI and data fluency as a baseline skill
An enablement budget with no standard funds random courses. Step two sets the standard.
Define a fluency baseline: the floor of skill every role is expected to hold. Reading data without a hand-hold. Framing a question to a model and judging the answer. Running the core workflow of your stack end to end. Write it down per role, so it is a clear expectation rather than a vague wish that some people are good with tools.
Mandate it, the way high-maturity organisations do. Fluency left to the keen few produces two experts and a queue of people waiting for them. A mandated baseline sets the floor every new tool lands on, so the next platform meets a team that can absorb it instead of one that ignores it.
Then wire it into hiring. Screen for the ability to operate your kind of stack, not just for having a tool on a CV. A new hire who cannot run the systems you already own widens the gap you are trying to close.
What to measure at this step: a documented fluency baseline per role, updated hiring screens, and a first read of who currently meets the baseline and who needs bringing up. That map tells you where the training in step three has to land.
Step 3: Build training around the stack you actually run
Generic AI training teaches a tool nobody uses. Step three makes the training specific to the systems your team opens every morning.
Build three things around your own stack. Internal training on your actual tools and workflows, run by the people who use them well. Peer learning, so the two experts you already have teach the rest instead of hoarding the knowledge. Reference material, short and specific, so someone stuck at the second screen has a place to look that is not the vendor’s manual.
Sequence it against the map from step two. Start with the tool that has the most paid capability and the fewest trained users, because that is where the return is largest and already funded. Train the workflow end to end, not a feature tour, so people leave able to do the job, not just recognise the buttons.
What to measure at this step: reference material for each core tool, a running peer-learning cadence, and utilisation of the trained features rising against the baseline you took in step two. Trace the same numbers the martech utilisation playbook tracks, because adoption is the proof the training worked.
How to train your team to hold the fix
Three steps set the fix. A habit holds it, because new features arrive every quarter and the baseline decays without one.
Teach one rule first: every tool comes with a rollout, or it does not come at all. No login goes out without training attached and a reference page written. This is the behaviour change that matters, because the gap reopens the moment a platform lands with a demo and nothing else.
Run a short quarterly fluency review alongside the stack review. The owner walks what got trained, where utilisation moved, and which new features need enablement next. It keeps the baseline live and stops the skills gap quietly reopening while everyone was busy.
Make peer teaching part of the job, not a favour. Recognise the people who bring others up, because in a team that shares fluency the mandated baseline holds itself. That is the human-in-the-loop layer this pillar is built on: AI agents that work, and people who can lead them.
Where Morphy helps
Most teams know their people cannot run half the stack. They do not have a spare month to build fluency training on their own tools, and generic courses have already failed them.
That is a four to eight week engagement. We map the fluency gap across your stack, build training and reference material on your actual tools, and set the baseline and the enablement line so it holds after we go. You leave with a trained team, a fluency standard written into roles, and utilisation moving on the capability you already pay for.
The metric is honest and simple: the share of your paid capability that people can actually operate, measured before and after. No new platform. The gain comes from running the stack you already own.
Go deeper on customer data maximization
Three ways forward. Pick the one that fits where you are.
- Get the playbook. Practical notes on turning the customer data you already own into revenue, straight to your inbox. Join the newsletter at the foot of this page.
- Take the assessment. Score your customer data maximization in four minutes and see your top revenue blockers. Start the assessment →
- Book a meeting. Bring your data problem. Leave with a prioritised fix, not a platform pitch. Book a call →
The playbook companion to Why can’t your team run the tools you bought, and how do you fix it?. Post 24 of 25 in the Customer Data Maximization series.
Frequently asked questions
How much should I spend on enablement?
Ring-fence a set share of every tool budget for the people who run it, and protect it when the quarter tightens. The exact figure matters less than making it a standing line rather than an afterthought. Buying capability and assuming usage follows is what leaves half the stack dark (2X).
What does an AI and data fluency baseline look like?
A defined floor of skill every role is expected to hold: reading data, framing a question to a model, and running the core workflow of your stack. High-maturity organisations mandate it rather than leaving it to the keen few. It is a baseline expectation, the way spreadsheet skill became one.
How long does closing the skills gap take?
The decisions take days: fund the line, set the baseline, name who trains on what. The first fluency lift across a team takes four to eight weeks of training built on your own tools. After that it is a standing rhythm, because new features keep arriving and the baseline has to hold.
Why train on my own stack rather than a generic course?
Because a generic course teaches a tool nobody in your building uses. Recruiting and training are the top barriers leaders name to AI value (Wharton, 2025), and training that never touches your stack does not move either one. People learn the systems they open every morning.