This buying checklist is published by Day Seven, an AI training provider. Use it to compare practical exercises, trainer experience, delivery and follow-up against your team's brief. It is not an independent review, a ranking, or a scored comparison of other companies.
It sets out seven criteria, with the question to ask for each one and the evidence worth requesting. Apply them to us as strictly as you apply them to anyone else.
Find out who will stand in front of your team, not who runs the company. Ask what they have delivered to similar teams and in similar sectors, and whether the person named in the proposal is the person delivering the session.
Ask: Who is delivering, what have they delivered before, and will they be there on the day?
Evidence: Named trainer, described delivery history, a conversation with them before you book.
A demonstration is not training. People retain what they have done themselves. Ask what proportion of the session is participants working in the tool, and what they will physically produce during it.
Ask: How much of the day is people typing, and what will they have made by the end?
Evidence: A run sheet or agenda showing exercise blocks, not just topics.
Training built around a generic syllabus lands differently to training built around the reports, tickets and emails your team already produces. Ask whether the content is shaped after a scoping conversation, and whether it targets the tool you actually license.
Ask: Will the exercises use our tasks and our tool, or a standard example set?
Evidence: A scoping call before design, and draft exercises that name your own workflows.
People need something realistic to work on without putting confidential or personal data into the wrong tool. Equally important: whether the training teaches people to check what the AI produced, rather than trusting it.
Ask: What data will people use in exercises, who approves it, and how is checking taught?
Evidence: A written data rule for the session, plus a checking or review step inside the exercises.
Group size decides whether people get individual help. Delivery mode decides whether shift patterns and multiple sites can be covered. Accessibility needs — screen readers, captions, materials in advance — should be asked about before you raise them.
Ask: How many per session, what happens with larger headcounts, and how are access needs handled?
Evidence: Stated cohort size, a plan for multiple cohorts, and a straight answer on accessibility.
Ask how participants will apply the learning after the session and what support is available. Ask what people keep — prompts, guides, an adoption plan — and whether anyone is available when the team hits a wall.
Ask: What do participants keep, and what support exists after the day?
Evidence: Sample materials, and a clear statement of what follow-up is included versus extra.
A logo on a website is not a reference. Ask for a client who has given permission to be contacted. Separately, ask how the provider expects the training to be evaluated — a provider with no view on measurement is telling you something.
Ask: Can we speak to a client who has agreed to that, and how will we know it worked?
Evidence: A permissioned contact, plus a proposed measurement approach agreed before delivery.
Print it, or copy the rows into a spreadsheet with a column per provider. Nothing gated, no form.
| Criterion | Evidence to request | Included or extra? | Open question |
|---|---|---|---|
| Relevant trainer experience | Named trainer, described delivery history, a conversation with them before you book. | Confirm in writing on the proposal | Who is delivering, what have they delivered before, and will they be there on the day? |
| Actual hands-on practice | A run sheet or agenda showing exercise blocks, not just topics. | Confirm in writing on the proposal | How much of the day is people typing, and what will they have made by the end? |
| Fit to your roles and your tools | A scoping call before design, and draft exercises that name your own workflows. | Confirm in writing on the proposal | Will the exercises use our tasks and our tool, or a standard example set? |
| Approved practice data and output checking | A written data rule for the session, plus a checking or review step inside the exercises. | Confirm in writing on the proposal | What data will people use in exercises, who approves it, and how is checking taught? |
| Cohort size, delivery and accessibility | Stated cohort size, a plan for multiple cohorts, and a straight answer on accessibility. | Confirm in writing on the proposal | How many per session, what happens with larger headcounts, and how are access needs handled? |
| Materials and support afterwards | Sample materials, and a clear statement of what follow-up is included versus extra. | Confirm in writing on the proposal | What do participants keep, and what support exists after the day? |
| References you may contact, and evaluation | A permissioned contact, plus a proposed measurement approach agreed before delivery. | Confirm in writing on the proposal | Can we speak to a client who has agreed to that, and how will we know it worked? |
Set out separately so you can hold it against the same criteria, rather than taking our word for it in the middle of the checklist.
More detail: about Day Seven, AI training, consulting and enterprise AI training.