Adoption advisor
Your 90-day AI adoption plan
Answer four questions and get a costed starter stack, a phased rollout, a training path, a usage policy checklist and the numbers to judge it by.
Not sure how mature your team is?
Take the 3-minute AI maturity assessment and we will size this plan around your score.
About your team
What we would do
For 10 people focused on writing & content, with no AI in regular use yet and about $40 per person a month to spend, we would start in AI Chat Assistants, AI Writing Assistants, AI Presentation Tools and pilot with 3 people before buying seats for everyone.
Pilot cost
$100/mo
Annual equivalent
$1200
Your stated budget
$400/mo
Recommended starter stack
Buy this stack for your team
$190per month
Enterprise covers the writing & content rollout end to end for everyone on the team: the 3 recommended tools set up in a shared workspace, the 90-day plan, the training path and the usage policy. Billed at $19 per person per month. Tool subscriptions themselves (about $100/month) stay on your own accounts.
- Shared workspace with this writing & content stack pre-loaded
- The 90-day rollout plan assigned to owners
- Training path and certificates for every seat
- Usage policy, spend tracking and stack reviews
Watch these tools before the kick-off
All video tutorialsA short walkthrough for each tool in your stack — use one in the 45-minute kick-off session.
The 90-day rollout
Start with 3 people, not everyone
- Pick 3 volunteers who already do writing & content work daily and give them ChatGPT.
- Name the two tasks each of them will use it for, and write down how long those tasks take today.
- Agree the data rules before the first login: what may be pasted in, and what may not.
- Book a 45-minute kick-off around the tool’s video walkthrough and written steps on this site.
Measure the saving and fix what got in the way
- Compare the before-and-after time on those same two tasks and record the difference in hours.
- Collect the failures as well as the wins — that list becomes your review checklist.
- Add the supporting tools (QuillBot, Gamma) only where the pilot showed a clear gap.
- Publish one short internal write-up from a colleague, not from management.
Extend to the remaining 7 people
- Buy seats in one batch so billing and admin stay in one place.
- Run the same 45-minute onboarding for each new group, using the prompts your pilot proved.
- Set a monthly review of spend against value in the cost calculator, and cancel anything unused.
- Assign one owner per tool who answers questions and watches for pricing changes.
Training path
- What is AI, really?Beginner · Everyone in the pilot, week one
- AI Fundamentals & TerminologyBeginner · Pilot group, month two
- Prompt Engineering that worksIntermediate · Owners and power users
AI usage policy checklist
- Write down which tools are approved, and that anything else needs a quick check first.
- State plainly what may never be pasted into a tool: customer records, credentials, unreleased financials, health or ID data.
- Require a named human reviewer for anything that reaches a customer, a regulator or a contract.
- Turn off training on your data where the plan allows it, and record that decision per tool.
- Keep AI-assisted work labelled internally so reviewers know what to check harder.
- Review the tool list and spend once a quarter; remove what nobody used.
What usually goes wrong
- Confident-sounding errors: every tool here scores below 100% on accuracy, so the review step is not optional.
- Shadow usage: if you do not approve something usable, people will use their own accounts with company data.
- Seat creep: paid seats keep billing after people stop using them — the monthly review catches this.
- Overreach: starting with a company-wide rollout is the most common way an AI project quietly dies.
How to tell it worked
- Hours saved per person per week on the two named tasks.
- Share of the pilot group still using the tool voluntarily in week eight.
- Monthly cost per active user against your $40 target.
- Quality: number of outputs sent out that needed correction after review.