- The right AI training depends on the size of the team and its main bottleneck: the manual tasks the business is trying to address.
- Start with an AI audit that finds the main AI opportunities and bottlenecks, puts a cost on each manual task in resource time, and lets leadership align on guardrails.
- Then follow a 90-day plan that maps the first 90 days: a 1 to 2 day AI training on real processes, the ChatGPT or Claude setup so everyone works the same way, and the start of implementation with an AI specialist. Implementation runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days.
What training should a team adopting AI do, and in what order?
The right training for a team adopting AI depends on the size of the team and its main bottleneck, meaning the manual tasks the business is trying to address. That is why the recommended path starts with an audit, not a course.
Most teams start in the wrong place. Someone buys ChatGPT or Claude licences, a few people watch tutorials, and usage stays patchy while the output still sounds generic. The training was about tools. It should have been about the team's work.
For most 5 to 50 person businesses and marketing agencies, we recommend this path. It follows our Identify, Upskill, Integrate method.
- An AI audit of the team's main AI opportunities and bottlenecks, the cost of each manual task, and the guardrails leadership agrees on.
- A 90-day plan that maps the first 90 days: the training, the ChatGPT or Claude setup and the start of implementation.
- A 1 to 2 day AI training, depending on time and team size, built around the team's real processes.
- Training on ChatGPT or Claude, so everyone is set up, understands the rules and guardrails, and works in the same way.
- Implementation, where an AI specialist helps build AI workflows. It starts during the first 90 days. Implementation runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days.
The goal is capability. Good training makes your people better at the work they already do well, a point we expand on in Is AI replacing marketers? If you are comparing formats such as in-person sessions or online courses, our guide to AI training for employees and corporate AI training covers the options. This article covers the recommended path.
Why should AI training start with an AI audit?
AI training should start with an AI audit because it shows where the team's time goes before anyone is trained. An AI consultant, such as Daniela Valencia, goes in and looks at the team: the main AI opportunities and the main bottlenecks.
Most importantly, the key difference Reclaim Creative brings is identifying the cost of each manual task and the potential savings in resource time. A tool tour teaches features. An audit shows you that your account managers rebuild the same client update every Friday, or that your in-house team rewrites product copy three times before it is approved. Those jobs become the training material, and people see the point straight away.
Put a rough number on each task. For an in-house team, use staff cost, which is what a team member costs the business: about $70 an hour. As an illustration, if four people each spend 10 hours a month formatting campaign reports, that is $70 × 4 × 10 = $2,800 a month, or $33,600 a year in resource time. For an agency, use the rate card instead, about $275 an hour on average. If an agency's account team spends 40 hours a month in total on a similar task, that is $11,000 a month in resource time at the agency rate.
This is how Reclaim Creative began. Our founder, Daniela Valencia, was spending 14 hours every week on social media wireframes for 12 clients at the marketing agency she worked in. Manual, repetitive and completely automatable. She brought it down to 10 minutes. The win came from knowing the task inside out, not from knowing the most tools.
Leadership guardrails
The audit also lets the leadership team align on the main guardrails, the do's and don'ts for how the team operates with AI. Without these decisions, staff training loses momentum. People wonder whether they can use AI on client work, which tool to use and who to ask when something goes wrong. If nobody answers, most people quietly stop.
The Australian Government's Department of Industry, Science and Resources, Guidance for AI Adoption starts with deciding who is accountable for AI use in an organisation. The earlier Department of Industry, Science and Resources, Voluntary AI Safety Standard places internal capability inside its first guardrail on accountability. Both are free to read and worth an hour of a leader's time.
For a 5 to 50 person business, the do's and don'ts can fit on one page: approved tools, data that stays out, where human review is required and who to ask. Name one owner for adoption, ideally a senior person who uses AI in their own work.
Leadership involvement also surfaces the bigger opportunities. After a full-day AI workshop, one marketing agency's leadership team identified an estimated $600,000 a year in resource time, and $950,400 a year projected across the agency. Reclaim Creative helped create AI workflows that solved those problems.
If you are unsure what belongs on your list, read how to decide what to automate in your marketing team, and what not to. For the wider picture, see Is my business ready for AI?
What goes into a 90-day plan for AI training and the start of implementation?
A 90-day plan maps the first 90 days: the training, the ChatGPT or Claude setup and the start of implementation. It turns the audit's findings into an order of work.
It sets out which workflows the team trains on first, which guardrails apply from day one and which manual tasks move into implementation first. Starting with the tasks the audit identified as the biggest bottlenecks keeps the training tied to the work that costs the team the most resource time.
Implementation runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days.
What should a 1 to 2 day AI training cover?
A 1 to 2 day AI training, depending on time and team size, should be built around the team's real processes, so people leave with AI workflows they use, not a list of tools.
Train role by role. A social media strategist, an account manager and an operations lead repeat different tasks, so they need different workflows. In a mixed session, each person should build at least one workflow on a task from their own week, using real briefs with sensitive details removed.
An in-house team example. Imagine a six-person marketing team inside a service business that spends hours each month turning sales notes into case studies and newsletter content. Hands-on training starts with a shared context document: brand voice, audience and approved claims. Then each person builds the workflow for their part of the job, with a checking step before anything is published.
An agency example. Account managers, strategists and creatives each bring a task they repeat across many clients. The training builds those workflows live, then sets the review points where a senior person signs off. Brand integrity stays in human hands.
The results can be large. During one workshop, a 42 hour a month marketing task at an agency was automated in 20 minutes. After another team workshop, an agency got 200 hours a month back immediately, an estimated $50,000 a month in resource time. Reclaim Creative helped create AI workflows that solved those problems.
How should a team be trained on ChatGPT or Claude?
Training on ChatGPT or Claude means setting everyone up and making sure they understand the rules and guardrails, so everyone uses it in the same way and works collaboratively.
The rules and guardrails everyone learns sit alongside the do's and don'ts leadership agreed in the audit. They cover three foundations: how to give AI the right context, how to check what it produces, and what information must never go into an AI tool.
Context
Most generic output is a context problem, not a tool problem. Ask Claude for a social caption with no brief, no brand voice and no audience, and you get an average of the internet. Teams need to brief AI the way they would brief a new junior: who it is for, what good looks like, examples, constraints and tone. We go deeper in why your AI output sounds generic.
Checking
AI writes confident drafts, and some of them are wrong. Teach people to check facts, numbers, claims and brand fit before anything reaches a client or customer. That habit is the line between useful drafts and AI slop. A person stays accountable for every piece of work that leaves the business.
Data safety
Staff need clear rules on what stays out of AI tools. The OAIC, Guidance on privacy and the use of commercially available AI products recommends that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools. For an agency, apply the same caution to client data under NDA, unreleased campaigns and commercial terms.
This is where DIY tends to stall. Subscriptions and tutorials get people using AI. They rarely get people using it well on their actual work, or using it the same way as their colleagues, because nobody has taught the fundamentals in the context of that work.
What does AI implementation look like after training?
In implementation, an AI specialist helps build AI workflows that can help teams save 20+ hours a week. This ranges from 2 to 6 months of engagement.
A workshop creates energy. The weeks after it decide whether that energy turns into practice. Put the new workflows where people already work: the project brief template, the client onboarding checklist, the weekly content process. Review them in team meetings. Ask each person which workflow saved time and which one they dropped, then fix or retire the ones that do not work.
This is also where custom builds make sense. Once a team understands where AI fits, a valuable process can be built into an AI agent around the business context. In our view, training comes first, because a team that knows the fundamentals briefs better, checks better and gets more from anything built later.
A proposal automation built in a founder session addressed an estimated $100,000 a year in resource time. Daniela Valencia of Reclaim Creative helped create AI workflows that solved those problems.
What can your team do this week to start AI training well?
You can start this week with a small version of the audit: find your team's main bottlenecks and put a cost on them.
- Ask each person to list the three tasks that take the most repetitive time each month.
- Put a number on each one using staff cost or your agency rate.
- Ask leadership to agree the do's and don'ts on one page: approved tools, data that stays out and where review is required.
- Pick one workflow and write a proper context brief for it before you prompt anything in ChatGPT or Claude.
- Name one person to own adoption.
Our AI-ready agency checklist and AI resources help with the audit. DIY will get you this far. Teams usually stall when they try to turn the list into reliable workflows across every role, and keep them in use. That is the part a 90-day plan, professional training and implementation are for.
How Reclaim Creative helps
Reclaim Creative follows the same path. It starts with an AI audit: Daniela Valencia identifies your team's main AI opportunities and bottlenecks, the cost of each manual task and the potential savings in resource time, and helps your leadership team align on the guardrails. Then comes a 90-day plan that maps the first 90 days: a 1 to 2 day AI training built around your team's real processes through our AI Team Workshops, the ChatGPT or Claude setup so everyone works with the same rules, and the start of implementation. Implementation runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days. During implementation, an AI specialist helps build AI workflows with your team. For agencies, this is our Agency AI Transformation, working with project leads and account managers. You work directly with Daniela Valencia, a marketer with 6+ years of in-house and agency experience, so the training respects the craft it is changing. Start with a complimentary Intro Call: see our services and get in touch.
Frequently asked questions
What training is recommended for teams adopting AI?
It depends on the size of the team and its main bottleneck, meaning the manual tasks the business is trying to address. Start with an AI audit, then follow a 90-day plan that maps the first 90 days: a 1 to 2 day AI training on real processes, the ChatGPT or Claude setup and the start of implementation. Implementation runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days.
What does an AI audit look at?
An AI audit looks at the team's main AI opportunities and main bottlenecks, and puts a cost on each manual task with the potential savings in resource time. It also lets the leadership team align on the main guardrails, the do's and don'ts for how the team operates with AI.
How long should AI training for a team take?
We recommend a 1 to 2 day AI training, depending on time and team size, built around the team's real processes. Implementation with an AI specialist starts during the first 90 days and runs over 2 to 6 months of engagement, so for some teams it continues beyond the first 90 days.
Does everyone on the team need the same AI training?
Everyone needs the same setup on ChatGPT or Claude and the same rules and guardrails: giving AI the right context, checking its output and keeping sensitive data out of AI tools. After that, training should split by role, because a strategist, an account manager and an operations lead repeat different tasks. Leaders align on guardrails and ownership of adoption during the audit.
What information should staff never put into an AI tool?
The OAIC recommends that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools. Agencies should also keep client data under NDA, unreleased campaigns and commercial terms out of these tools. A one-page AI use policy makes the rules clear to staff.
Can a team learn AI from tutorials and ChatGPT subscriptions alone?
Tutorials and subscriptions get people using AI, which is a useful start. They rarely get people using it well on their own work, because the fundamentals of context and checking are skipped. That is why DIY output often stays inconsistent and generic.
Is AI training about reducing headcount?
Good AI training is about capability, not headcount. It helps people spend less time on repetitive tasks and more on strategy, creative work and clients. Reclaim Creative frames every programme around making people better at their jobs.
What should leaders decide before rolling out AI training?
Leaders should agree the guardrails, the do's and don'ts for how the team uses AI, and name one person who owns adoption. The AI audit is the place to make these decisions, before training starts. The Australian Government's Guidance for AI Adoption starts with deciding who is accountable.
How do you make AI training stick after a workshop?
Implementation makes training stick. An AI specialist helps build the new workflows into the templates and processes the team already uses, over 2 to 6 months of engagement. Keep the workflows that save time and fix or retire the ones that do not.
Should agencies and in-house teams train differently on AI?
The path is the same, but the workflows differ. Agencies train on tasks repeated across many clients, such as briefs and content plans, with senior sign-off to protect each client's brand. In-house teams train on their own brand's processes and usually measure gains in staff cost rather than rate-card time.
