AI in Digital Marketing
Six live sessions that leave your marketing team genuinely able to use AI, rather than aware that it exists.
Most AI training is a tour of tools. This is a working method. Your team learns one prompting framework, then applies it across every stage of the content process, in your own AI platform, on your own real work. Between sessions they run experiments and report back to each other, which is the part that makes it stick.
Delivered in Microsoft 365 Copilot or Google Workspace with Gemini, depending on what your business actually runs.
Two frameworks, applied together
Everything in the programme hangs off two simple structures. Teach those properly and the tools become interchangeable, which matters when the tools change every quarter.
The content creation workflow
Ideation, research, composition, editing, distribution. Five stages your team already works through, whether or not they name them. The course maps AI onto each stage rather than treating AI as a separate activity that happens somewhere else.
CRAFT
Context, Role, Action, Format, Tone. A five-part structure for writing a prompt that gets a usable answer first time. It is taught in session three and then deliberately applied again in every session that follows: CRAFT for ideation, CRAFT for research, CRAFT for composition, CRAFT for editing, CRAFT for distribution.
That repetition is the design. People do not retain a framework they meet once. They retain one they use five times on their own work.
The six sessions
1. AI in digital marketing: an overview
Where AI actually fits in a marketing team’s week. Tools by function, how roles are changing, an introduction to the content creation workflow, and an honest look at ethics, hallucination and privacy. Five short activities, plus a pre-session questionnaire so the rest of the programme is shaped around what your team already worries about.
Between sessions: test one AI tool and one Copilot or Gemini feature. Record a hypothesis, test it, report back.
2. Experiments: show and tell
The team demonstrates what they tried, what worked, what failed and what surprised them. Peer learning does more here than instruction, because the use cases come from inside the business. Followed by live demos of the AI features inside the tools they already use daily.
3. Prompt engineering: the foundation
How language models actually interpret what you type, why vague prompts produce vague output, and the CRAFT framework. Hands-on: take a weak prompt, rebuild it with CRAFT, compare the outputs side by side. This is the session people cite afterwards as the one that changed how they work.
Between sessions: rewrite one of your own recurring prompts using CRAFT and document the before and after.
4. AI for ideation and research
Traditional ideation methods and where AI genuinely helps rather than just producing more options. Reverse brainstorming, theme expansion, competitor and trend research. Includes a group brainstorm sprint run against a real brief, comparing manual, AI-only and hybrid approaches.
Between sessions: take a group idea and research it properly using AI.
5. AI for composition, editing and distribution
Drafting without losing your brand voice, editing for structure and tone rather than just grammar, and turning one idea into content for several channels. Includes building a brand voice prompt your team can reuse, and live demos in your own suite.
Between sessions: take one piece of content and adapt it for three channels.
6. Tools, ethics and what comes next
The CRAFT framework revisited across all five workflow stages so the whole programme connects. A tour of the current tool landscape by category. Responsible use and human oversight, including a “would you post this?” exercise that tends to generate the most useful discussion of the whole course. Then where this is heading and how each person plans to use it.
What your team can do afterwards
- Write structured prompts that produce usable output on the first or second attempt
- Apply AI at each stage of the content process rather than only to first drafts
- Build and reuse a brand voice prompt so AI-assisted copy still sounds like you
- Turn one piece of content into several formats without starting again
- Use the AI features inside the tools they already have, which is usually the fastest available win
- Judge when not to use AI, and spot output that should not go out
Delivered in your stack
The frameworks are platform-agnostic. The demos are not, deliberately.
Teams running Microsoft 365 get Copilot demos in Word, Excel, PowerPoint and Teams. Teams running Google Workspace get the Gemini equivalents. Everyone works in ChatGPT alongside it, because the prompting principles transfer and it is useful to see the same task done in two places.
We have run this course in both. The structure does not change; the screenshots do.
Formats
| Format | What it covers | Best for |
|---|---|---|
| Full programme 6 x 60 minutes |
All six sessions, with activities and between-session experiments | Teams who need to actually change how they work |
| Foundations 2 x 60 minutes |
The overview and prompt engineering sessions, which carry CRAFT | Testing the appetite, or a team that only needs the fundamentals |
| Tailored | Sessions selected and reweighted around your team’s roles and tools | Teams with an existing baseline or a specific gap |
Sessions run weekly or fortnightly. Fortnightly works better, because the between-session experiments are where most of the learning actually happens and people need time to do them.
What is included
- Six live 60-minute sessions, recorded so people who miss one can catch up
- Slides, worksheets and activity templates
- A prompt library your team builds during the course and keeps
- The CRAFT prompt generator
- Pre-session questionnaire and between-session activities
- Email support between sessions
- An AI tool directory, kept current
Investment
[ INVESTMENT ] for a cohort, with additional places priced per head. Cohorts work best between eight and eighteen people, because the show and tell sessions need enough voices to be interesting and few enough that everyone gets a turn.
Get the course outline
The full session-by-session outline, activities and outcomes as a PDF, plus a sample from the prompt deck so you can see the standard before you commit anybody’s Tuesday mornings to it.
We will email you the outline. No sequence, no sales call unless you ask for one.
Questions we get asked
Does the team need paid AI subscriptions?
Free ChatGPT is enough to follow along, though a paid tier is better. Copilot and Gemini demos need whichever licence your business already holds, and Copilot in particular requires a subscription separate from base Microsoft 365 licensing. We will confirm what you need before we start rather than discovering it in session two.
Is this for marketing people only?
Mostly, but it travels further than you would expect. We have had photographers, video teams, salespeople and operations staff in cohorts and the content workflow applies to all of them. The examples stay marketing-led.
What if our team is at very different levels?
That is the normal case rather than the exception. The pre-session questionnaire tells us the spread, and the activities are structured so confident people go deeper while everybody else still completes the task. The show and tell format helps here too, because the advanced people end up teaching.
Will this be out of date in six months?
The tool demos will age. The frameworks will not, which is exactly why the course is built on them. CRAFT and the content workflow apply to whatever assistant your business is using in two years.
Can you run it for more than one team?
Yes, and we usually recommend separate cohorts rather than one large group. Above roughly eighteen people the discussion sessions stop working, and those carry more of the learning than the instruction does.
Do you tailor the examples to our business?
Yes. Activities run against your real briefs and your real content wherever possible. A course taught on invented examples produces people who can prompt about invented examples.
See some of our work
