AI enablement, or how to get your marketing team past "I tried it once"

This is for anyone running a marketing team where two people use AI all day and everyone else tried it once. AI enablement means putting AI inside a weekly job, set up with your brand rules, then training people on their own work.

The enthusiasts are brilliant. I’m one of them. But to everyone else they can look like they know a secret, and meanwhile there’s a campaign that has to go out on Friday.

The best example I have of a team getting past the early adopters is TCS’s social team. Our case study counts more than 80 posts a month from them with a tool we built and trained on their brand.

Why does AI stay with two enthusiasts?

For most people AI is still a blank chat box. It hasn’t been told the team’s rules, and people haven’t been given time to learn it, so the enthusiasts fill those gaps in their own time. Everyone else waits for someone to set it up for them.

BCG surveyed 11,749 workers across 14 markets this year (BCG press release ). 74% of frontline employees now use AI every day or a few times a week, but only a third say leadership’s communications about AI are clear, and across everyone surveyed only 36% feel they’ve had adequate upskilling (BCG ). Back in 2024, Microsoft and LinkedIn found 78% of AI users were bringing their own AI tools to work (Microsoft Work Trend Index ).

Your two enthusiasts are probably in that 78%. They found a tool and worked it out in their evenings, and now they’re quicker than everyone else. The rest of the team never got the same head start.

When I look at teams where AI hasn’t spread, the same four things are usually missing:

• The tool doesn’t know your rules. Every prompt starts from nothing, so it has never heard of your brand voice or which clients you can’t name. The output sounds generic, someone has to fix it, and fixing it takes longer than writing it.
• There’s no job attached. “Use AI more” isn’t a task anyone can do on a Tuesday.
• Nobody’s been shown. People get a licence and a link, but never an afternoon on their own work.
• The boss doesn’t use it, or uses it quietly, so it never looks like part of the job.

What does AI enablement mean for a marketing team?

AI enablement for a marketing team is the work of getting AI into everyone’s normal week. It has four parts. First, pick the jobs where AI will help, such as first drafts of social posts or turning a webinar into copy. Second, set the tools up with the team’s own brand voice, tone and rules, so the first draft already sounds like the company. Third, put those tools where people already work, so nobody has to remember to open something new. Fourth, train people on their own briefs and channels, and keep a person checking the output before anything is published. AI sales enablement is a separate job, using AI to get the right content and data to a sales team (we’ve written about how marketing teams can use AI to help sales close deals ).

How did TCS’s social team make AI part of the week?

They started from work they already had, with a tool that was trained on their brand. TCS’s CMO, Ashish Babu, wanted to see what AI could do for the efficiency and creativity of their content marketing, and asked us. We began with workshops and demos on “The art of the possible”. Then Suhail Adam, TCS’s Head of Social, asked us to help his team with content creation.

So we set his team up on Compass Agents, the AI product we built for enterprise marketing teams, and trained it on the TCS brand and tone of voice. By the time we wrote the case study, the social team relied on it for social posts and for turning video into copy quickly. It was creating more than 80 posts a month, saving more than 60 hours of production time a month, and getting them from idea to posting three times faster (case study ).

Compass Agents has significantly enhanced my control over the social post creation process. Whether I need to plan well in advance or make last-minute adjustments, Compass Agents enables me to surpass our objectives.

Anmol Patel Social Media & Insights Manager, TCS

What I take from it for any marketing team:

• The team asked for it. Suhail came to us with his own brief after the workshops.
• The brand went into the tool. We trained it on the TCS brand and tone of voice when we set the team up.
• It went into work they already did, their social content and turning video into copy.

It wasn’t perfect at the start. In our 2024 interview, Anmol mentioned “occasional challenges with nuance and complexity” and some repetition in the posts during the London Marathon. Anmol’s answer was to get better at asking: “It’s about training employees to ’talk to’ these tools” (interview ).

Seven ways to get the whole team using AI

Get one job working for everyone before you add a second. These are the seven things I’d do, in roughly this order.

1. Pick one job. For TCS’s social team the job was social content, so that’s what the tool does. Choose something your team does every week and moans about, with an obvious before and after. Adapting one campaign for six channels is a good one.
2. Put your rules in the tool. Write down what usually lives in people’s heads, like banned words and which clients you can name. We started writing “author files” for everyone we write as, and they grew to 400 lines (how to build an AI writing style ). Once the first draft sounds like you, people stop treating AI as extra work.
3. Put it where the team already works. If people have to remember to open a new tab, most of them won’t. Our resourcing app, the one that books everyone’s time, is connected to Claude, so I can ask for next week’s plan in one line. Claude even has its own Teams account at Fifty Five and Five now, so it can talk to people of its own accord.
4. Train people on their own work, and give it real hours. BCG found in 2025 that regular use is “sharply higher” for employees who get at least five hours of training and have access to in-person training and coaching (BCG ). Spend those hours on your own briefs and brand. In Microsoft’s 2024 research only 39% of people using AI at work had been trained by their company.
5. Use it where people can see. In BCG’s 2025 research, the share of employees who feel positive about generative AI rose from 15% to 55% with strong leadership support. Share your own prompts , the bad ones too. When Claude Opus 5.5 came out in September I told the team it feels like talking to a normal person again.
6. Turn the enthusiasts into builders. Your two fans are the obvious people to build what everyone else uses. Barnaby Ellis, our Head of Strategy, built a playable game for a client pitch in a day with Claude Code. Then Owen Steer tested the gameplay and Fergus Hannant, one of our designers, turned it into a polished demo in under two hours (Barnaby’s write-up ).
7. Keep a person on the last mile, and count the hours. AI does the first draft and a person checks it before it goes anywhere. We build AI tools with a person checking the last mile, because people trust output they know someone has checked. Then measure something the team cares about, like the posts and hours saved in our TCS case study.

What did we get wrong with our own team?

We asked a lot of people to change quickly, and not everyone came with us. I wrote about this earlier in the year: there was staff turnover, because not everyone got it, and not everyone wanted to. Some people had built good careers on the old way of working, and I was asking them to learn to build things in a completely different way (what an AI-powered marketing agency actually looks like ).

I also talked about the future before we were living it. In June 2024 I sent a newsletter called “The Marketing Agency is Dead”, and looking back it was about six months ahead of where we actually were.

What worked was much less exciting. We wrote our rules into the author files and connected AI to tools we already used, and people started building things for each other, like Barnaby’s game. Now people on the team run AI agents overnight on research and analysis, and pick up the results in the morning.

So how do you get a whole marketing team using AI?

Pick one job the team does every week, and teach the tool your brand and your rules. Put it where people already work, train them on their own briefs for proper hours, and use it yourself where they can see. Keep a person checking the last mile and count the hours it saves.

If you’ve got two enthusiasts, ask them which weekly job they’d hand over first, and make that the first piece of AI enablement for everyone. If you’d like help building it on your own brand and rules, get in touch.

Frequently asked questions