Three months at ActionRocket: new tools and an AI colleague

Three months into life at ActionRocket, I’ve clocked 3,386 emails, made peace with Asana, acquired an AI colleague called Alan, and sweated through one very warm work trip. So far, so good. When I joined back in June, I had a pretty good idea of what I was walking into. We’d been open during the interview process about the culture, ways of working, and our thoughts on AI. But there are some things you only really learn once you’re in it. Most importantly, there’s still no P45 in sight. Always a good start.

Good people make good work better

One of the biggest highlights so far has been the people, I get to work with clients who are collaborative, transparent, and open to trying new things. There’s trust to challenge ideas, suggest different approaches, and work things out together, rather than simply doing what’s always been done. And then there’s ActionRocket itself. We’re a small independent agency, which means there’s nowhere to hide. Conveniently it also means there’s plenty of opportunity to get involved, share an opinion, and have a real say in how things work. Three months in, I already feel like I can make an impact. No pressure, then.

Getting comfortable with being uncomfortable

I’ve worked at a few agencies over the last 10 years, they’ve all been different but the goal is usually pretty similar: do great work for great clients. What changes is how you get there. One of my first challenges at ActionRocket was getting to grips with new tools, particularly our project management tool which we use to build plans and allocate work.

I’ll admit, the first few weeks were a lot. There was a new platform to understand, new processes to learn, and the small matter of working out how I could use it all to support both myself and the wider team. There’s still plenty I don’t know, and I’m actually quite happy about that. As the tools and platforms we use keep evolving, so will the way I work with them. I can already see the potential, and getting better at using them feels less like something to ‘complete’ and more like part of the job.

Turns out I have an AI colleague

And then there’s AI. Somewhere in the last three months, I’ve gone from being slightly sceptical about it to having an AI teammate who supports us with a few tasks. His name is Alan. Alan is great. We love Alan. Alan is our friend.

The more I’ve seen how the team works with AI, the more I’ve realised that it isn’t about adding another shiny tool for the sake of it. It can become part of the way work actually gets planned and delivered – helping with project setup, campaign planning, and spotting opportunities to make the next project run even better. Sally our Operations Manager, recently took a deeper dive into exactly that with Asana, including Alan and a couple of his AI teammates. Watch Sally’s on-demand session: The guide to faster campaigns: A live look at AI-powered marketing execution.

Having the chance to actually use AI as part of my day-to-day work has changed the way I think about it. I’m learning where it can help, where it fits into our processes, and like everything else, how to get the best from it. I’m okay with this new working relationship, provided Alan doesn’t go full Skynet on me.

Busy periods teach you a lot

August has been very warm, very busy, and a useful reminder that agency life doesn’t always arrive in a neat, perfectly organised package. One thing I’ve really appreciated at ActionRocket is that we make time to reflect after projects: what worked, what didn’t, and what we could do differently next time. It sounds simple, but it means there’s always an opportunity to learn and improve rather than immediately charging into the next thing.

And sometimes, those lessons are slightly less strategic. Picture the scene: a client trip to Switzerland, trying to look reasonably presentable, and then the heat happened. Thankfully, I’d packed an overshirt, which quickly went from ‘nice extra layer’ to ‘essential crisis-management equipment’.

So, three months in: good people, interesting work, plenty still to learn, and a newfound respect for strategic layering. I’ll take that.