My role covers e-commerce and digital marketing for a size-inclusive fashion brand we sell direct-to-consumer and through ASOS, Selfridges, Fenwick, and Harvey Nichols. That mix of agencies and in-house channels comes with an account structure that grew more complex as we scaled: two Google Ads accounts split by region, three Meta accounts split by currency, and GA4, Search Console, and Shopify sitting on the side, with GBP and USD figures I couldn't just add together. A lot of data, spread across a lot of platforms, none of it talking to each other.

For years I held that together in Looker Studio, rebuilding the dashboard by hand every time someone asked a new question. It cost me hours every week, and I still didn't fully trust what I was presenting when I was done.

Handing off the "why," not just the "what"

The biggest shift for me wasn't a report I read, it was an instruction I gave once. A drop or a spike that someone was eventually going to ask about used to sit unexplained for a day or two while I found the time to dig into it, and a real problem left unexplained that long can turn into lost revenue before anyone notices. So I told DOJO: don't wait for me, go find out why and tell me what happened. It took that and ran with it. The next time a channel dropped, it had already traced the cause, worked out we'd been pushing sale through that channel the week before, so the drop was a natural pull-back rather than a real problem, and had it summarized and ready before anyone even asked.

That's the part I push back on when the DOJO team describes DOJO as a "junior marketing executive." The way I use it, it's operating a level above that: it's carrying a standing responsibility I handed it, not waiting around for me to have time to ask a question.

Time that used to go into firefighting now goes into decisions that actually move the business.

The same shift shows up outside paid media, too. Manual product analysis for our merchandising team that used to take a couple of hours in Excel now takes a few minutes, roughly a 90% cut in the time I spend on it, the last time I ran the same analysis myself, with a bit of refinement.

Built to be honest, not agreeable

None of that would carry weight if I couldn't trust what DOJO was telling me, which is why the one thing I asked for when I set it up was a personality: stop being so agreeable, just be honest, and correct me when I'm wrong. It delivered, to the point that it's genuinely a bit relentless. I'd forgotten a report was running in the background, went to update it with current data, and it told me, plainly, that this was the fifth week running the same issue still hadn't been fixed. It caught me off guard the first time, but it's exactly the tone I wanted: something that keeps score instead of just being agreeable.

That directness paid off on something with real budget implications. In its very first weekly review, DOJO flagged that one of our brand campaigns was running well under Google's own recommended budget, and instead of mentioning it once and moving on, it kept surfacing the same finding, week after week, for six weeks straight. Most tools would have raised it once and let it sit unread in a report nobody reopened. DOJO kept it in front of me until I actually acted on it, which is the difference between a dashboard and something that actually changes what you do.

My colleague Sophie, our Senior Ecommerce Manager, picks up the story from the content and site-experience side:

A year of backlog, turned into a plan

I came back from a year of maternity leave to a site that had moved on without me: new pages added, old ones left untouched, and nobody quite sure which ones still mattered. We're also mid-migration to a new Shopify theme, so working through a year of backlog by hand wasn't realistic.

I ran a full SEO audit with DOJO first. Some of it we already knew from previous audits, but there was a lot it highlighted that I could just tidy up, roughly 40 things I could act on directly, on top of dead pages with no traffic and pages quietly competing against each other for the same terms. That mix of familiar and new is what built my trust in it. I used the same process to simplify our top and bottom nav, cutting duplicate pages we'd meant to clean up for ages.

It's become a loop: I update a collection page or a piece of content, then ask DOJO what else is worth optimizing. I've also had it build a UX audit and an A/B test plan, and I'm running some of those tests now.

GEO is the part I couldn't have tackled alone. I don't know how to optimize for how AI engines read us, so I asked DOJO to help, and it did.

It's useful for smaller questions too. A teammate asked if our knitwear pages were outperforming wedding guest yet. DOJO said not yet, and told me to check back in early-to-mid September, rather than just reporting where we stood that day.

Back to Fabio, on why that groundwork matters:

Generic AI got me halfway there

None of that persistence or context happens by accident, and it isn't something I take for granted, because I've used the alternative. I'm not new to AI. I'd spent two or three years running ChatGPT, Claude, and Gemini for analysis before DOJO, so I know their capabilities and their limits. What I found myself doing was still manual: dumping data in, asking for recommendations, getting a nice structured answer back that was missing context. I'd have to be very specific and feed it extra information every time. It helped, but it stayed time-consuming.

DOJO was different from day one, because it already understood marketing and e-commerce as its backbone. I don't have to explain myself. I can use acronyms, use any shorthand I want, and it's straight on point. When your numbers are split across regions, currencies, and channels the way ours are, that saved context is the difference between a tool you have to manage and one you can actually rely on to manage that complexity for you.