What a Fast-Moving Brand Can Do With an Intelligent Marketing System
Chris Reynolds, CEO at Thrudark
I'll be honest: I didn't expect to be writing this.
When someone from DOJO pitched me on a marketing AI platform, my initial reaction was the same one most people in my position have. We've tried the tools. We've read the case studies. And usually, a few weeks in, you realise the tool either does one narrow thing well, or does ten things passably, and none of them quite fit how your business actually operates.
DOJO is different. Not because it does everything (it doesn't), but because the things it does, it does without needing you to hold its hand through each one.
Here's what we've actually used it for. Not a brochure list. The real stuff.
The broader picture first
Yes, it's becoming a cliché. AI tools are changing the business landscape and the brands that don't react fast enough will lose ground to the ones that do. I'm not saying anything new there.
But I think most people are still reaching for the wrong tools. What's out there tends to be modular and task-specific: one tool for social copy, another for data, another for SEO. Each does its one job reasonably well. None of them know your business. The hidden cost of that fragmentation is something most marketing leaders underestimate until they've lived it.
DOJO works differently. It's an intelligent marketing system, connected to your data sources and carrying the business context built up across every conversation, from every team member who has used it. It's as comfortable planning a social campaign as it is producing a P&L summary for the board. The same platform that helped us build a Marketing Mix Model across £40m of historical revenue also writes product copy that sounds like ThruDark, because it's learned what ThruDark sounds like. That breadth, grounded in genuine context, is what separates it from a collection of specialist tools stitched together.
It reminds me of the big data revolution. A decade ago, the brands that invested early in data infrastructure pulled away from the ones that didn't.
I actually spent five years early in my career implementing ERP systems, and I remember clearly what happened when businesses pooled all their departmental data into one connected platform for the first time. The efficiencies available were significant. But not every client truly capitalised on them. Some resisted the change, adapted slowly, and watched competitors move faster. The technology was never the limiting factor. The willingness to commit to it was.
This is the same shift. I think of it as the Big Resource Revolution. It isn't an abundance of data this time: it's resource. Every team in the business can now do more with less. Not marginally more. Materially more.
We are hiring less. I'll say that plainly. The team we have is more capable than an equivalent headcount would have been two years ago, because the production work — the analysis, the reporting, the drafting — is increasingly handled by the tools. That's not a comfortable thing for everyone to hear, but pretending it isn't happening doesn't help anyone.
What I will say is that the fear among the team has been far less than I expected. Some of the most sceptical people have been the quickest converts once they saw what the platform could actually do. When the tool saves you four hours on a report you were going to spend a Friday afternoon building, the scepticism tends to dissolve.
Daily trading during a promo
ThruDark runs occasional promotional events called Strike Ops. A few days, deep discounts across the range, significant paid media investment. It's the kind of period where you need to know what's happening every single day: new customer revenue, repeat customer revenue, paid efficiency against our NCR target, market-by-market splits, product-level performance, cohort profitability through to CM3.
Previously, building that picture took hours of manual work in spreadsheets.
Now I upload the day's data files and DOJO builds the full report: interactive charts, YoY day-aligned comparisons, efficiency flags, market breakdowns, anomaly callouts, in the time it would have taken me to open Excel. The March 2026 Strike Ops report alone had seven charts, a full product performance table across the top 20 SKUs, paid efficiency recovery modelling for the final four days of the promo, and a strategic framework section separating tactical optimisation actions from structural market decisions.
That's not AI saving me a bit of time. That's AI doing a job that previously required a dedicated analyst.
And it doesn't stop at the numbers. The part that actually changes decisions is the contextual layer on top: why a market is underperforming, whether an efficiency breach is structural or tactical, which products are genuinely in decline versus which are being cannibalised by a newer line. A dashboard tells you what happened. DOJO tells you what it means and what to do about it. That distinction matters more than any chart.
Marketing Mix Modelling from scratch
This one surprised me most.
We'd wanted to run a proper Marketing Mix Model for a long time. Budget had always been the obstacle. External agencies charge significant fees for this work, and I wasn't convinced the output would be actionable at our scale.
DOJO built the model in-house. Saturation OLS, 128 weeks of data, £40m+ in revenue modelled across Meta, Google, organic search, email, and promotional events. The final model hit an R² of 0.914 on held-out test data. That means the model explains 91% of our weekly revenue variance, which is the difference between a useful planning tool and an expensive spreadsheet.
More importantly, the output changed how we think about budget allocation. We can now see, with reasonable confidence, what each channel actually contributes versus what it's claiming credit for in last-click attribution. Those are very different numbers, and the gap between them has real commercial implications.
AI search visibility — before most brands are paying attention
DOJO has SEO and AEO (Answer Engine Optimisation) capabilities, and one of our analysts ran a citation gap audit looking at how ThruDark appears — or doesn't — in AI answer engine responses. Tools like ChatGPT, Perplexity, and Gemini are changing how people discover brands. The audit identified specific content gaps where competitors were being cited and we weren't, along with a prioritised list of content topics to address.
This isn't a future problem. It's happening now, and most brands haven't started. The race to be cited, recommended and visible in AI search is already underway. Having the capability to audit, identify the gaps and act on them immediately is exactly the kind of edge that compounds. We're moving on this while others are still debating whether it matters.
Monthly reporting that doesn't consume the team
Executive summaries. Monthly trading reports. Promo wrap-ups. Brand awareness measurement. Media monitoring. Social listening. Allies programme reporting.
All of it runs through DOJO now. The playbooks are saved and reused. The data formats are understood. When I need a March exec summary, I don't brief an analyst and wait three days. DOJO pulls the data from connected systems, flags any gaps and prompts me to provide what's missing, then delivers a structured report back in the same session, formatted to our own reporting conventions, with the right revenue framework, the right YoY methodology, and the right period splits.
The consistency that comes from having one system that knows your reporting standards is something I hadn't anticipated valuing as much as I do.
Where we are right now, and where it's going
I want to be transparent about the current state, because I think honesty serves everyone better than a polished case study.
Not every integration is in place yet. Some data still gets uploaded manually because the direct API connections aren't there. That's a real friction point. The upside is that the data sits in a shared library, so the whole team benefits from each upload, not just the person who ran the report. It's collaborative in a way that a spreadsheet saved to someone's desktop never was.
When DOJO closes the remaining integrations — and they're on the roadmap for this year — the manual step disappears. The platform becomes significantly more capable, not because the analysis changes, but because the data flows in automatically and the team can stop spending any time at all on data management.
The value we're already getting, before the full integration stack is even in place, is genuinely significant. And we're nowhere near the ceiling. That's a reasonable description of where most AI tools are in 2026. The question isn't whether to wait for the perfect version. It's whether you're building the capability, the habits, and the institutional knowledge now, so that when the integrations close, you can make the commercial most of them immediately.
The brands that are halfway through their learning curve when the tools mature will be the ones that pull ahead. The ones still watching from the sideline will be catching up for years.
That's the version of this story I think actually matters.
I'll be honest: I didn't expect to be writing this.
When someone from DOJO pitched me on a marketing AI platform, my initial reaction was the same one most people in my position have. We've tried the tools. We've read the case studies. And usually, a few weeks in, you realise the tool either does one narrow thing well, or does ten things passably, and none of them quite fit how your business actually operates.
DOJO is different. Not because it does everything (it doesn't), but because the things it does, it does without needing you to hold its hand through each one.
Here's what we've actually used it for. Not a brochure list. The real stuff.
The broader picture first
Yes, it's becoming a cliché. AI tools are changing the business landscape and the brands that don't react fast enough will lose ground to the ones that do. I'm not saying anything new there.
But I think most people are still reaching for the wrong tools. What's out there tends to be modular and task-specific: one tool for social copy, another for data, another for SEO. Each does its one job reasonably well. None of them know your business. The hidden cost of that fragmentation is something most marketing leaders underestimate until they've lived it.
DOJO works differently. It's an intelligent marketing system, connected to your data sources and carrying the business context built up across every conversation, from every team member who has used it. It's as comfortable planning a social campaign as it is producing a P&L summary for the board. The same platform that helped us build a Marketing Mix Model across £40m of historical revenue also writes product copy that sounds like ThruDark, because it's learned what ThruDark sounds like. That breadth, grounded in genuine context, is what separates it from a collection of specialist tools stitched together.
It reminds me of the big data revolution. A decade ago, the brands that invested early in data infrastructure pulled away from the ones that didn't.
I actually spent five years early in my career implementing ERP systems, and I remember clearly what happened when businesses pooled all their departmental data into one connected platform for the first time. The efficiencies available were significant. But not every client truly capitalised on them. Some resisted the change, adapted slowly, and watched competitors move faster. The technology was never the limiting factor. The willingness to commit to it was.
This is the same shift. I think of it as the Big Resource Revolution. It isn't an abundance of data this time: it's resource. Every team in the business can now do more with less. Not marginally more. Materially more.
We are hiring less. I'll say that plainly. The team we have is more capable than an equivalent headcount would have been two years ago, because the production work — the analysis, the reporting, the drafting — is increasingly handled by the tools. That's not a comfortable thing for everyone to hear, but pretending it isn't happening doesn't help anyone.
What I will say is that the fear among the team has been far less than I expected. Some of the most sceptical people have been the quickest converts once they saw what the platform could actually do. When the tool saves you four hours on a report you were going to spend a Friday afternoon building, the scepticism tends to dissolve.
Daily trading during a promo
ThruDark runs occasional promotional events called Strike Ops. A few days, deep discounts across the range, significant paid media investment. It's the kind of period where you need to know what's happening every single day: new customer revenue, repeat customer revenue, paid efficiency against our NCR target, market-by-market splits, product-level performance, cohort profitability through to CM3.
Previously, building that picture took hours of manual work in spreadsheets.
Now I upload the day's data files and DOJO builds the full report: interactive charts, YoY day-aligned comparisons, efficiency flags, market breakdowns, anomaly callouts, in the time it would have taken me to open Excel. The March 2026 Strike Ops report alone had seven charts, a full product performance table across the top 20 SKUs, paid efficiency recovery modelling for the final four days of the promo, and a strategic framework section separating tactical optimisation actions from structural market decisions.
That's not AI saving me a bit of time. That's AI doing a job that previously required a dedicated analyst.
And it doesn't stop at the numbers. The part that actually changes decisions is the contextual layer on top: why a market is underperforming, whether an efficiency breach is structural or tactical, which products are genuinely in decline versus which are being cannibalised by a newer line. A dashboard tells you what happened. DOJO tells you what it means and what to do about it. That distinction matters more than any chart.
Marketing Mix Modelling from scratch
This one surprised me most.
We'd wanted to run a proper Marketing Mix Model for a long time. Budget had always been the obstacle. External agencies charge significant fees for this work, and I wasn't convinced the output would be actionable at our scale.
DOJO built the model in-house. Saturation OLS, 128 weeks of data, £40m+ in revenue modelled across Meta, Google, organic search, email, and promotional events. The final model hit an R² of 0.914 on held-out test data. That means the model explains 91% of our weekly revenue variance, which is the difference between a useful planning tool and an expensive spreadsheet.
More importantly, the output changed how we think about budget allocation. We can now see, with reasonable confidence, what each channel actually contributes versus what it's claiming credit for in last-click attribution. Those are very different numbers, and the gap between them has real commercial implications.
AI search visibility — before most brands are paying attention
DOJO has SEO and AEO (Answer Engine Optimisation) capabilities, and one of our analysts ran a citation gap audit looking at how ThruDark appears — or doesn't — in AI answer engine responses. Tools like ChatGPT, Perplexity, and Gemini are changing how people discover brands. The audit identified specific content gaps where competitors were being cited and we weren't, along with a prioritised list of content topics to address.
This isn't a future problem. It's happening now, and most brands haven't started. The race to be cited, recommended and visible in AI search is already underway. Having the capability to audit, identify the gaps and act on them immediately is exactly the kind of edge that compounds. We're moving on this while others are still debating whether it matters.
Monthly reporting that doesn't consume the team
Executive summaries. Monthly trading reports. Promo wrap-ups. Brand awareness measurement. Media monitoring. Social listening. Allies programme reporting.
All of it runs through DOJO now. The playbooks are saved and reused. The data formats are understood. When I need a March exec summary, I don't brief an analyst and wait three days. DOJO pulls the data from connected systems, flags any gaps and prompts me to provide what's missing, then delivers a structured report back in the same session, formatted to our own reporting conventions, with the right revenue framework, the right YoY methodology, and the right period splits.
The consistency that comes from having one system that knows your reporting standards is something I hadn't anticipated valuing as much as I do.
Where we are right now, and where it's going
I want to be transparent about the current state, because I think honesty serves everyone better than a polished case study.
Not every integration is in place yet. Some data still gets uploaded manually because the direct API connections aren't there. That's a real friction point. The upside is that the data sits in a shared library, so the whole team benefits from each upload, not just the person who ran the report. It's collaborative in a way that a spreadsheet saved to someone's desktop never was.
When DOJO closes the remaining integrations — and they're on the roadmap for this year — the manual step disappears. The platform becomes significantly more capable, not because the analysis changes, but because the data flows in automatically and the team can stop spending any time at all on data management.
The value we're already getting, before the full integration stack is even in place, is genuinely significant. And we're nowhere near the ceiling. That's a reasonable description of where most AI tools are in 2026. The question isn't whether to wait for the perfect version. It's whether you're building the capability, the habits, and the institutional knowledge now, so that when the integrations close, you can make the commercial most of them immediately.
The brands that are halfway through their learning curve when the tools mature will be the ones that pull ahead. The ones still watching from the sideline will be catching up for years.
That's the version of this story I think actually matters.