- The Publisher Newsletter
- Posts
- How Mumsnet harnessed AI to become a data and insights powerhouse
How Mumsnet harnessed AI to become a data and insights powerhouse
The story of AI at Mumsnet is primarily a data-insights story. With users generating 1.5 million fresh words every day and a 25-year old archive of more than 6 billion words, the site is a treasure trove of consumer insights. AI has ‘turbo-charged’ Mumsnet’s ability to unlock that value for brands.
Welcome to the Publisher Newsletter by Media Voices: profiling the people and products powering publishing.
- ad -
Inside Leading Publisher’s Engagement Strategies
Learn how leading publishers are building reader habit with interactive formats. Real engagement case studies, practical takeaways — now compiled in one playbook.
How publishers are increasing reader engagement and build habit with interactive content
Practical examples of formats that encourage repeat visits
How interactive experiences support subscription growth
Actionable ideas you can apply quickly in your own newsroom or product strategy
How Mumsnet harnessed AI to become a data and insights powerhouse
The story of AI at Mumsnet is primarily a data-insights story. With users generating 1.5 million fresh words every day and a 25-year old archive of more than 6 billion words, the site is a treasure trove of consumer insights. AI has ‘turbo-charged’ Mumsnet’s ability to unlock that value for brands.
Before I spoke to Mumsnet CEO Sue Macmillan at last November’s Definitive AI Forum in London, I thought of Mumsnet as an old-school forum where parents went to find advice on teething and tantrums. I genuinely did not have ‘Mumsnet bosses AI’ on my 2025 bingo card.
However, on the stage at Stationers’ Hall, Sue quickly set me straight, enthusiastically telling me how the 25-year-old parenting forum is harnessing AI to power stronger pitches and deliver more actionable insights to brand partners. “I feel like AI has just changed everything,” she said, “and I think that is really exciting.”
If I had thought about it logically, Mumsnet is the perfect publisher use case for AI. Members of the parenting forum generate 1.5 million words every day and since it was founded in March 2000, it has grown to become a repository of more than 6.5 billion words. That’s quite the dataset.
Mumsnet hired its first data scientist nine years ago after a prospective hire told the management team that analysis of what had been posted in Mumsnet in the run up to Brexit would have predicted the result. “We said, ‘You’ve got the job!’” laughed Sue.
“People are not just talking about parenting. They’re talking about brands. They’re talking about products and politicians,” she explained. “We knew there was unbelievable data and insights in there. What Gen AI did was turbo-charge our efforts to turn that into something that could be really useful and actionable.”
Just six months after ChatGPT was launched, Mumsnet had developed the first proof-of-concept for MumsGPT, an internal chatbot which analyses all of the discussions on Mumsnet to determine what forum users think. Sue said: “It allows people to [ask] what do Mumsnet users think about Keir Starmer? What do they think about Sainsbury’s? What do they think about any question you want to ask?”
The benefits of Mumsnet’s AI-driven approach to data analysis range from the sheer size of the dataset to the fact that it can be run more or less in real time. “We are talking to brands, saying you don’t really need to go to the expense of running focus groups with mums anymore,” said Sue.
Anonymity and honesty
Sue highlights one crucial advantage for sentiment analysis in Mumsnet data – the honesty of the conversations taking place on the site. “People are having really honest conversations because it is anonymous,” she explained.
She describes Mumsnet forum posts as the opposite of social media’s ‘stealth boasts’ featuring pictures of angelic kids doing amazing things. “You don’t put the picture up of the moment they wouldn’t get in the car seat and they threw a massive tantrum,” Sue said. “On Mumsnet you are talking about those and you’re talking about the mother-in-law and the neighbours who’ve parked in your space and everything and anything. It is real, honest, authentic data, and there are incredible insights within that.”
As an example of real-life insights that undercut conventional wisdom, Sue points to the ‘twinning’ trend that sees parents dressing the same as their children. “Brands got really excited about it a few years ago. But two-thirds of people on Mumsnet hate that trend,” she explained.
They have also debunked the long-held belief that people book their holidays in January: most Mumsnet users book in July, August or September. Also the annual ‘spring clean’ is actually done in January, straight after Christmas.
Sue underlines the importance of user privacy and explains that AI is only used to synthesise insights from anonymous, aggregated data. However, she also highlights the community’s enthusiasm for sharing their opinions. “Our users love to tell brands and organisations what they think,” she added.
Real-world results
Looking at success factors, Sue says AI-powered data insights are helping with every brand pitch Mumsnet makes. “We can see it in our pitch conversion rate,” said Sue, “It’s going up as a result of being able to say, ‘Look, we are the experts on this audience. We can tell you what you can’t get elsewhere.’”
And while the platform has always been a rich source of insights, Sues sees potential to leverage AI to grow Mumsnet’s data-to-business services. “One of the things we’re working on at the moment is segmentation,” she said, highlighting the value in knowing what different types of mum are thinking at different life stages.
But whether its sentiment analysis, trend tracking over time or behavioural insights, the key for Sue is actionable insights for brands. “That’s what we hear loud and clear from them,” she explained. “Everybody has dashboard fatigue… there’s so many sources of data out there.
“The focus of what we’re doing is how we take that data and turn it into something that can be used by everyone who works in these organisations to improve what they do.”

Sue Macmillan, CEO, Mumsnet, talking to MediaVoices’ Peter Houston at The Definitive AI Forum. Picture by Simon Crompton-Reid, Confex Media.
Culture change
Sue described herself and Mumsnet founder Justine Roberts as AI evangelists and explained that they see AI as a cultural shift as much as a technology shift. They were early adopters of ChatGPT and have encouraged staff to share how they are using it across the business. “We have an email that goes out every week from the CEO, from me, and it celebrates a great use of AI this week,” she said.
The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”
She emphasised that the speed of change means that people need to be using the tools regularly to be able to understand them: “They are moving so quickly that people have got to use them on a day-to-day basis.”
AI is currently absent from the Mumsnet user experience, but Sue can see a role for it in improving search functionality and personalisation. “We are thinking about what it means for discovery on Mumsnet. There are possibilities around AI answers, like Reddit answers, but Mumsnet answers.”
However, while Sue sees AI assisting content discovery, she doesn’t see any place for AI in content generation. “A big part of the reason why people come to Mumsnet is there are amazing stories every single day about other people’s lives. What are human beings more interested in than anything else? Other stories from human beings.”
She suggested people should think of AI as a thought partner, an electric bike rather than a self-driving car. “You can’t just let it run for you, take your eyes off the wheel and play a video game. It’s an electric bike. You pedal and it goes faster, but you’ve got to pedal. You’ve got to steer.”
Listen to the full session in this week’s episode of The Publisher Podcast, available wherever you find podcasts.










