…to think Mumsnet data could be fascinating!

[JustAnotherNameChange2026]

AIBU to think Mumsnet data analysis would be fascinating… and to wonder if it already exists?

This might be a bit nerdy so please be kind 😅 but I was wondering whether Mumsnet ever does proper data analysis on posts over time for anything other than marketing?

Given how many years of posting there is, it feels like a goldmine for understanding how women are actually feeling about the world in a given moment.

Things I’d love to see analysed, for example:

Sentiment analysis over time – are posters generally more negative/positive in certain years?
Do big national or global events show up clearly? (pandemics, elections, cost-of-living crisis, etc)
Are there spikes in “I’m miserable” / “at breaking point” type threads during certain periods?
Do women post more negatively during major football tournaments like the World Cup? 😂 (half joking… half not)
Seasonal effects – January gloom vs summer optimism
School holidays: more threads about burnout, relationship strain, or mental load?
How language around motherhood, work, or relationships has changed over the last 10–15 years
How certain boards skew more anxious/angry/supportive over time

AIBU to think this would be genuinely interesting (and potentially quite powerful), rather than creepy or pointless?

And does anyone know if Mumsnet already does this sort of analysis behind the scenes, or releases anything like it publicly? Other than for marketing/advertising purposes which is all I can seem to find online.

So AIBU, or would you read the hell out of this too?

Maybe I’m a bit nerdy too as I also think that would be fascinating!

As a fellow data wonk, I’d be all over this!

What methods would they use

Totally fascinating!

I completely agree it would be interesting, but not actually particularly valuable. At my work we would call this "data tourism". Dead interesting, but not actually of value to the company. Which is a shame because I bet there would be some surprises in there!

Would the data be given to the government though?

it probably funnels its way to them anyway I suppose

yanbu, interesting op x

I definitely believe there is a way to commercialise this data, it’s just finding the right source of funding. And I personally think work like this would be in the public interest so there could be public funding or funding from colleges and researchers to take a project like this on, which would be mammoth I think!

I guess my angle is less “this would help MN/the company” and more “this would be a fascinating cultural snapshot”, even if it never went beyond curiosity value. And yes, I suspect there would be some really interesting surprises!!!

This website has definitely been crawled and used in AI training. As has just about every freely accessible website (and many paid for ones too).
How we stop this is more important in my opinion. We can plant "fake" posts like the A to Z map company used to do with fake streets but I doubt that would be enough.

Probably fairly simple stuff to start with, segmenting and cleaning the data and then things like tracking word usage and overall tone over time (more positive vs more negative language), looking at spikes in certain themes or phrases, and comparing volumes around big events. Not deep AI mind reading stuff, just pattern/trend spotting at scale. Enough to show trends without over interpreting individual posts.

Then you’d get more detailed and start drilling down into that data. For example, the data shows that actually women were far more miserable during the World Cup. So then you drill into why… finances being used, partner absenteeism or drinking or drug taking…. Whatever!

Well I don’t believe there’s any stopping AI, that’s definitely here to stay.

But what I’m talking about is data analysis… people have been analysing data since the dawn of time. AI just makes it faster and more efficient.

Go to Aston University and fill yer boots. 😠

So what would people love to see analysed, given Mumsnet is 20+ years old?

Oh? Tell me more….

Oh please someone do this and write a Dataclysm style book about it!

I haven’t read that! Is it worth a read? Is it very much focused on how people find relationships etc or is there wider insights?

Wasn't the data scraped by someone doing 'research' and a thesis into how transphobic MN is? I seem to recall there was quite the furore about it and rightly so. I mean, fancy a lot of women standing up for the rights of women and girls. How very dare we. 🙄

Ahhh, that doesn’t surprise me, I’ll have to go look it up. But that would obviously make users wary of how the data could be used.

But, there’s a big difference between someone scraping posts to make a point about posters, versus high level, anonymised trend analysis that doesn’t single out users or push a particular narrative.

What Id love to see wouldn’t be about judging or labelling the community, more the kind of broad “what was being talked about and how did it shift over time?” stuff you see.

Dataclysm is brilliant, definitely worth a read. It’s really clever what he does with the data (it’s a few years old now but still cool). It’s not really about relationships, it’s about how people present themselves and the disparity between what they say and what they do.

Oooohhhh… I just thought…. What about an annual personal Mumsnet Unwrapped!!!

Peak posting times
Sentiment analysis of personal posts over the year - positive, negative, fuming!!! 😅
‘You read 312 threads about in-laws and still felt shocked every.single.time.’
You told 19 people to LTB, 7 to wobble their head, and 1 to change the locks
Your most typed phrase was ‘You’ve got a DH problem’

😂

This study from Drexel University uses facebook and Twitter to analyse posters. There is one for Reddit too. MN cant be far behind
https://theconversation.com/narcissism-on-social-media-tells-us-a-lot-about-ourselves-18308

This research from Liverpool University uses Reddit as a source though it could be MN too.
https://theconversation.com/people-with-personality-disorders-often-use-language-differently-our-research-reveals-how-271109

I just about have the skills to do that. But I won't as it feel like a betrayal of people's confidences.

That's why MN is so valuable. All that data to sell - what changes minds, what reframes discourse, how perceived peer (ie, is it 'just another woman like me and not a fake poster at all') pressure works with this lucrative demographic.

It's certainly not the user interface or server space on a 14 year old PC running Windows XP by the looks of it that's made it valuable enough to be negotiating with private equity firms, after all. It's the most valuable thing of all. Data.

How do Universities get away with studying social media posts then? I suppose once they are in the public domain......

Justine licensed OpenAI to crawl Mumsnet, so they could easily slice and dice the data like that. No doubt they already are.