Below are some questions but first the context:
Last night I was working on the first of a series of conversations about how Birmingham will feed itself way into the future. The New Optimists Forum is organised by Kate Cooper who has the very powerful idea of getting groups of scientists from different disciplines and policy makers to think about this thorny problem. She argues, I think rightly, that getting practical about problems and places helps us understand best what we need to change now.
One of the scientists was Ian Nabney who talked about the opportunities that will come to make better decisions about complex problems when we have more data and more power to crunch and use that data. Here’s what he said.
It made me ask the question what if we created a new form of planning gain: supermarkets share their data with us rather than build a new badminton court.
Could knowing what they know about our eating habits help us lead healthier and better lives?
Mark Braggins Tweeted this this morning:
And it also tickled a local MP’s curisosity. Richard Burden (who’s Northfield constituency may have a few urban “food deserts”, another idea kicked around at last night’s forum) tweeted this question about half an hour ago:
So here are some questions:
- Is asking supermarkets to share their data a good form of planning gain?
- If so in what form would we want it – opendata, depersonalised or maybe full data to be share just with civil servants
- What would be the arguments against (so we can anticipate) or just how naive is this! ?
- How would Kate Cooper of the New Optimists go about talking to sainsbury’s about this?
- Would you rather have a new pavilion at the local park?
Odd what comes out of combining real world conversations with online stuff!
Looks like Adrian Short was thinking about supermarket card data as a public good back in April – scroll to the bottom of this post.
An interesting topic. However, one thing you can’t do in discussions about supermarkets and planning is seperate out ideology (as Nabney argues). In a sense this isn’t a problem that ‘science’ or ‘data’ can solve. In my area of Brum, around Stirchley/Bournville, we’re increasingly getting hemmed in by supermarkets (an enlarged Sainsburys to come at Selly Oak, an Asda and Tesco at Stirchley to go with an existing Co-op). There’s some discussion on my hyperlocal blog where you’ll see the word ‘ideology’ used and you’ll note the emotional tone of those commenting.
Potentially, access to data might allow both sides to strengthen their cases during the planning phase. Indeed one might find that those opposed to supermarkets are a lot more savvy in this area and able to utilise data to shape alternative visions beyond the glossy 3d fly-throughs and corporate press releases that the supermarkets offer up.
Whichever way, let’s not pretend that this is not a ‘hearts and minds’ issue, or worse, pretend that a ‘scientific’ response to this issue is itself free from ideology.
Oh and I should add that linking the release of data to section 106 stuff wouldn’t wash in most cases. Folk really want the physical stuff.
I’m intrigued by this. Partly through personal interest via an arts project I’ve bee mulling over connected with supermarket loyalty card data, but also more generally. I guess immediate question that springs to mind is being specific about what data you’re looking for the supermarkets to share and from what source. I assume you mean that shopper’s card data, right? Secondly, the only way I could see this even being considered is as depersonalised data, both from the supermarket’s legal standpoint and from an ethical point of view. Why would you want or need personalised data anyway?
Hello Jason – want to tell us more – or just want to mull this over? (how are you btw?)
Interesting idea – and to my mind instinctively seeming to be more honourable than the current system, which has always felt a bit like more of a bung which developers offer to get their controversial proposal through. But offering data instead of cash (or instead of paying for a new road to be built), being non-financial strikes me as a much cleaner alternative.
Really interesting idea this. My quick thoughts:
Has to be de-personalised or else can’t / shouldn’t happen for obvious privacy / data protection reasons; publish as open data, not just for civil servants but freely available.
What data? Loyalty card data seems most obvious: what’s being bought, when, where, how often etc. Customer footfall by location would be interesting.
Longer term publishing as linked data and linking to weather data and accessibility, transport networks, traffic conditions etc could be fascinating and of public interest as well as being useful to planners.
A similar idea was raised at the Society for Location Analysis (SLA): http://www.thesla.org/ annual forum a few weeks ago and I think it’s likely to be on the agenda for an event due Spring 2012 in Hampshire – will check that’s correct and post an update.
Will more considered comments when I’ve given it a bit more thought.
It seems to me that what Ian is describing is essentially a transport management issue. Now, the Travelling Salesman Problem has been solved for over 30,000 nodes and Ian states that he doesn’t want us to restrict the range of food on offer to us.
So, my best guess at what he might have in mind is some way of us seeing the data held by suppliers, transport providers and retailers so we can start to do some crowdsourced planning which optimizes the whole process of getting food from provider to our homes.
I’m making some large assumptions here. But if I’m right it will only work if there is a centrally planned operation across multiple commercial interests.
The other discussion is around access to either everybody’s de-personalised data or our own individual data.
Supermarkets invest a lot of money to create these data warehouses and I can’t see them letting us get our hands on them, more’s the pity.
Access to our own information might be more achievable, and something that Kate could potentially go to Sainsburys’ about.
If it could be demonstrated that people would build interesting applications that someone with a loyalty card could choose to run their data through *and this had commercial value* for Sainsburys, then it might be a goer.
For any of the other things above we probably need a revolution.
Great idea. Lots of points to attack.
1. I agree with Jason – depersonalised is the only way. Privacy issues.
2. The thorny “where-is-the-data-stored?” question. If this was to become law, could the supermarkets bypass it by moving their data warehouses to another jurisdiction?
3. A really cool spin-off would the multi-brand loyalty card. Could we have a single card that worked in multiple outlets? A bit like Nectar, but with big wings. Why do I have to carry 6 of them in my wallet?
4. An implementation thought: If we gave consumers the legal power to share THEIR OWN spending data, and made it an ‘opt-out’ rather than an ‘opt-in’ right, then the supermarkets could be required to share the data of everyone that hasn’t opted-out. I’d love to get hold of my spending data in an easy to use format.
That’ll do for now
On privacy: anonymised data can often be reverse engineered, this idea would certainly need to be passed by the ICO.
One option would be to have a trusted 3rd party that would host and match the data.
It seems (although I haven’t been able to hear Ian’s comments) that there are two types of interesting data – logistics data for supermarkets, and consumer data (split into two: purchases and demographics).
An alternative approach that would potentially avoid some pitfalls might be to crowdsource consumer data: build an app that allowed users to submit information themselves…
So, supermarkets are to ‘share’ with the public the data they took from us in the first place? Can’t we have the pavilion, the badminton court and loads more as well, please?
Hello Ladybird – sorry it took me so long to spot you comment. Hell yeh – lets go for the lot – and free dinners to please.