Leeds Bins: the challenges and rewards of building an app powered by open data.
Leeds Bins is an app that tells you when your bins will be collected in Leeds. It's simple, it's been used over 12,000 times in just a few months, and it's free.
The app relies on two large datasets on Data Mill North that are automatically updated every week. One file holds a list of the 330,000 addresses in Leeds. The other holds a list of bin collection dates. A unique property identifier links the two.
So far so good. If everyone in Leeds knew their unique property identifier there wouldn't be a problem. But they don't. Of course they don't.
What almost all of us do know is our postcode. Since they were created in the late 60s this is how we expect to enter our address when we want to receive information by post, on the web, or in an app.
Because it's what users expect, Leeds Bins works that way too. The user enters their postcode and then selects from a short list of addresses. They get their bin dates and save them to their calendar.
Postcodes were designed, rolled-out, and promoted by The Post Office back when it was part of the UK government. They are the most widely-known and one of the most widely-used parts of the UK's data infrastructure. And yet the postcode system's restrictive licensing options limit its applications in business and public service innovation.
Since the privatization of the postal address file (PAF) in 2013 the hope that postcodes might be available under a more powerful license — or even as open data — has substantially reduced.
This affects the Leeds Bins app in two ways.
- It costs more. We have to pay 2.4p per search to get an address from a postcode.
- It works less well. We often struggle to match the returned address from our postcode search with a unique property identifier. This is because the postcode of each property cannot be released as open data alongside its address by Leeds City Council.
Since releasing Leeds Bins we have spent over £600 on postcode lookups. Partly that's our fault; we didn't cache postcode lookups properly during development. But partly it's just the way that the system is supposed to work.
Our costs are currently manageable but they will keep rising as the app gets more popular.
We don't know what we haven't been able to do
Open data on bin collections released by Leeds council has let us reach different people in a new way. The council has been able to save money by reducing the frequency of traditional bin-day mail outs. This Christmas we are testing whether it reduces contact centre calls (Leeds release open data on that).
Our app has been developed in an innovative way with much lower costs than if it had been developed internally or by a large contractor. Most importantly, lots of people are really happy with their app and it is improving communication and reducing costs.
But while open data has made this possible we are frustrated that the reliability of our app and the cost savings that it could have generated have been reduced by postcode data remaining closed. We have spent much more development time than we needed to and produced a worse product than we could have.
Data infrastructure like the national postcode system needs paying for somehow. We're happy to pay for it but there has to be a better way than this. If the UK had a widely-used, free and open postcode system as in France the Leeds Bins app would be six months better by now. We remain hopeful that the government's efforts to create such a system will be continue and be successful.
I explain more in a video!
If you want to learn more, I spoke about the Leeds Bins app in my ODI Lunchtime Lecture in 2017. Skip to 5 minutes in this video, or watch the whole thing.