Our company is called [Hebe] Works. We’re a data and technology company based in Leeds, so not too far away from here, and this is us. Yes, it’s an iPhone picture site a bad one, so forgive me.
We were successful with the bid we placed with Nesta and Innovate UK to do 18 months’ worth of exploration R&D about how to blend media with open data and technology.
These are my business partners, Lee, and Simon with the bowling ball. Lee looks after The City Talking newspaper and the media side of the business. Simon does the technology, and I run the data side.
The other guys, we’ve got videographers, two storytellers, musicians, data scientists.
We’ve really grown. This is not the full team. This was a few months ago. This time last year we were six people, now we’re sixteen. We’ve really gone to town on the work that we’re doing.
We describe ourselves as T-shaped people. Each of us has a really deep understanding of one area of work, but what we try and do is get the guys working across the boundaries, blending those skills with each other, so they have a good understanding of how they can spread their [real big talents] with the rest of the team.
That works quite well. It can be quite intimidating when people come on-board, because they suddenly realise they’ve got to up their game quite a bit, but it brings everyone along.
Yes, as I mentioned, The City Talking is our newspaper. It’s been going now for about two years or so. We are in – well, we started off in Leeds. We’ve just published our 29th issue in Leeds. These are some of the covers from them.
We have done things differently. We were doing it by ourselves at the start, and then went into partnership with Yorkshire Evening Post to really – well, it was for a few reasons really. It was to really expand the distribution and the readership of what we were doing, but also to take away some of the pain as well.
Traditionally we’ve been running our own advertising, so the guys would be phoning up people, trying to place a £100 advert here, a £500 advert there.
The partnership worked really well at that time for us, because it meant they support the advertising, with a team that was set up, and they just paid us, and we got complete creative freedom to keep doing what we’re doing. It just meant we could focus on actually making the paper, rather than chasing sales all the time. It took a huge amount of pressure off. We could just accelerate quite quickly.
The way people view the newspaper we’ve always found quite interesting.
These are some of the Instagram shots that people have took, and we’ve noticed that when people are talking about The City Talking they try and frame the shot and create something quite pretty with it.
It’s always like people taking a picture with a coffee, or with a dog, or strategically placed record player. People time and effort to place these shots, and almost create a bit of art. It’s really nice for us to see that sort of thing coming back, and realise that people are viewing us in that nice way.
Yes, we’ve done quite a lot in Leeds, but we kind of had moments last year when me, Lee, and Simon were talking about, “Well, what are we going to do with the business? Are we just going to keep it in Leeds, or are we going to really go for it?”
All the best decisions come from just saying, “Okay, we will do it anyway”, and this is one of those decisions. (Laughter) So we’ve really accelerated over the course of this year. We’ve hired more storytellers. We’ve hired more people to do the video.
It’s meant that now we’ve gone into Sheffield, we’ve gone to Manchester, York, and now we’re going to Liverpool [as well] in December. So, the guys are working really hard back at base doing five editions over Christmas.
That’s amazing, and then next year we’ve got all sorts of other plans as well about further expansion, because people seem to really like it.
I thought I would talk about the media that we have done, and how we’ve actually used it in this project. I thought it would probably be good to start talking about data platforms that we’ve got that are focusing on the hyperlocal.
Leeds Data Mill is a platform that I set up with Leeds City Council about 18 months ago now. Yes, crikey.
Because there are a few data platforms out there, what we wanted to do is not just do the city council data, we wanted to get the private sector data in there as well, and tell the full picture of the city.
I sort of pulled out quotes from York, [Calderdale 0:05:41], quite a few cities around [about us] that are doing some interesting things, but everyone is talking about the same thing. It’s about engaging with new audiences, pushing out the data, and seeing what people do with it.
Some people do really good stuff, other people do things that aren’t so good, but actually they’re all embraced. That’s the general feeling, that everyone is good to have a go, from bedroom developers to big agencies.
There’s a lot of support for it with events, and just engagement really. It’s been a really good movement that’s been happening.
It means producing fun stuff. This is something that I did yesterday when I had a spare ten minutes.
Its based on information on taxis. I just had a quick dig around and found the most popular taxis is a Skoda in Leeds. I would say the most popular is black and white, closely followed by silver.
Then we got a bit more into it, the newest taxis in the city, the best number plates ___. The best taxi, a BMW 530d, which are beautiful.
I was just having a bit of fun, but that’s the lightweight stuff that we don’t really do in the newspaper at all. I was just having a bit of fun. But it’s there. Why not?
So, yes, Ones and Zeroes. This is what everyone asks. “You work in data, so you deal with ones and zeros. Ha ha ha.” This is a screenshot from my computer when I was actually doing Ones and Zeros, so, [yes it’s true].
But when I look at data and data platforms it is a badly made package. it’s horrible to work with sometimes, and you’ve got to know what you’re doing. You’ve got to be able to ___ [Excel], let alone do other exciting things. And it’s a bit elitist, it’s a bit geeky, but actually the stuff that’s inside these spreadsheets is really good. You can get lost for days in datasets.
We’ve just had some prescribing data, and we’re just getting really into it, finding all sorts of things that we wouldn’t have known .
Some of the health things. My background is health as well. Some of the things that we tried to do is around loneliness in Leeds and trying to define what it might look like.
We found, again, all sorts of things, but nothing that really could help us define it.
I was working with the A&E department in Leeds, asking them for data that isn’t open but wasn’t patient identifiable, we’ve anonymised it to an extent, and we could find all sorts.
Some of the interesting things were around self-harm from 10 to 20 and 20 to 30 year olds, where that kind of peaks. The thing that I sort of had a preconception about was that self-harm happened with that age group admittedly on Friday and Saturday night at about midnight.
Actually, there’s no correlation between time. People were getting admitted at ten o’clock in the morning on a Wednesday or three o’clock on a Friday. There’s no correlation.
Some of the other things, confusional state, as you would imagine, [in the] 80s and 90s.
Anxiety. Now, this was a really interesting for us, because this is our demographic for the newspaper, 14 to 35 year olds.
My business partners say whatever age I am the newspaper is supposed to be really cool. So whatever age I am the newspaper is targeting to one year less. So I’m never cool. Because they’re arseholes…(Laughter)
Anyway, our target market, 20/30’ish year olds are being admitted into A&E and being diagnosed as being anxious. There’s no treatment for that, other than to calm them down, help them, and then maybe refer them back to GP practices.
Another thing our analysis detected, as you would imagine. Parents going in with kids zero to ten years old, they’re most likely going to be zero to one years old spike Then, again, 20 to 30 year olds.
So, it was quite interesting. By looking at the data it could be quite exciting.
When we’ve tried to blend the data with newspapers we’ve had mixed success, I would say. This was the first attempt at doing it, where we were talking about Christmas in Leeds.
We were trying to combine – well, it ended up being an infographic, and I’m not a fan of infographics at all, but that’s how it ended up. It was our first attempt. It was good.
We’ve got things like largest footfall the week before work? When is the worst time and the best time to do Christmas shopping? What are the most tracks played at Christmas? What sort of drinks are people having?
We kind of just built up a picture, and then a writer Danielwrote up a poem to go alongside it.
So, it was initial stab at combining hyperlocal data with a newspaper.
The more exciting stuff…
As a company we try and do things with depth and beauty. The first attempt was not depth and beauty.
This is getting more towards that, where we were doing video to promote Thought Bubble, which is a comic festival that happens in Leeds.
We were sort of teasing people with it, and putting little bits and pieces out, so I thought, “You know What? I will create a diagram where you’ve got your own superheroes, so Superman, Batman.” These are the top ten superheroes according to The Superhero Database.com which really exists!
Then we were looking at the superpowers that they have. So as you roll over this one all Superman’s powers become available. If you look at Hal Jordan, the Green Lantern, his only skill was having a green lantern. (Laughter) Whereas Superman and Spiderman…
It was just trying to blend data with media, and just trying to do something different.
Question: What’s that called, the graph?
Mark: That’s a chord diagram using DJ.js
Then we started to do things that I got really excited about.
This was a project that I did over a month, where we were looking at air quality in Leeds, and trying to blend data with video. We done it in the newspaper. We’ve done it in digital, like a graph that you can mouse over.
I will kill the sound, because there’s a lot of tracks [going on at the same time]
It was just basically trying to visualise a lot of data. Air quality in Leeds, we’ve got a really rich dataset where air quality sensors are taking measurements every 15 minutes or every 30 minutes, so I thought we would use that and try and create a compelling story.
We’ve got four different areas of the city. We’ve got Kirkstall, Temple Newsam, the City Centre, and [Headingley]. Headingley is like student central. That’s like semi-urban, city centre.
This is [Giuseppe], our videographer. We roped everyone into doing this. [Joel], our designer. Joel’s mum and my wife. We roped everybody into it. (Laughter) Giuseppe did the music as well.
It’s trying to respond to pieces of data. [Not that] ___ [measuring] contest, but it’s a lot of data ___ [to play this from] start to finish.
What we’re trying to do is show the effect of that on prevalence of heart disease in different settings.
On the website you can mouse over and flip between the videos and see what the differences were. We tried to make it the exact same scene, but with different characters.
Up here you can see the ranking ___.
Question: It looks utterly amazing
Question Is it available on the website, just on the website?
Mark: Yes, just on ___. Yes, we ___. The stuff that we did was ___ [lots and lots of] data ___ many different platforms as we could and it was very difficult to work out what the narrative was.
Working with storytellers in the company meant that I could basically chop out loads of scenes that I had made where the data wasn’t really relevant.
The main thing was around asthma and the air quality. One would spike and you can compare those. There’s a huge spike there. Why is that?
You can see who has passed and who has failed the air quality. Leeds City Centre fails with 28 parts per billion.
Oh, yes, [this was a really] ___ [over the whole] ___ was that when the city council closed down all the roads for the Tour de France air quality improved by 20%.
[We want to] tell good stories, positive stories in the newspaper. We don’t want to beat up the council for that. We just want to send out a message, like, “What can we all do to help us achieve that more often?”.
So, It was great fun. [We’re] hanging out of the back of cars, running down the streets, and just having a [little bit of fun] filming this. That was our first attempt at doing video as well.
Question: You said it was a month in total to do that. How many people were working on that ..?
Mark: I did the data. We had two people filming. Giuseppe did the music. [Oscar] did the – underneath it is a story about the city. I think about five of us put that together.
We filmed it in about – well, one day for each story. We were just flying around in the car. [I get to do the] locations ___.
I thought I would show the process sometimes for how we work on video.
This is a video we’re doing in partnership [with the BBC], which again…post it notes on the wall. It’s a film all about the Leeds music scene. It’s an hour long feature film so if you go to that URL you’ll see the whole film.
So, yes, a lot of it is about what information exists around us. These are just on the information the exists next to our office in Leeds.
There’s so much of it that’s available, and really not a lot of this data is collected, but when it is we can turn it into things that are interesting, and that’s our area of expertise now that we’ve found.
We’ve done really simple stuff like findmybinday.com. Imaginative URL it does what it says on the tin. (Laughter) You [type in] and it tells you when it’s your bin day. You can put it in your calendar, and you can find out what sort of things to put in there.
We’ve been constantly playing with different things that we can do with this data to see what sticks. We’ve been very fortunate. We’ve had that freedom.
We’ve also done some interesting work around footfall. Again, there’s a really rich dataset in Leeds, where we’ve got eight cameras in the city that record information every – I can’t remember. What was it? Every hour in the city. That data from those sites for the past 10 years, so we’ve got this huge amount of data that we can play with.
One of the things that we’ve made, that we haven’t released yet, because we’re still finishing it off, but as [you mouse over the bottom these circles] get busier, so you can tell – sorry, that’s should be playing. (Laughter) Anyway, you can tell [by mousing over] how busy the city is.
So, we’ve done some other interesting things around that, predictive analysis with footfall data. We’ve had data scientists use those cameras to try and predict what the city centre is going to look like, and they’ve got pretty accurate.
The original is the green one, and our forecast is blue. On a yearly prediction we were about right. When we draw down into a daily analysis we get pretty accurate as well.
That’s where it gets really exciting for me, is when we’re predicting by hour, and now we can predict, using the data, using the data science capabilities, to plus or minus 50 people in the whole city, for any hour over the next year.
So, if you were asking, “Well, how busy is it going to be at three o’clock on 3rd February?” we can tell you within 50 people how busy it’s going to be.
With all this information that we’ve got, we’ve tried to create something out of it.
So, rather than doing little bits and pieces, that aren’t joined up, so people have to go all over the place to find out actually what we’ve been doing, we’ve put it all together in the dashboard called Solomon, and it looks like this. We’ve got a range of different things on there that tell stories.
We want to basically visualise the data, so take it out of that horrible packaging, find the gem that’s inside it, and then give it back to people so they don’t have to be data analysts to understand what’s going on in their city.
We’ve done things like where’s the nearest bike bay? How many parking spaces are available?
We’ve even done something on statutory notices, where we’ve plugged them all into our system, into our system, solomon, so you can view all the statutory notices, see how you interact with them, and you can see where that is on the map.
Basically we’ve just tried to take a digital solution, rather than it just doing something that goes into the newspapers.
Each story has data. You can flip them over and see the data source, where the data comes from, and who has created this story, and it all sits together in a dashboard.
The dashboard itself you can have multiple canvases. I might be really interested – because at home I might be a school governor, so I want stuff about schools in the local area, especially about my school. But I also might be really interested in culture and interesting things that are happening in the city.
So we have these multiple canvases, where you can go in and see different information that we’ve created.
Also as a user I might want to just have my own dashboard, where I just have one story from there, and one story from there, and forget the rest of them, I don’t care. I just want something that I’m interested in. So we’ve created it like that.
You can get to the dashboard. You just get a plus icon. As you click the plus icon you can add stories in. Or you can click the toolbox, and it will let you add a story into that, so you can create it that way. It will bring up a whole range of stories that are available to you, and you can just start adding them in.
You can equally feedback, and that will come up in our slack Channel, so you can interact with us over that, and get general help, and create all your canvasses.
It just [means] you can have all this information at your fingertips, so that even if you’re just going into a meeting about a certain topic you can just [get the relevant] information for you.
The other things, this was in Leeds. This is just around the corner from our office, where there was this huge swarm of bees that happened.
That’s really newsworthy in a really hyperlocal setting but it only lasted for about two hours/two and a half hours. For this presentation I was desperately trying to find an image that represented it, and this is actually where it happened.
It just so happens that that’s actually my dad, which is weird. (Laughter) I didn’t know that he dealt with this, but he’s a beekeeper, so it’s not…
In a whole twist of irony, it turns out I’m allergic to bees through my massive exposure to them when I was a kid. Now I have to be admitted to hospital if I get stung. Thanks, Dad. (Laughter)
That’s provoked some of the stories that we wanted to tell as well around allergy advice. This is a very selfish story, just for me, where I want to be able to know when people are talking about bee stings [on Twitter].
It’s basically a reminder for me to go and get an EpiPen, so that my wife can stab with me in the leg with it if I get stung by a bee. Which she is desperate to do. (Laughter)
Yes, I forget to buy one of these EpiPens, so what we’ve done is we’ve taken social media data, we’re basically scraping it all the time on set topics, this one we’re scraping for bee stings.
Every day we might have 10 people talking about bee stings, but all of a sudden 50 people are talking about it. It’s never going to be something that’s trending in Twitter, but it’s really significant at a hyperlocal level.
So what we’ve done is basically create that algorithm, suck up all this information, and then that will alert us on the dashboard. So, it just sits there, alongside all the over stuff.
We’ve [had real fun on this] journey over the last about 13/14 months now. We’ve been really fortunate to get the funding, which meant that we could accelerate the stuff that we had in our heads that we really wanted to do, and we’ve got lots of plans for other things that we’re going to be doing with this.
It’s all open source, so this is on GitHub anybody can make it. We work with developers that can create their own stories, so academics and people come along [and] they’ve got some interesting things that they want to show.
There’s a guy in London that’s just downloaded it and is using it to track all his health stuff, and his gym stuff. He’s just got it as a closed down platform so nobody else can see it.
It’s responsive, so it works on every device. We just tried to make it useful. We just tried to turn this horrible data into these gems.
That’s it from me. This is a quote that has really resonated with me over the last few months. “We know that data is powerful. [Everyone is] talking about it like it’s the new oil, it’s the new everything, but actually it’s not until you turn it into something that people actually can interact with and enjoy you get more from it than you would from just looking through a spreadsheet.” That’s what we’re trying to do at HEBE .
Thank you. (Applause)
I’m happy to take questions.
Question What’s the business model around it? What do you think is the economic value that you’re hoping to extract and how?
Mark: That’s a good question. With the newspaper, obviously we’ve got adverts going in there, so there’s a really clear relationship there between revenues coming in, so that’s straightforward 0:26:26.
What we’re trying to do is work with brands, to have brand partnerships with them, because we’re talking about things in a really positive way.
Like with [Tom’s Trainers]. The company came to us and wanted to just put an advert in the newspaper. We said, “We could do that, but let’s use the money and commission five artists in the city to crystallise them, to paint on them, to graffiti on them, and then we will do a profile of the artists”, and it just so happens that the Toms are in there.
Question: What around the data specifically?
Mark: The data, we’re commercialising Solomon Dashboard at the moment, so we’re working with Yorkshire Water, we’re hoping to engage with with a few other companies, like big companies in the country. Doing that kind of thing.
Question: Someone would pay to have a little part of the dashboard that they’ve got a story in there, with their dataset?
Mark: Yes, exactly. We worked with NHS England when they wanted a locked down version which showed them their KPIs every month. We’ve designed it so that we can use it with anything, really. It’s not tied to just working with a city. It’s for anything that we think we can apply it to. We’re going to be putting video in there. We’re going to be putting more imagery more customisation. We’re going to be doing like a WYSIWYG editor in there, so you can create your stories around the data. We’re really expanding the functionality of it.
Question:So, on the Solomon Dashboard are all of the little [tiles] automated to connect with the data [so you can continuously update]?
Mark: Yes. Some of them are actual live, so they’ve got the data. You don’t need to refresh it, it just automatically updates [with that data, so it continually taking data feeds from it.
Some of the other things, yes, we push straight into the Data Mill, put straight onto the data platforms that [are APIs 0:28:32].
Often we’ve found that actually what we need to do is suck in the data, pull it for our own API, our own engine [based in the back-end, that] then spits it back out as an API.
A lot of the data is messy, so columns will be named differently every month. There will be things like errors in the data, so it won’t load.
There are all kinds of issues, but we’re basically trying to get rid of all those headaches by creating an engine that fixes it all for us, really. That’s our data scientist has done that.
Question Is that a custom-built bit of software?
Mark Yes, we built that over about two months, because we were discovering all these issues, ___ fix it.
Question How did you avoid malicious businesses, like rivals, obviously putting stuff on? Say you had one of the taxi firms, you would get rival taxi firms, putting deliberately bad messages about their rivals?
Mark: Well, the company [will] do that.
In the case of the taxis, it’s Leeds City Council that are putting that out, so the taxis are legally obliged to give them that information to get their licence. They’re just collecting that licence information and then spitting it out. It’s just I’ve gone in and found which taxi firms they are. It’s not the taxi companies themselves submitting the data
Where companies are [re-submitting] data onto the [Data Mill], where companies are doing it, the information goes to the city council, and then they give it a quick once-over, to make sure there’s nothing identifiable in there.
So, you’re not identifying people, basically. You’re anonymising the data, and you’re checking, when you’re looking through it, that it looks right, and there’s nothing malicious in it.
Often just pushing out the data, when people pick it up, they like armchair auditors go back to the city and say, “There’s an error in this”, or, “I don’t think this is right.” Actually, we’ve seen data quality being driven up by that engagement with the city council and the office
Mark: It’s just licensing data that the city council ___.
Question You’re not actually [saying anything about] people’s experience [of that taxi firm]?
Mark: No, not at all. It’s just a summary of the base facts. Yes, just like which company is licensed, which cabs, what colour they are, what registration plate they have, that kind of thing.
Question Can you do how many drivers are CRB checked ___?
Mark: We just collect that data, ___. I imagine they collect that stuff, but [the] city council [do it]. It’s not something that we have any involvement in. We just take the data. Hello? Yes? ___. (Laughter)
Question: That’s interesting. Relating to that point, this is a quick question followed by a long one. Your experience with GPs, [the potential for being sued]. Have you now just thought ___?
Mark: No, not at all. That’s a time and place. This was five years ago. It was just learning of lessons from it, so that’s not really an issue. The data is out there, it’s published, it’s available.
Question: [Crosstalk] the open data ___ publishing the reviews of taxi firms, as you’ve mentioned there ___ I think ___. Thank you. The other one was about experience of working with the council to open their data, because we have tons [of the stuff
Mark: Yes, that’s right.
Question: How was that process? how did you encourage them? Was there enthusiasm internally? What were the key parts to [it]?
Mark: Well, it started off with just an idea, and I pitched it to a guy I know at Leeds City Council, and said, “Why don’t we try this as a city? It could be really exciting. [Nobody is doing hyperlocal data like this].” He got excited about it, and I pitched it to the board at Leeds City Council, and they all thought it was a really good idea.
It was a little bit of funding. I think it had about £30,000 initially, just to try something out. Then I managed to secure another £200,000 for them to push it forward
It was really their chief executive saying, “Yes, this is a good thing. Let’s do it as a city. Let’s see what happens. Let’s just publish whatever we can, and let’s see who picks it up, and let’s see what we do with it.”
I’m sure there was a little nervousness from them, but they didn’t really show it. They just embraced it.
This has been big success in Leeds. There’s loads of different movements that happen [around it]. They’re leveraging that, and that success that they’ve had with that and smart cities building on those initial foundations
Question:Do you think the chief executive gave it his blessing?
Mark: Oh, yes. Yes, [Tom] ___ was such a key thing to happen in getting the services involved, and saying, “This [is a thing the city council [now are doing].”
It started off with one guy, Stephen, at the council, but now he’s got 10 or maybe 20 people that are working with services, just as part of their job, for an hour a week, or two hours a week, or whatever, just data and help them [through that]. Even going into the [back-end] of the site, and hand-holding them through that process. So, yes, ___.
Male: ___ [loads of questions for you]. As we know, all councils are being strangled by central government cuts so the justification of opening data that takes 10/20 people, it becomes increasingly hard to justify unless there is a visionary at the top who is saying, “This is going to be good for the city.”
Facilitator: Yes. I think ___[0:35:30] is that ___ do something different, and in their case [they were seeing] open data as a way for the people to find services in the city, like [communities led].
Mark: Yes, outsourcing the knowledge and the insight to give back to the council.
Question: Thank you.
Question: How important is your [print] product to the success of City Talking?
Mark Oh, massively, yes. That is City Talking, That is the city talking. It started off as a Facebook group initially, so it’s been an interesting transition going from digital to print.
We hardly touch Facebook anymore, and the print is its so important for all the business. It’s not only just about sharing skills, but because we [tunnel] things in such a positive way it means that when we’re going into a new city then they embrace us, they’re happy for us to be there.
Because usually they’re dealing with perhaps journalists that beat them up all the time, saying, “This is going wrong. You need to do it better. Why hasn’t this happened yet?”
We’re not like that. We tell stories about underground movements that are happening, like what art galleries are opening , new coffee shops, just people doing good stuff in the city, and trying to…Often they don’t get a voice in the city, so because it’s seen as such a positive thing it means that you have really grown-up conversations with city councillors now, and it leads on to other things with the wider business It’s just good fun to do as well. The amount of cool people that we meet ___. Yes, it’s just a really creative thing we felt that we could all get excited about.
Question: Do you have a bigger audience for that in print than you do online?
Mark: Our reach is – it’s hard to say. I think we’re distributing to about 70,000ish Digital… but it’s hard to measure. We don’t really measure things like that. It’s just like with the Instagram shots, where people are really into it. That’s how we measure it, people liking what we do, rather than looking at the number.
Maybe it would be easy to do that with data with data analysts in [business].
What we don’t want to be is a company that’s just creating content, [like buzz-feeding], because we know it will get lots of people looking at it and it will go viral. It’s not of any interest to us, that kind of thing.
We’re telling beautiful stories, [with beautiful photographs], and we’re just doing it our own way.
Question: How often do you publish your newspaper?
Mark: In Leeds it is monthly, and in the other cities we’ve got them quarterly, pretty much. Its distributed in like cool trainer shops, or coffee shops, or art galleries. Just places that we think our audience are.
Question: Do you have to pay ?
Mark :For the distribution ___?
Mark: No, I think [they’re] keen to have it, because people [aren’t] going in just to pick up the newspaper, are they? They’re having a coffee there at the same time. So people are going there to read it for half an hour and then grabbing a coffee, and reading it
Question: Is not all this information already online on the council website?
Mark:No, it wasn’t. Until I started the Data Mill it was just in the council . Then when we launched [at least 50 and now it’s over 200].