This week I picked up some comment around the last in a series of Open Data Master Classes. I think a lot of the content of these classes was around the use of OS OpenData™ but the comments and following thoughts probably apply to all public sector data released as OpenData. The conversation was about the constraints on startups building businesses on Open Data. Jonathan Raper had recently written this blog post explaining the challenges that small businesses experienced in trying to work alongside/in partnership with government in this area. While I have some sympathy with his view, I am not sure that he really makes the case for the economic benefit and opportunity from OpenData. It seems to me that the promised innovation is slow or non-existent (particularly in a commercial sense), data.gov.uk has been open for nearly 2 years and the London Data Store has been around for about 18 months, surely we should be seing some shining examples by now?
Quoting from Jonathan’s list, the principle players are:
- OpenlyLocal promoting local government transparency
- Placr aggregating public transport information
- CycleStreets open access to cycle routes
- Level business opening Companies House information
- MySociety campaigning with open data on transport, housing and transparency
- Open Knowledge Foundation creating standards and technologies for data releases
- OpenStreetMap crowd sourcing a map of the world
- The Stationery Office government data publishers
No doubt you can point to some others that Jonathan has missed (itoWorld for example, maybe they are too competitive to placr?) but just for a moment let’s take stock of the organisations that are listed.
I am going to discount The Stationery office because I don’t think the innovation and new business theme of advocates of OpenData intended the new businesses to start up within government agencies or be funded by government until they could be privatised! Let’s have a look at the remaining 8 including itoWorld and categorise their business models. Incidentally 2 of the organisations (CycleStreets and OpenStreetMap), to my knowledge, are not using public sector OpenData, they are based on crowd sourced data which is rather different.
I have broken things down into a few broad brush categories and then assigned companies to one or more category as they may have a mixed revenue model.
- No apparent business model at this stage – 2 companies.
- These companies may not be seeking to build revenue, the sites could be a showcase for development skills and they may earn money from contracting.
- Alternatively they may be searching for a revenue model and it just isn’t yet clear to them (or at least to me from their sites) what that model is.
- Grants or other funding from the Public Sector – 4 companies
- In a time of austerity, relying on Public Sector to fund you with research grants or project based fees for manipulating OpenData may be a tad risky, funding may be curtailed or opened up to competitive procurement.
- Consulting – 2 companies
- Of course consulting is a viable business model but if the consulting is to help public sector bodies get to grips with the challenges of OpenData this is hardly the creation of the vibrant innovative high growth businesses that some advocates have predicted.
- Charity/Donation funding – 3 organisations
- Of course we want a 3rd sector but is the Open Knowledge Foundation really an example of an innovative startup that will create employment and new tax revenues?
No doubt my scepticism will be greeted with howls of “traitor”, “judas” etc and perhaps some of the organisations listed by Jonathan may want to follow Jonathan’s lead and set out the opportunities to build a business around OpenData in the way that he has done in his blog.
So what’s the point of this blog? While I have been a strong supporter of opening up data for 21st century democracy, supporting Transparency and even possibly Accountability (but see some reservations below and my previous posts and presentations on this subject – accountable to who?) I have always doubted the fantastic numbers (where on earth does that preposterous but much quoted £6bn come from?) that some advocates of OpenData have placed on the innovation gain.
With OS OpenData™ (can I say how much I dislike OS’ attempt to attach a trademark ™ to OpenData) the economic benefit seems to be largely down to the cost savings that some previous purchasers of the data have made and of course the important democratic gain of having some core reference data sets that can be used to visualise much of the other OpenData (the uncopyrighted or trademarked government type). Now if someone at BIS would give the Royal Mail a righteous kicking and get the PAF out of their grasping fingers we would have the missing part of Dr Bob Barr’s Core Reference Geography (yes I know NLPG/AddressPoint/AddressLayer/GeoPlace would be even better but PAF would be a start). I know an exercise has been commissioned to try and establish the economic benefit of opening up some OS data but the results have yet to be published, I look forward to considering the results if they ever see the light of day.
Personally I am very doubtful that there is a great deal of opportunity to generate profitable new information products and services on the back of public sector OpenData. I just can’t see colossal opportunities in bus departures, live train info, government expenditure data, planning, environment, health and crime data which are some of the topics that have attracted the most clamour from advocates. Of course there will be some small businesses that will survive or flourish but this is not the great information nirvana that we were promised by Rufus Pollock, Sir Tim Berners-Lee and Nigel Shadbolt. Maybe I am too impatient but in the fast moving world of digital, with all that suppressed entrepreneurship surely we should be seeing a bit more by now?
One last thought. OpenData is not without cost, infrastructure, API’s, web feeds, data maintenance, cleaning etc all cost money and require some level of human involvement at the moment. That is cost to the public sector that is making it available, possibly some innovative startups can do this more efficiently than government IT folk but there is still a cost to doing it.
Tony Hirst has done some great stuff visualising OpenData and would be considered by most to have impeccable credentials in this domain. In his blog post “So What’s Open Government Data Good For? Government and “Independent Advisers”, maybe?” he summarises some of the arguments for OpenData and looks at how businesses might monetise OpenData. He reviews the commercial models and suggests
“When all’s said and done, though, the biggest potential is surely within government itself? By making data from one department or agency available, other departments or agencies will have easier access to it.”
Tony manages to wrap up the accountability bit (remember that) with the commercial stuff with
“Maybe then, there is a route to commercialisation of public facing public data? By telling people the data’s there for you to make the informed choice, the lack of knowledge about how to use that information effectively will open up (?!) a whole new sector of “independent advisers”: want to know how to choose a good school? Ask your local independent education adviser; they can pay for training on how to use the monolithic, more-stats-than-you-can-throw-a-distribution-at one-stop education data portal and charge you to help you decide which school is best for your child. Want comforting when you have to opt for treatment in a hospital that the league tables say are failing? Set up an appointment with your statistical counsellor, who can explain to you that actually things may not be so bad as you fear. And so on…”
Maybe the PDC are right to be cautious about the economic benefit of opening up more public sector data.
Enough said, several hostages to fortune and the future left here. Perhaps someone will offer a convincing explanation of why I have got this completely wrong (sits back and waits)
8 thoughts on “#OpenData – who profits?”
Very interesting post but I think you’re way too sceptical!
I’m starting to see OS OpenData, in particular, all over the place. I’ve been using it for months to make canal maps, and yesterday I found someone else is using VectorMap District to do the same. The iPhone App Store has lots of OpenData-powered products. People are tentatively starting to make stunning outdoor maps by mashing up OSM and OpenData. And, of course, OSM itself uses OS OpenData as one of its many sources. None of these people were using OS paid-for products before.
As Tom alludes, the main value of open data is to the little guys. Big guys like MySociety have a good chance of getting grant funding, or sponsorship (in kind) from a Government department, or what-have-you. Little guys don’t know how to do that.
That said, there is certainly room for improvement, and I’d like to see OS take a look at how they can make their data more accessible to users. Two suggestions in particular. Firstly, provide vector data in parseable text format (XML or CSV or whatever) as well as just shapefiles: shapefiles make sense for OS’s existing customers, but are a bit, well, you know, paleo. Secondly, reverse the recent change from 10km tiles to 100km. 100km x 100km of VMD is so huge that no browser has a chance of loading it, whereas 10km x 10km can be parsed clientside.
Richard
I think we are at cross purposes.
I was not disputing that OSOpenData(TM) has been a big success in terms of takeup. I know that loads of people have downloaded the data and quite a few have actually made use of it, particularly the CodePoint. No doubt there will be some commercially viable apps or services built on OSOpenData(TM) but whether they will generate enough economic activity to cover the cost to government of releasing the data remains to be seen. However commercial apps are not the main point or the big gain that we all got from Gordon’s Day when he unlocked the minor jewels from the royal jewellery box. The big gain for me is the democratic gain from enabling anyone to visualise and analyse all of the other data that has been opened up, that’s the non (TM) OpenData.
My scepticism was directed towards the argument from Jonathan in his article (referenced in my post) and comments here that if only government would make “the really good stuff” available the startup community would start to flourish. Jonathan is not referring to OSOpenData(TM) (I hate that TM and am declaring war on people who TM ordinary english because they remove a space or two), he is focussed on transport and other government departments data.
A question:- How long do you think it will take before OpenData based startups will be generating tax revenues that exceed the costs to government of opening up the data?
Cheers
Steven,
I agree with much of what you say, but I think you – along with many open data advocates – are perhaps slightly missing a crucial point.
I doubt much of the innovation, many of the opportunities for business, democracy, etc. will come from organisations whose central interest is data. I suspect the open data world will continue to be dominated by hip smartphone apps and snazzy visualisations.
For every ito and Placr doing useful things with timetables there will be lots of companies providing sustainable transport planning services who will be able to provide more sophisticated services using the data, and doubtless IT companies helping them to do this.
For every My Society and OKFN there will be lots of voluntary and community organisations working on neighbourhood planning and tree conservation who may have the skills, or the skilled support, to make a more sophisticated case for the causes they are advocating.
Those organisations won’t be attending hack days or unconferences, and except through initiatives like MADWD and some good journalist their work with open data probably won’t surface.
The problem with the data dumps (such as data.gov.uk and data.london.gov.uk) is that they only deal with two problems: they make the data available and remove restrictions. But organisations who might find the info useful usually don’t know about them and don’t have the skills to make use of them. What’s more, most wouldn’t be able to afford to additional cost of paying intermediaries to help them use it, so you’re back to square one – practical barriers prevent wider use of data.
Tom,
Great points.
I too doubt that “hip smartphone apps and snazzy visualisations” will prove to be a sustainable long term business model.
In your last para you sum up a dilemma – raw data is not information let alone intelligence, however it takes time and expertise, i.e costs money, to get much from the raw data and those costs don’t go away even if the data is free
Jonathan
Good response, thanks
I am not sure that I agree but I think this is a conversation that would be better conducted by voice rather than keyboard
Will try to write my positive side of OpenData before I go on internet free holidays
Cheers
Steven
“Stay sceptical” I tweeted the other day in reply to your critique of open data business models. I need you going “nih” in my earhole to keep me focused. But your nay-saying misses the real barriers and so cannot see where growth can come when they are removed.
Barrier 1. We still don’t have the good stuff. Without access to postcodes, addresses, school catchment boundaries, planning applications, company records, weather data, train timetables/ departures etc etc, the key open data opportunities are restricted. Opening these data sets up would give Placr (and everyone else) much more lucrative margins on which to build a business.
Barrier 2. App markets are immature and still not consolidated as they are flooded with vanity publishing projects. This means there are too many apps chasing the available income at the moment. There are signs of some apps being abandoned given the absence of a long term business model (new purchasers cannot fund the installed base forever), and this will increase margins in the market. Placr cannot earn enough by app publishing in this market at present, which is one reason that we have decided to wholesale data to other providers in an attempt to drive up margins on the core data-as-a-service product.
Barrier 3. Open data distribution needs to use the open source stack to minimise costs, but many organisations that we see still have lots of legacy systems that cannot easily consume data-as-a-service through an API. We have just designed and built an API (Pearson Plug and Play, launched this week) to allow content to be harvested from corporate databases and then served out to apps and web services. We hope to develop this as a line of business, but the very fact that it is necessary tells you that the market is not yet quite ready.
There is also the challenge of identifying value to business and consumers from commodities that in many cases have never previously been traded… and it takes time to experiment (and fail sometimes). So we all need a little patience to see how this will pan out. Keep on being sceptical because it is still not guaranteed to work out… but we can also use support whenever you have spare positivity to spread around.
Jonathan
https://www.placr.co.uk/gallery
Hi,
Plus ca change – when the 2001 census data went from bonkers expensive ~£100K full unlimited use, to completely free everyone was running around speaking at conferences about how there would be an opendata (TM) revolution (TM). Not much happened.
A counter example is how QAS built a 100M business off the back of new licensing arrangements around the PAF.
The difference in these examples is how one grew into a set of solutions that solved an existing huge, expensive and pervasive headache for businesses – address management, the other not so much.
Blair
ps. Did you notice OS, the Met Office and Land Registry all got moved under BIS two weeks back?