The French Government has launched an initiative called in June 2013 to provide digital services for the common good through a process that is intended to be more agile and hopefully more efficient. It has been labeled an incubator for “State Startups” in order to convey the willingness of facilitating the ideation, development, and launch of web services to prevent administrative frictions and facilitate the access to critical information to the general public. Since then, this incubator claims to have created 30 services, among them 6 that are highlighted as successes, thanks to the collaboration of 54 team members and a budget of €2.1 millions for 2016 (see 2016 annual report in FR) and a total budget since its creation of possibly €4 millions. I leave you to the interpretation of the outcome, after having checked the services that are live and would like to challenge the usage of this public financing to enable the development of the social services of the future.

What we know:
The French Government has been running for 3,5 years an initiative to build new services that would operate as a single window to facilitate the access to the relevant information from across all the government entities. To do so, they have spent a potential total amount of €4 millions of public financing to build a 54 people team divided in small units of up to 4 members to work their own ideas that would help the general public. The global budget is built out of a team growth of 8 people in Y1, 15 in Y2, 30 in Y3, 54 in Y4 and a smoothed yearly average fully charged salary of €39K taken out of the 2016 report. Moreover, the strategy seems to be focused on generating services out of the collective brainstorm of civil servants and the team members, with a low to none market testing before releasing if we accept they launched 30 services. If we only take into account only the ones that are in “consolidation phase” and “transfer phase”, that means there has been 15 services that have been released, and 6 are worth of highlights according to the annual report.

These numbers would state that the services that have been launched have, on average, received €133K of “funding” if we were to use the VC terminology. And 6 out of 30 have been “successful” if we accept that the fact they are highlighted in the report means they are reaching the goal that was required. Which is a good ratio. Yet, with €4 millions and 3,5 years of operations, we could have expected differently with another way of working and a mindset that is not focused on directly delivering services. Indeed, if we were to leverage and harness the power of the crowd, these numbers would potentially have been higher with the same budget. But it would have required proceeding with a vision that is disconnected to the one that is shared by the current leadership team.

The incubator for State startups focuses on shipping services to the end customers, like a simulator, or an API for a company to use them. As a non-tech user, the simulators would be interested if there were no already existing ones (like the simulator to identify the cost of a fully charged salary based of the gross salary) and as a tech user or company willing to leverage a database that is owned by the Government, I would love to be able to access more than what I can through existing services (let me not highlight any). All in all, we can assume the whole team is doing whatever they can to provide the best services to the general public. But how can you deploy so many services at the same time so little budget? A standard startup would work with work constraints that are out of the spectrum of what employees of the State can imagine, in order to deliver a service that have a very few chances of attracting end users. It is not the fact that it is a tool provided by the State that would make it massively used. And funding each project with €133K would only dilute the impact the project could have.

The option would then be to focus all the energy of this initiative into providing a single point of access for private companies to build on the data the government can provide, including an API related business model to enable more hiring and funds without sucking out the public finances. This means the Government would need to collaborate with private companies, or at least the incubators and accelerators it is co-funding, to build public services that would really be game changing, like the partnership between the French unemployment service Pole Emploi with the Bayes Impact. This way of working would create synergies between the private and the public sectors instead of relying on the ability of the French Government to deliver web services. That is definitely not part of its prerogatives and out of its capabilities and it is essential the leadership team accept this situation and collaborate hand in hand with the existing and successful startup ecosystem.

Practically speaking, you would need a 10 people development team to operate a real Open data platform that would provide APIs to any startups willing to work on generating added value for the general public. It would require an additional team of 5 to 7 project managers who would collaborate closely with the startup ecosystem to identify what they would need and how to efficiently release those APIs to be useful. It then makes a team of about 20 people to create a springboard for new and spot on services to appear. The one question would be the business model, as API platforms may quickly become costly. Would the Government subsidies the usage or price the access or enable a revenue sharing model. That would all depend on the priority that would be set.

Be it either a State government or a company, regardless of its size and financing capabilities, it is critical for the success of a venture to keep from overextending.
This means that a company would need to understand that in the case of the innovation, it can’t directly compete with light and agile startups that are focused in a single expertise. It would need to focus on assembling the expertise and completing the toolboxes that are needed for it to thrive in its complex market. Startups won’t outcompete large companies unless the latter forget this because once the former have successfully set up their operations in one field, they would extend their reach and from here on would become potential competitors of the incumbents. It is up to the top management teams to digest this and collaborate with the startups to use their infrastructure instead of struggling for a piece of a business they, in the first place, don’t need to own. Taxi G7 in France did understood this when Accor still does not.