Pierre Pezziardi
Transcript (Translated)
So 80% of the job market is a hidden market. Yet, 99% of Pôle emploi's energy, and moreover, you see it in a very structured way, you arrive on the site, the Pôle emploi homepage, it's thus a window on 20% of the market. You have job offers. Now, if you are job searching or if you have loved ones who are, since job postings only represent 20%, you can probably suggest something else to them. Instead of looking for job offers, look for companies that are going to hire. Okay? And send them phone calls, spontaneous applications. And so, they will tap into this hidden market. 80% of the market. So, we created a state start-up called La Bonne Boîte. A state start-up is a small team, it's 2.5 FTEs.
La Bonne Boîte for Pôle emploi involves big data, it searches a past database of hires, hiring flows, because you know that in a very Jacobin country, we have very nice central databases, since each of you, for employees, well not business owners, but employees, you are all a line in a database where there is your social security number, the company's SIREN number, your start date, your activity code. And so we can exactly reconstruct the past flow of hires and you know what, there is a theorem that is quite... simple which says that those who have hired will hire again provided they are in good health, and this theorem is extremely predictive, it predicts with 80% accuracy so that is to say, if you go to La Bonne Boîte, you type forklift operator in Auvergne and you will see that there is Amazon, etc., which has a whole bunch of warehouses, hello and welcome, and
And these results, 8 out of the 10 results that will be displayed, 8 of them would welcome you very well since this model works.
What is amusing in this story, why I dwell on it, is because it crystallizes a bit of everything that makes state start-ups. The first thing is... Me, this story of the hidden market, I didn't know it, but I realized that at Pôle emploi, everyone knew it.
Everyone knew it. But everyone stayed in the rut of job postings.
People had internalized failure, the fact that we couldn't...
It's a fatality. We can't tap into the hidden market since it is hidden.
While the employment problem, we've had it for 30 years. It's a real concern. The big data I described to you, using a hiring database, is a linear regression, so these are algorithms that those who have done a first year of science master. So there is no particular innovation, the data has existed for 15 years. So we have a major problem for France.
Data that exists, that was available, a simple algorithm, a linear regression, and it wasn't done.
Why? What is it that meant it was never done? You see? I don't have a single answer for you. I think that structures, organizations create their own myths, their own limits, their own taboos.
They are resilient in reinforcing the rules, the practices, the customs that saw them born and grow. And so we have this form of resistance. A rut mechanism that makes it very difficult to get out of ruts. That's why there is a very important point in this state start-up method, since to connect it to the theme that brings us together today, Incamban, what we want is to streamline innovation, to streamline radical innovation. Today, the time between... The time between visualizing a problem, what I told you, but we have nothing to see the hidden job market. There is no tool in France that allows targeting these spontaneous applications. And this problem is the moment when we put a service online for users. For the state, the average time is six years. To do something that doesn't work in the end. So our subject is State Startup. It's to streamline this innovation. That's why we will take six months.
And why do we succeed in this portfolio? Since today, the portfolio that you will find on beta.gouv, you will hear about la bonne formation, it's a good twenty startups.
For all of them, we deliver within six months. So, what happens? The first new fact... What I'm telling you is that if you really want to innovate, you will... In fact, innovation is iconoclastic. It breaks rules, it breaks customs, idols. That's how you recognize innovation. It screams. People scream. They say, 'But it's not possible!' 'We've always done it this way!' 'You'll never make it!' 'It's shameful, you're going to cannibalize our market!' 'You're going to kill the poor!' It must scream. Otherwise, it's not innovation. It means you're not breaking any rules. Now, if you try to innovate within the very structure, within the IT department, within the business line, within the business as it runs, for this very reason, it will be impossible because you will face too much legitimate resistance, moreover. To your innovation. And so the first good idea we had is this.
The step aside. The step aside. That is to say, we tell ourselves that, you know, to innovate, let's innovate outside the structure. La Bonne Boîte was not made in Pôle emploi's IT department. La Bonne Boîte, the project ownership, is not Pôle emploi's project ownership. La Bonne Boîte, the server, is an OVH server.
Okay? La Bonne Boîte, we hacked the database I told you about because we signed directly with the director general to avoid dealing with all the lawyers who would have said it was impossible since this database was intended, as usual, for control.
And there, we... We repurposed it to make a prediction that is useful to you, to help you find. So the first idea, if you want to streamline innovation, by virtue of the theorem that innovation is iconoclastic, that you will face too much resistance if you try to do it right in the middle of the business, radical innovation must be done on the side. On the side.
And once again, I'm not describing a world where there are nice innovators who are prevented by nasty conservatives who don't want the rules to change. That's not the world, actually. There are some who want you to believe that, politicians in particular. But the truth is, there are two legitimacies. The innovator is legitimate because the guy who created La Bonne Boîte was legitimate in telling you, as I recounted, we don't have tools in France for the hidden market. But inside, if you will, the main orientation inside is not to promote innovation, it's to keep... the system running as it is today. Imperfect as it may be, we keep it running as it is today. So the main orientation is definitely not innovation. I felt it, if you want, I was IT director at BRED, I still remember it, it's so obvious, you're really on the side of the Empire, that is to say... They tell you that you are the CIO, oh Pierrot, you then have both the studies portfolio, new projects, and then you have production, you send billions in payments to pay the unemployed, etc. Well, the flows, the bank, what do you think? When they tell you, Pierre, your innovation project is two and a half years behind. Yes, but it's not serious.
You never have to fire a CIO for that. However, the day there's a small file with a thousand payments to unemployed people that doesn't go through on Friday evening, and as a result, the unemployed aren't paid on Saturday, and it's a Pentecost weekend, and so on... But then, you're in the CEO's office, and he tells you, 'I'm going to appear on Claire Chazal's show on TF1 on Sunday—you need to do something!' You see? So, it's important to understand that the main orientation and objectives of a structure are first and foremost to keep the system running as it is. And therefore, any innovation is perceived as something new that could jeopardize this primary mission.
So, the great theorem of what I'm telling you here is the dual legitimacy. You have to preserve the systems as they run, the legacies, the ways of doing things, imperfect as they may be, but...
I imagine you're here in this room of experts to make continuous improvements to these legacies, step by step. But this continuous improvement will never challenge the core principles of the system. After continuous improvement, we can cut the cost of servers managing job ads in half. We can create a better search engine for job ads. But you'll still be in the job ads business. And if you want to do something else, if you think the world is different, you take a step aside. And that step aside is very important. It means you cut everything off. There's no longer any support function, you're not dependent on marketing functions, you're not part of those functions... Because the slightest attachment to the structure, and you'll be absorbed by 'yes-buts,' which correspond to the legitimate thinking, once again, of the center of the Empire. So, the first theorem is the step aside.
So, creating zones of autonomy—and that's my job—to create zones of autonomy within the State, meaning finding...
Contracts without breaking the law with OVH, being able to hire two developers for six months to support a product manager, bringing in a coach, etc. So, building small teams of two to four full-time equivalents and asking them to launch something in six months.
The second good idea we had, still with the aim of facilitating innovation, is for things the State can't manage to do. The State is burdened with problems that have been around for hundreds of years.
Written by Louis XVI, saying it's still a pain for public procurement, for royal markets, etc. We create bureaucracy, etc. Even back then, bureaucracies were created for royal markets.
Today, 200 years later, it's still the same thing. I don't know if some of you have dealt with public procurement, but well, we have... I see some of you smiling. We wonder why it's become so complicated, actually. Why, when I respond to a call for tenders at Société Générale, I send an email with a PDF and say, 'Here, you'll find my response attached.' And when I do the same thing with the State, well then, I need an electronic signature, I have to fill out four forms, I send DNA samples, a urine analysis.
I'm barely exaggerating. So, we have recurring problems, and these are actually the business models of all State startups.
The interest of this initiative is that we have the chance to work for what seems to us to be the public interest, meaning solving the real problems of our fellow citizens. I'm proud to have set up the right tool. There are over 500,000 people. We did statistics on cohorts of 100,000 people, which allowed us to detect a slight difference between those who use the right tool and those who don't. And we have a... So, that's very important. First order, what's the impact? Was it worth it? I can tell you, yes, I have plenty of hits on my site, but that's vanity, it doesn't mean anything. The goal is to increase the employment return rate. And so, when we looked at cohorts of 100,000 people, we observed that the impact is greater than 10%. That is, in the employment return rate at 5 months, for cohort A, which didn't use it, compared to cohort B, here, they are 10% more in employment after 5 months. That's huge, actually. It pays off, if I consider... An unemployed person costs society 3,000 euros per month. So, just considering that they've gained a month of unemployment benefits, it pays for the entire incubator. And it has impacted tens of thousands of people. So, we can have a social impact, and that's the business model of all State startups. Understood? It's not about profitability; there's no capital. You see, you've understood that a State startup is simply recreating the in vivo conditions of a growing SME that can fail.
And with a business model focused on solving an irritant.
The pitch for a State startup is actually very simple. And you'll see on the beta.gouv site if you're interested, if you need to be concerned with radical innovation.
There's a section called 'how to write a product sheet.' How to write a pitch, actually. A pitch always has three parts.
First part, the drama. Every year, most unemployed people don't know how to route their spontaneous applications. Every year, millions of SMEs self-censor when it comes to public procurement. It's too complicated. Every year, 40% of people don't claim their social benefits. 40% of French people are entitled to things and don't go after them.
5 billion euros in unclaimed benefits. And they talk to us about 200 million euros in RSA fraud.
So, we start with a drama. Then, the second part, the idea: what's the new development? Thanks to my XML terton-chmure technology, or thanks to this website, this digital thing,
we can change the world. And the third part is: what's my strategy for... growth, deployment—that is, how I'm going to do something, how I'll have my first user in six months, because that's the rule.
When I tell you an SME can fail, I really need to recreate those conditions within the State itself. So, I have to tell people, 'Listen, this team we've set up—in six months, if you don't have users saying they're satisfied,' you're going back to your old jobs. We don't build software; we build satisfied users. Otherwise, we don't settle for vanity metrics like... Ah well, thanks to my tertant schmurt XML technology, or whatever, thanks to this site, thanks to this digital thing, we can do it, we can change the world. And the third is how can I, what is my growth strategy, my deployment strategy, meaning how am I going to do something, how will I have a first user in six months, because that's the rule.
When I tell you that an SME can die, I really need to recreate these conditions within the State itself. So I need to tell people, listen, this team we set up in six months, if you don't have users who say 'I am satisfied,' you go back to your old jobs. We don't manufacture software, we manufacture satisfied users. Otherwise, we don't settle for vanity metrics like... So one could do it.
And so these conditions create a product sheet, if you will. So the drama, the new fact, the good idea we had, hey, we're going to allow SMEs that respond to simply enter their SIRET number, and we will retrieve all the supporting documents from the State's systems, we will make a gift package and put it in the public buyer's system. That way, the SME just enters its SIRET number and its proposal.
That's the second block. Third block, deployment strategy. Well, in six months, the entire French public procurement system will not be transformed into MPS, into simplified public procurement. We're not magicians either. Radical innovation, you have... There's Uber, there weren't 60 million French people who had it the day after Uber's launch.
You always have an early adopter strategy. You have an excellent marketing book called Crossing the Chasm that will explain this very well. It's about first going after... Friends and family and the early adopters. That is, after six months, we had... So, you always have to go after the volunteers. You shouldn't go after people, especially the biggest ones, who will tell you what's the point. And so, how do you get your first satisfied users after six months? And what is your growth strategy to go from 10% market share? Today, we have between 15 and 20% market share of public procurement that is now simplified, that is simplified public procurement. But that means that still 80% of the markets you face are still paperwork. It's still not satisfactory, if you will. And so the growth strategy of these startups is always a real issue, meaning what leverage effect can we count on. When Zuckerberg launched Facebook, he launched Facebook and was constantly looking for leverage effects. So he said, hey, I'm going to rely on schools.
I'm going to promote it in schools where I'll have ambassadors, etc. So, what do I do here? Hey, I have exemplary pilot communities and then I'll hold seminars. And then, I have marketplaces. What is my business deployment strategy to take this startup from my early adopter population to a national rollout?
The new fact for the State, though, is that's why I somewhat circled beta.gouv, it's not nothing in fact. Because what I'm saying could be interpreted as an inequality problem. There are plenty of people who will tell you, 'No, but wait, sir, it's not possible.' You're creating a... We do that often in the national education system. It's not possible. You're creating an inequality between those who have MPS, who are entitled to this added value, simplified public procurement, and those who are not. It's shameful.
So it's a truly new way of thinking to be able to say, but massive innovation across the entire French territory in one go, it's not possible. So, for the egalitarians, it's a tragedy, in fact, because it's just impossible to convince everyone, but even to... You've seen, if I remove half an hour of Latin, I have 500,000 people in the streets, you know.
So necessarily, if we expect innovation to be massive and uniform, it's precisely the opposite of the definition of innovation.
It sneaks in, it fails most of the time. When it succeeds, it grows, it takes market share and it cleans everything up. So the State has admitted... It's not nothing as a stance to admit that not only will we be able to create public services that are better in some places and then quickly deploy them,
There's a second aspect that's interesting, you know.
I talked about royalty earlier. The State has always been about grandeur. I am the State, you are my subjects, I provide you with a public service.
And you see that a beta approach is a much more humble approach. It's a low posture. You know, the banners we put up, it says here the State in off mode to improve the public service. Help us, by the way, if you find any bugs, don't hesitate to click. We seek people's contributions. So it's necessarily a much more modest posture of a servant State. More than a State, how to say, grand.
That's the context. In this step aside, there's a second theorem, well, an empirical lemma, let's say, that I wanted to share with you.
Innovation, often, you've probably seen... It translates to, we hold a seminar in a company, then at the end we say, 'Yeah, we have a good idea.' The good idea is that we're going to create an idea box. And we're going to put all the important ideas that we should implement in it. Okay? And of course, it smiles, because you all know it doesn't work. It doesn't work.
People who make lists and lists, I've seen them, you're going to see a lot of them, lists of things to do, of stuff that would be good to do. It would be good if it were simpler, it would be good if we had finished... Etc. So, searching for innovations is futile. It's stupid, it's futile. Because what you need to look for is not innovations, but innovators. What you need to look for are people who not only have an idea but who have the desire deep in their guts for it to become reality.
Why? Because despite everything I tell you, despite the step aside, etc., innovation still requires stubbornness. It requires profiles with strong commitment, who have decided to make a matter a personal affair. Okay?
The product managers of State startups are men, women, young, old, that's not the issue. But they all have one bias, though, they're all indignant. They're all people who tell me, 'But it's really not possible that...' That drama, you know, from their product sheet. They all have it deep inside. And for us, it's very important.
So, to recruit innovators, not innovations. But then, follow the thread. That means, in that case, we set aside all the IT professionals in companies, in administrations. Not one of my product managers is a statutory project owner, a guy from the inner circle, normally the person designated to create innovations. You see, in an organization, it's set up as a specialist in modernity or computerization, IT departments, project management departments. And well, it's never them.
Instead, the right person is a 'toothless one,' as the other would say. It's a Pôle emploi advisor in Hayange, National Front town hall. A Pôle emploi advisor in Hayange, National Front town hall, in a small agency. So how did he get there from there? I didn't know him.
At the center, working on a product with a team of 2.5 FTE. How can we manage to find this person? That's what it is... I'm telling you, we won't use the pyramid, the hierarchy... No, I'm giving you some... Because he didn't come through a public tender, he's an employee. So, he remained an employee of Pôle emploi.
Well, it's always the same. We make a call for volunteers.
So, we have a standard format, in fact, since it works the same way every time. What's needed is to send it from... Who is speaking in this invitation? It's the CEO. So, Jean Bassère sends an email saying... Dear friends, you are probably witnesses to regular irritants for users or agents. You remember, the irritant, we need to find the business model. It's not, dear friends, you have probably observed this new technology that you would like to implement. No, we don't care. We don't care. People who want to implement tablets or people who want to use XML or blockchain, I don't care. I couldn't care less. Besides, everything I'm saying is important, it's about focusing people on value. So, value is solving the irritant. And how are we going to measure it? How are we going to deploy it? So, you are probably witnesses to...
Imagine receiving this email, you're in a large organization. So, dear friends, you are witnesses to regular irritants that occur for users and agents. You have a good idea that involves digital technology. That you would like to defend and implement, you are ready to invest yourself in it. This is not an idea box. You are ready to personally invest yourself in it for 6 to 18 months. So come on that day, at that time. Imagine if you receive it—I don't know if any of you are in large companies—because we've talked more about large companies than small ones, in small ones it's already the case. I'm talking about a process that is of interest in large organizations.
So you receive this email, on the day of the hackathon, 120 people, which is actually too many, it's read at Pôle emploi, horrible, because we end up kicking 118 out, so we took two. But that's where you manage to recruit these people who are freaks, odd people, who have a form of indignation, an idea. Who are not IT specialists, so you will need to surround them with digital skills to build a team. So generally, when you recruit someone like that, what they will lack is one or two full-stack programmers and a lean coach, that is, the person who tells them here's everything you can remove that serves no purpose in your critical commercialization chain. And who constantly applies lean methodology, all the time, all the time, all the time. Why are you going to this meeting? How does it help your commercialization of your first unemployed people? Why are you doing this? What are we going to—everything we can remove, and we realize we can remove a lot, meaning things like the right box, we put them on hold after three months. Imperfect, always, it didn't work in the overseas territories, etc. But that's not a problem, you're not commercializing in the overseas territories. You make noise around you.
So the second empirical lesson: look for innovators, not innovations.
And that has a very particular power, because what I'm indicating today, we were talking about organization, flattened organization,
the fact that it designs a world... Where we have disregarded hierarchies, established balances. We said the person who has the energy and the idea has the power. That's what we're saying. And we give them power. When I sold the innovation approach to Jean Bassère, CEO of Pôle emploi, and it's no small thing that he accepts, because I tell him, listen, Are you satisfied with your innovation at Pôle emploi? No.
Are you ready to try something else? Yes. Well, I have a process that is interesting, it's really great because it's not expensive.
We can fail, but if we fail, we lose 200,000 euros. Because the team I described to you costs less than 200,000 euros for 6 months.
There's just one problem: we don't control the subjects. Since we're looking for the innovator, not the innovation. And he accepted that. So it's very interesting to think... In large organizations that are big fans of control,
That's why, by the way, hierarchy becomes so complicated, coordination efforts and the share of added value devoted to control and coordination only increase. It's because we are really control enthusiasts, and that becomes the real problem of large organizations. So this thread, if you will, and I'm going to reveal the goal of Beta.gouv, the goal of Beta.gouv is not to create 25 products and make noise by saying we're doing digital.
It's not that, it's a societal ambition to say: what prevents us from structurally implementing this mechanism,
and thus making thousands of autonomous, self-organized teams emerge, led by leaders who are not necessarily part of the hierarchy, but who derive their legitimacy from what they imagined, from the drive they had.
So, you imagine that the first...
Important fact: this scales. It scales very simply.
Today, at Pôle emploi, we hold a competition every year, with two startups starting. So there are two startups starting per... We could double it, and it would still work.
But you can well imagine that we can do this at the Regional Council of Burgundy.
At the Ministry of Agriculture, at IGN, at INSEE,
Everyone has problems, you see. Everyone has problems, and the State is the first to need to reform itself. But deep down, this goes even further. It goes further because...
We create small autonomous teams. But they are somewhat homothetic to what happens on the web. You know, what is very inspiring about the Internet is still Open Source, Wikipedia, Linux. That's what's super inspiring, it's that big things,
that make big products, Linux is still no small feat, it's one of the largest code bases used. Wikipedia is no small feat either, it's an encyclopedia, I mean, it has its problems, etc. But it's no small feat. And they organized themselves with a DNA that had nothing to do with what you can find in large organizations.
Which was based on prior control and specialization. Here, we have the opposite: we have subsequent control and generalized mutual aid, despecialization. And all our State startups have a lot to do with that. That is, creating an autonomous team and... open, okay, and self-organized and open to the outside, which means that, in essence, they will tackle the root causes of organizational problems. that she has her problems, etc. But it's not nothing. And they organized themselves with a DNA that had nothing to do with what can be found in large structures, which was based on prior control and specialization. Here, it's the opposite: we have subsequent control and generalized mutual aid, despecialization. And all our state startups have a lot to do with this. That is, the fact of creating an autonomous and open team, okay, self-organized and open to the outside, means that, in essence, they will tackle the root problems of organizations.
These are the attributes of large public and private organizations, which involve the weight of control and coordination linked to the siloing of all departments and the fact that the marketing department clashes with the sales department, which clashes with the back-office department, which clashes with the accounting department, and we coordinate.
And we create big things, we create the national education system. And we go in circles with these things. And where our society... We reproduce... Customs in which, ultimately, the room for maneuver for each person—I don’t know if you were a teacher, in schools or advisors at Pôle emploi—your mass margins are very narrow.
In these industrial organizations, we truly replicate Taylorism, even in the service sector.
And so this is happening in companies while... At your home, you have room for maneuver and action; you can contribute to something meaningful, like Wikipedia,
And here, inside, you have something that is much less meaningful, which is a local optimum of our sub-hierarchy. You have to minimize the unit costs of the parts we buy because we are in the purchasing department.
This doubles the number of accidents and vehicle returns. But that’s not a problem. We have the local optimum, so we...
I’m trying to show you a reality that perhaps the broadcast highlights, which is this somewhat harmful territorial atmosphere that you must find in all large organizations, that you must endure.
These industrial horrors we see, sub-organizations polarized around sub-objectives. We will maximize the baccalaureate success rate. There are 80% of baccalaureate graduates, that’s great. But actually, we don’t care; that’s their order. There are 25% of school vouchers, 25% of kids who enter and leave without... That’s the indicator we need to follow. We need to maximize the number of arrests of hashish smokers because it makes lots of numbers, lots of arrests, and also resolutions. Have you felt safer since we’ve been arresting guys who sell joints?
So, we must minimize the costs of purchasing parts. But which then destroy, which cause problems with efficiency and vehicle returns. So all large organizations have these problems: local objectives in silos, and very little collaboration with the rest of the company and even less with the outside. And new organizations, the web, are very inspiring from this point of view, but we can interact with the outside, And we can create systems that accept and precisely... truly welcome the contribution of everyone. If you go to one of the state startups, I was talking about non-recourse earlier, with 40% non-recourse, meaning people don’t know the rights they have. No wonder. To know your rights, in fact, you have to go to the CAF, to the CNAM, to the CNAV. As well as to Pôle emploi, and you also have local aid that can be from the department, the region, etc. So in fact, there are many places where, if you want to know what you’re entitled to, including the national education system for certain scholarships, and I’m forgetting some, of course. So social security today is insecurity because... To find your rights, you have to read texts, go through very complicated things. And you’ll go to the mesaides.gouv site, and there, in one place, you say who you are,
and if you live in Paris or Seine-Saint-Denis, and you have very low income, etc., you’ll even see that it will tell you the national benefits you have, but also extra-legal benefits, meaning those from the Seine-Saint-Denis department. Or aids like Paris Logement Famille, because you live in the 75014 district, and so it will activate, you’ll see that you’re entitled to Paris Logement Famille. Why am I talking to you about this? Because we were on the chapter of open organization, contribution. Managing a system like this is impossible. That is, if, at the next deliberation of the Paris Council, where they will... Raise the threshold for Paris Logement Famille. My teams, my aids, it’s two FTEs, have to say ‘hop, hop, hop, hop, hop, hop, we’re going to change this, we’re going to change the rule, etc.’ It doesn’t scale because there are social aids in Brest, Rennes, Montpellier, in Burgundy. And so you understand that if we want to organize this social security, we need to find a way to collaborate with all these people. So that when this famous deliberation takes place, it’s not my problem, but the problem of the city of Paris, to contribute to an open-source software, which represents executable law, which is in one place, and towards which all converge... The algorithms of all rights, whether they are regional, local, etc. And that’s what we did. So what does that mean? It means that the code is on GitHub, of course, that there are manuals, that there is even what we call, which is a big word for some, a conversion tunnel for contributors. A contribution is code, but it’s also tests. There are as many tests as there is code. And so we are very proud that people in the basement in Melun, they enter the conversion tunnel, they say ‘click, at the end of the simulation, in the app itself, saying ‘there’s this,’ there’s a little button called ‘ if you see that this simulation is wrong, please let us know.’ There, you click, it creates an automated test.
It sends it back to us, we created a social network, I don’t know for those who know, fitness, things like that, automated tests. We made one, it was so important that we recreated one, we called it Ludwig, Wittgenstein, the great philosopher who extensively discussed the abysses of meaning. A test makes things concrete. And so, with one click, you see, you’re in the app.
I’m going to catch you. You, because you have knowledge that I want, in fact. You might be from the basement in Melun or you might be the one in the neighborhood who helps everyone with social benefits, so you know. And if you have knowledge, you find that this simulation is wrong, boom. And there, it creates an automated test.
But what does it describe? Very concretely, it describes an organization that is interested, that starts to breathe with its ecosystem. You’ve never seen this on a pot.gouv, you’ve never seen this on Ameli, on other public services.
So it proceeds from a different posture. So now, I’ve moved into the social realm. What’s really happening at the heart of all this? All this is not about IT. Our subject is organization. Our subject is even value scales, since we’re saying that small artisanal teams create more wealth than a bureau chief or a coordination director, etc. There is a... The great pride we have is that the Pôle emploi advisors who came and created products, after 18 months, we implemented an HR rule: they actually become managers, they cross the glass floor. We recreate value through craftsmanship. Today, in a large company, doing is not noble; doing won’t advance your career. You must very quickly extract yourself from the iron to find some fists that will do the work so that you can say you have lots of people around you, and that’s how you’ll move up. And that’s the only way to move up in a large organization today. That’s what makes... Agile practitioners, artisanal IT workers, great coders—there are ultimately few in our ranks. And large IT departments are made up of Excel project managers who struggle to create much value. But we can’t blame them. They followed the signals of their organization that ultimately instructed them to take this type of position if they wanted to advance their careers. There was no career for master craftsmen capable of bringing products to market, meaning having broad IT skills, capable of not being afraid of data production to the... to the frontend, as well as being able to conduct conversations with users, since that’s what the methods require. Incamban or Agile, it’s still about having interactions with users, so a form of empathy, so that level of maturity. These master craftsmen who are 30, 40 years old are very undervalued. So I’m talking to you about a world where precisely this movement,
with a re-internalization already of IT know-how that the State has completely lost. And this re-internalization is in fact accompanied by a revaluation of these professions. There is a beauty in craftsmanship.
In conclusion, should I conclude? Yes, we will leave some room for questions. The world, what we think today, and we will likely have a new president in May 2017, what we will tell him is that Beta.gouv is a prototype. It's 25 startups, it's 25 teams, it's an extremely new way of organizing within the State. It overturns a certain number of value scales.
But it is highly replicable. Highly replicable. That is to say, we have shown that it can be deployed everywhere tomorrow. We can reserve what we estimated at 0.1% for innovation. 0.1% for innovation amounts to 500 million. This could result in approximately 5,000 people working in this type of team, solving problems at all levels. That is, these problems that we cannot manage to solve.
This would translate into things involving digital technology, things with less digital technology, it would create autonomous middle schools, it could do this kind of thing. Because in the end, what we believe is that this digital revolution is nothing less than a revolution of organizations and value systems. That is to say, we have systems of organization and values today centered on control. Division and control, mass production. And 80% of our value added to the GDP in France comes from services. 80%. And we have applied these methods to them. You have call centers, back-office experts, regional education authorities that supervise middle schools. This entire form of organization has inspired all our services. And yet, perhaps not the nuclear bomb system, etc., but most services can be redeployed, and we were talking about this with the experience of Mr. Ducadong, can be redeployed within autonomous collectives with other, more organic forms. Autonomous, artisanal, responsible teams that ultimately rediscover the meaning of their action through the autonomy created around them. This is what we have modestly tried to do with these initiatives, and I hope we will be allowed to continue this movement and make it a... state dynamic. There you go, thank you very much for your attention.