---

Transcript of the Video:

Roger Martin: How would you describe the strategy that you’re following?

Nandan Nilekani: So, I think let’s go back to this issue of how do we get scale and speed? Now, one option that was there for us was to create our, look at the issue, we had to collect, because we wanted the uniqueness of every number the only way we could do that was using biometrics or using fingerprints and the iris. So we collect all the 10 fingerprints, both the iris and the picture.

What that meant was that every man, woman and child in the country at some point has to come to a location and give this information, so that’s a massive logistical challenge to get 1.2 billion people to come and realize that if we start building a infrastructure for doing this we’re going to end up creating a huge organization with huge costs even for employees and that’s when we came out with a partnership model and we’re filling an ecosystem. When we don’t do that what we call as enrolment but we do that through partners. We have 65 partners today. Partner like state governments, banks, insurance companies, oil company and so forth. And the idea was that all these agencies are already dealing with the people and if we can somehow insert our enrolment process and embed it in their workflow then we can get enrolments out of that, that was actually the sort of critical thing, but we provide the technology.

We provide the software and the platform so whether you’re enrolling a person in a village in Bihar or a bank in Bombay the software and enrolment are the same even though the agencies may be different. Even though we’re using all these diverse agencies we’re able to standardize on the way the enrolment happen.

So this is a very different way of looking at the problem of leveraging existing infrastructure networks and that’s how we are basically driving this and this is a highly scalable model because we can keep adding enrolment agencies and enrolment points. So today, we have about seven thousand points already where enrolment is going on and they all use the same technology, they all use the same trained operators. So it’s all the same quality data coming from all those locations.

Roger Martin: What kind of incentive did you have to provide the partners to do this?

[00:02:27]

Nandan Nilekani: Well, two incentives, one is that the partners also benefit from these ideas because let’s say our state government is in the business of providing subsidized food and fuel to the poor they needed this database to make sure that it reached the right person and that there was no diversion of that commodity, it went to the genuine beneficiary. So they value this database because it allows them to make their basic database of beneficiaries much more, you know, spruced up. So that’s the win-win. If you partner with us, you know, we’ll help that process.

In relation to that, we pay our registrars a $1.00 per person for enrolment so that there’s an economic thing also there.

Roger Martin: But boy that’s pretty cheap. I mean that’s a good deal though. I would think that the cost per person to enroll, set up an enrolment, like I was thinking as you’re early on saying, hmm, you got to get them enrolled, I wonder what the cost per unit just for that?

Nandan Nilekani: Actually, it cost about $1.00, about 50 rupees is what we pay our registrars and the back end costs us another 50-75 rupees. So actually for our cost per person is about less than $3.00 total cost. And also the other…

Roger Martin: That was very low. That’s fantastic.

Nandan Nilekani: You should tell that to my colleagues.

Roger Martin: Yes. No. Well, again, your problem is you have to multiple it by such a big number, right.

Nandan Nilekani: Yeah, it’s 1.2 billion people.

Roger Martin: Yeah, 1.2 billion people. I could do that.

[00:04:05]

Nandan Nilekani: And the other important thing again, it comes back to the model is our registrars, our partners they in turn give this business to enrolling agencies. You’ve got companies that specialize in doing that and those enrolling agencies invest in the equipment and the pay them only on actual enrolments. So it’s a completely variable cost model. So we only pay when people are successfully enrolled. Another different way of doing it, it doesn’t buying a lot of hardware, it was creating a operating model.

Roger Martin: And it creates, at least the other thing that seems very attractive to me is it creates very little infrastructure that you don’t need that you’re going to have to be paying for later. It’s almost self, kind of decomposing.

Nandan Nilekani: That’s right…

Roger Martin: Too bad everybody enrolled.

Nandan Nilekani: Because enrolling agency buys their equipment and they do “x” thousand enrolments they recover the cost of the hardware, so we don’t have to have the burden of hiring a lot of infrastructure. So it’s a light model. Again, it’s a model done from, you know, the technology world.

Roger Martin: And it’s great. I mean I love this sort of living advertisements of the opposable mind. It is a great example, do you want speed or scale and, you know, what I wrote in the book and partially from interviewing folks like Nandan is the reaction is no, I don’t want to make that choice, I want to figure out something different and what it takes us a brand new model which you’ve dually created and that’s what makes the difference.

But that instinct, I mean I think it’s just such an important instinct to just say hmm, it’s never been done before, everybody tells me I have to make this trade-off, I’m just not going to do it. Where do you think that came from in you? Why…

[00:05:59]

Nandan Nilekani: Well, you know, the thing was that I took up the job because I wanted to get it done, right. So we thought long and hard and all of us said how do we do all these things at the same time? How do we get, you know, scale and speed? And one of the commitments were made publicly is within the next three years we’ll enroll six hundred million people into the system. That’s a public goal. It’s, you know, everyone knows this goal is on our website, so that’s a huge goal to achieve.

Now, to get to that kind of numbers we have to reach at least a million people a day have to come, you know, somewhere in the country. A million people have to walk into something somewhere and do this and do this and all that stuff. So that was the thing and I think the only way we could do it was what, finally there’s only one way to do it which was the solution that we came out with and we also had said that we launch within 18 months. We launched in 14 months.

Again, it goes back to the fact that, you know, speed is of the essence in something like this, that you can’t, you know, linger on for months and years, you need to get on the job and already we have enrolled about 7.5 million people and our current enrolment rate is at about 200 thousand a day and we expect to get a million a day by October. And it’s essentially by adding enrolment stations it’s scalable in the front end. By adding servers it’s scalable at the back end. So it’s scalable both front end and back end. It’s down to how well you architect and design the solution and how you leverage, you know, everybody in the ecosystem.

Roger Martin: But I still want to know why? Can you go back and think about, most of these things never happen, right. You don’t have Indians uniquely identified or it takes two decades because most of the time the people in charge say well it can’t be done, it can’t be done any faster than that, that it will take 20 years to do that or we can’t do it in a high quality way or it will take $50 billion dollars and we don’t actually have $50 billions dollars to spend. What got you to the place where you were to walk in, put an audacious goal out there 600 million, we’re going to get it for 250 an enrolment?

I mean, I’m not sort of saying, you know, tell me it’s because you’re wonderful which you are, but I’m more interested in how did you get to the point where that’s the way you looked at choices that you seem to be given because I think that’s what people might find very valuable? How did you get there?

[00:09:00]

Nandan Nilekani: No, I think this goes back to, again, I think how we do the envisioning and strategizing, right. I think if you start with a goal and work backward from the goal then you’re forced to come out with a solution that gets you to the goal. Now, you know, it is very clear that we had to get a substantial number done within five years and 600 million was, you know, as good a number as any. Now, then when you say how do you do 600 million and you assume it takes you a couple of years to build this stuff then it has to be at the rate of a million plus a day otherwise you won’t get to 600 million because let’s say 200 million a year for three years.

And then if you say if you want to get a million a day and if an enrolment station can do 25 a day that means we need 40,000 enrolment stations. And then you say I’m going to run our organization with 40,000 enrolment stations, I’ll allocate this gigantic bureaucracy and then you realize no, maybe that’s not the way to do it. Is there a way to do this in partnership so that you can leverage other people’s presence out there and take advantage, so that becomes the next conclusion. Then the question comes if you’re going to use other networks to enroll how do you ensure that they all do it in the same way? So then the logical conclusion from that is you provide the software. So, you know…

Roger Martin: You just work back.

Nandan Nilekani: You just work backwards from the goal and finally the answer comes out of that. So you work backward from a goal and that drives the answer.

Roger Martin: So what other things could you do in India, working backwards from a goal, when you’re done with this?

Nandan Nilekani: That’s a different question.

Roger Martin: Not a good question?

Nandan Nilekani: No, I don’t think that.

Roger Martin: No. Maybe.

[00:10:54]

Nandan Nilekani: But let me tell you something else. The other thing is that the point of this is certainly to give the ID, certainly to make sure that people who are poor and marginalized who don’t have IDs become part of the system, but it’s also about creating an innovation ecosystem on top of this. That’s I think the other big thing which we are trying to do which is that after we give people this ID, this ID will be available as an online ID which means I can authenticate a person’s ID online and we make that available as an open API, as an application programming interface which any application can imbed on their application.

So suppose you’re a banking application and you want to verify the ID you can verify with us. Or suppose it’s a Facebook application, they want to verify ID they can verify with us.

So this whole idea is to create an open infrastructure where people can now build applications which require ID verification and that is really the long-term goal of this. I mean the short-term goal is to get everybody in, but as you provide these kind of online IDs you can use that to re-engineer public services, banking, telecom and that we see as a real game changer in all this.

Roger Martin: And it seems to me you’ll end up leap frogging a lot of countries that are supposedly more developed.

[12.32 minutes]

---

This video was filmed at the Rotman School on December 13, 2011.