By Pradyuman Maheshwari
Instead of doing a drink or a lunch at a five star coffee shop, we decided on the headquarters of the Broadcast Audience Research Council in Parel. The office of the CEO Partho Dasgupta isn’t closed. Complete transparency assured, he laughed as he dived into some Pav Bhaji. Pav is something he loves, he told me, although he hails from Kolkata. The last two decades in Mumbai have made him a convert. Excerpts from an interview:
It’s interesting you’ve the Rubik’s Cube as the mascot for BARC. Any significance?
The industry has been posed with this puzzle of solving the audience measurement and numbers issue for so long. There’s a big puzzle to be solved and what better puzzle then what we’ve all grown up with – the process of solving which has been interesting, the Rubik’s Cube. With all its colours, dimensions, etc we thought it would be a good inspiration for us.
The issue with the audience measurement business – whatever be the media type – is that it’s always puzzle people cannot unravel with ease. The Rubik’s Cube is in a sense symbolizes that as it’s not easy solving the Rubik’s Cube. So are you trying to say that you won’t be able to join the dots very easily?
It should be signifying something which is very tough. We are trying to solve a tough puzzle.
One of the reasons why the new system exists is because in the earlier format, people reportedly were puzzled by a lot of outcomes. Isn’t that an issue, symbolically speaking?
Are you talking of puzzles being unraveled or the enigma that existed?
The puzzle and the enigma!
There are two parts. An enigma remains an enigma. We’re talking about a puzzle to be solved and we think we are solving that puzzle.
Symbolism apart, let’s get to the real issue and the million dollar question – are we on course?
It all depends on how you want to term it – on course or not on course. Let me tell you objectively where we are. We are fairly on course in terms of the four odd streams which on convergence will ensure that can start giving out the numbers commercially. The first is monitoring the channels that play out. That’s completely on course. The second being the placement of embedders which puts out the watermark, the technology we are basing the entire measurement process on. Roughly 250 channels have ordered and half of them have already been installed. We are looking at another 100-120 odd channels, which will come in and that also tells you that it’s completely on course. The third being the meters; the whole development of meters. Let me tell you this isn’t a very easy thing. This is embedded technology based on Intel chips that we are developing. This chip is new and hence it required a complete set of software and hardware development of all the other ancillary peripherals and stuff like that…which has all happened and they’ve already tested it out in the labs in different conditions. But now we’re trying to test it out actually on the field. From this month, the meters will be out in the field which will go up and up. We already have the back office ready to get the data back and see how the data is performing. So the third stream, the meter, back office and the related infrastructure is very well on course.
What about the others?
Yes, as for the fourth stream, which starts from establishment study data to the sample design to seeding of the boxes into people’s homes… here is where we have a small problem; rather a delay. This is because of the availability of data which we are trying to solve. Overall, yes we are on course. There’s a couple of months delay here and there which happens on a project of this caliber. It’s important to get it right, we’ll never want to commit a date; we’ll want to check, validate the data, be fully satisfied, get the stakeholders on board to understand this is the data and say this is good. Hence we will only announce it on a certain date when it will go live.
Would the disputed data be from IRS? You said there’s a problem with the data, is it the same IRS data which has been the problem?
I don’t want to name it because are a lot of data points we are using – IRS, Census data, NCR, MIB, TRAI data etc.
Looking back, would you say that it was a bad idea to depend on IRS data as your base data?
We’re still evaluating that. We haven’t come to a conclusion yet. Based on that outcome we’ll take a call for next year.
Coming back to the date of launch, when do you think can the industry expect it?
I’m not giving a date because I don’t want to make a mistake some of our other industries in similar situations have faced. I’m only telling you things of what’s going on as a process, which will give us data from the field as we go on to next month. When we see the full validated data fit for commercial use, we’ll announce and make it commercially usable. It would be towards the end of the year; around that time.
So will Santa Claus bring in a new audience measurement system?
(laughs) That’s for him to decide. Why should I?
I’m pushing for a date because there are people who’ve paid money for an audience measurement system. And then there was a government which was fairly aggressively wanting a new audience measurement system.
Uday Shankar, one of our directors, said at one of the meetings, “Understand we are doing something for a generation and not for a couple of years. Even if it takes few more months and few more iterations, do it right and at the right cost.† It would be very easy for me like to go with one supplier of data. Typically, the world over, people take more than three years to do such a changeover but it would have been so much expensive for the Indian industry to bear, which is not funny. As you know, we are talking about 20,000 meters that are needed based on the country’s size and it’s heterogeneity so it would’ve been stupendously expensive if we would have gone with something simple used by some other countries. We took the painful, tough path purely because we thought that’s the most wise thing to do based on both cost and the country’s condition.
Will you have 20,000 meters on Day One?
We’ll have more meters. We’ll report from 20,000 households, not meters. If you have a multiple TV viewing household, it will have two meters.
The government was exceedingly belligerent about wanting to have a new measurement regime. What has been your experience with the current government compared to the previous one?
It’s been nothing different. They are not pushing us for the date and things like that. We keep them updated regularly.
If a broadcaster asks you as to when do you think will the new system come into effect given there subscription monies to be paid to the old data system, what is your response?
Honestly, that’s a broadcaster’s own decision. The board knows where broadcasting companies are present and they know exactly what’s happening. So they have to take their own call on that and when they want to cease a subscription or start it is completely up to them I honestly don’t want to give any advice on what they do with the other subscription.
But clearly by this fiscal the new system will start, right?
Should be.
Are you worried about what happened with IRS?
It is a big learning but I’m not worried.
Let me elaborate. IRS was set up by esteemed members of the industry where members got together and selected a vendor. It was all done through various approved processes and when the result came in, because it wasn’t favorable to some people, it started facing issues. The same thing could happened with the BARC data too. You have people from the industry come together, appoint a committee, etc but since you can’t produce ratings or fake outcomes, are you worried that the broadcasters will not be mature enough to take aberrations or changes from the existing system?
I don’t want to speculate at the outcome. I’m concentrating more on the process now. My belief is the outcome should reflect the process. We’re trying to do everything possible and hence I’m not giving you a date or things like that. We’ll ensure all the different processes we’re talking about to ensure the output is right. I’ll take extra precautions to validate the data and carry the stakeholders along before we go commercial about it. At the same time I won’t speculate on the outcome. I’m more concerned about the process now so that the output is right.
What happened with IRS is that certain publications found that the data was not favourable to them. That could well happen here with BARC’s findings tooIn the validation process, how much would you want to alter in case if a certain channel was No. 1 earlier and suddenly falls to No. 3? We saw that when digitization and LC1 happened, there were some tectonic changes in the ratings. Is there a worry how the industry will take it given the fact that media people who otherwise try to sermonize from the rooftops are not very kind to being submitted to the vagaries of the world?
Let us cross the bridge when it comes. I’m not worried about it now. The bridge is being built on a very strong foundation. We shouldn’t be worried about the bridge falling very easily; people who are working on the bridge, know the bridge very well.
As an industry observer, isn’t it a genuine worry for all of us? Thanks to what happened with IRS, we didn’t have data for readership for a long time. Has this been discussed with BARC board on such an eventuality?
Not really in these terms but obviously we’re taking all kinds of precautions for that. The BARC board has been advising us how to take precautions from such issues. I’m reasonably confident things will be taken care of.
You now have 26 vendors. It must be quite a job to deal with the vendors and orchestrating the entire act together.
To typically manage the handshakes particularly since there were many technologies involved and a lot of international vendors too, the whole handshakes thing is quite a job.
Do you think it’s better to have it done this way rather than go with just one?
Absolutely, 200 per cent. It’s a tough path, but it would be any day better to do it this way than any other path you mentioned.
When you started, you selected the vendors. Has there been any change since? Have you looked at changing course if a vendor or technology doesn’t work?
The path remains the same. A standard evaluation process is trying out different vendors. We have evaluated three vendors for the same job. The path or the principle is the same. In this, the whole board and the technical comity  have been a huge help and encouragement.
One of the primary issues with the existing system was in terms of the integrity of the homes… how are you looking at addressing that problem?
There are two to three basic things. One is the whole solution of technology. There’s no way the people who know homes will get to see any data. Similarly, the people who get to see the data will never get to see the homes. There are completely different vendors, processes in that. Even with the agency that is going and actually visiting the homes, we’re ensuring it doesn’t flow top-down. That means nobody in the agency at the central level gets the names and addresses of 22,000 people where the boxes will be placed. It goes to the local representative who controls not more than 100 odd homes, and that’s an approximate figure. So, even a corrupted person can’t do substantial damage. The third big thing that we’re doing is putting dummy homes. There will be normal seeming dummy motors in a lot of homes that actually will not ping. That data will not even be used.
What percentage of it will be dummy homes?
I won’t give that figure. Of course, it would ensure that people don’t understand where it is.
You mentioned people who have access to the homes don’t have data but they have access to the previous week’s data.
How would they know where it’s come from?
The fact of the matter is that if an ‘X’ channel has not fared very well, they at least have that data and can influence people… bribe them, et al.
It doesn’t work like that. For example, would Indore be a unit of measurement? No. He’ll never be able to figure out if it has come from there. It would typically not be. It’ll be small enough for people who won’t be able to correlate it to larger data.
But 100 can be a significant number for a long tail channel?
Do you think all 100 homes will actually watch it? It would never work that way. The other big thing would be outlier. The moment you give a gift or something it will come out in my outlier. It’s very easy to do. Just take the outlier completely out from the samples. I have extra homes I’m sampling in any case. If I find some abnormal things happening in certain sections, my outlier analysis will pick up those homes and take them off. This basically means I can replace them with other homes which already have my meters.
In terms of data analysis, you have a mix of companies. Not just Hansa, but a few others too, right?
Hansa is just putting boxes in people’s homes whereas Magic9 Media headed by Praveen Tripathi is doing the analysis.
Are you looking to do any consulting work on how people should review the data?
We think it can be a conflict of interest.
For the industry, do you think there are some salient features you’d like to highlight so as to repose the confidence level that good times are here again…
The industry should look forward to a 180-degree leap in terms of technology, in measurement which by itself is a big one. It is already seeing a leap by three times the sample size than what it is used to right now. Industry should also be looking forward to the transparency we’re bringing along in terms of what we’re doing and how we’re doing. We started with roadshows in the beginning of what we’re trying to do. You’ll see us taking the road again to tell people what kind of sample design we’re doing. Industry should also look at the way we’re overall presenting the whole data which you’ll also see when the final output comes out. The GUI we use, I’m sure they would love it a lot more than what they’re seeing now.
There’s a fair amount to be done in terms of software the people have to install, the ones they have to read, the staffing at the organizations needs to be understood, a huge amount of education that needs to be imparted…
In fact I’m already talking to a couple of media schools. They want to introduce courses for using BARC’s GUI so that when these guys graduate they can be easily picked up from apart from the fact that we’ll also train the users in the customer’s premises.
Is there anything that you’re worried about? Something that’s giving Partho Dasgupta sleepless nights?
Yes, lots. As I told you this whole handshake of technology and different businesses, there are 12 processes and about 26 vendors. This integration is something to worry about. It’s not easy and we’re doing it, but we really have to get it right. I’m not saying it’s happening without any effort. It gives me a huge number sleepless nights.
What’s tougher to handle – technology or people?
Ultimately, it’s all about people. Technology does what people want it to do.
