It?s become all too familiar. Take any set of numbers?income, consumption, production, inflation, poverty, sales to television ratings?and you almost always have a big argument on it. All manner of consumer goods marketers, and lately even the head of one of the world?s largest consumer expendable companies, Unilever?s CEO Paul Polman, complain that market researcher Nielsen?s Retail Audit does not capture its market positions accurately. There is this long-standing doubt and dispute over the robustness of television rating numbers by TAM, with niche channels and rural audiences supposedly getting short shrift. Ditto with readership surveys for the print media, which somehow again favour big newspapers and magazines over smaller and specific-interest driven publications. Time and again, industrial production numbers, as represented by the Index of Industrial Production (IIP), have shown to be in divergence with not just company production and sales numbers on the ground, but even with another set of government statistics, the Annual Survey of Industries. And every passing year, the National Sample Survey?s private final consumption expenditure data is capturing less and less on what Indians consume, just 45-50% now compared to a high of 75% in 1970s.

One can go on and on, but you get drift. Numbers are not capturing the reality and trends in a rapidly changing and growing India. Where?s the problem? Well, let?s first take note of the issues raised by data users.

Marketers like Hindustan Unilever, Dabur and Marico point out that the Nielsen sample of 22,000 retailers covers under half-a-percent of over 11 million shops in the country. And then there is always the issue of capturing retail diversity in a country as heterogeneous as ours?from paanwallas, neighbourhood kiranas, rural haats and hole-in-the walls to fast proliferating swanky big-box retailers.

Similarly, TAM?s criticism has been its very small sample size of just 8,000 peoplemeters in a country with over 400 active 24/7 channels, 138 million television homes, with around 30 million connected digitally. Add the complexity of at least half-a-dozen languages big with television audiences, and clearly you have a currency that looks far from robust. IIP suffers from a combination of an outdated base year (1993-94), though it is being updated to 2003-04 shortly, a basket of 543 products that overestimate defunct categories like typewriters and underestimate host of others like cell phones and LCDs, and a general lack of data reporting compliance by even organised sector companies. The NSSO expenditure data and CSO?s National Account Statistics for National Income again trip at the altar of non-representation of the vastly changed dynamic of the Indian population, which is earning and consuming in fundamentally very different ways now compared to even a decade ago.

The key problem across all these data sets is non-representation of the universe, by acts of omission or commission, and the easiest solution bandied about is increasing the sample depth and spread to capture more robust data that may lead to a better quality of statistics for all manner of planning. That?s easier said than done.

Take retail audits or television ratings, for instance. By some estimates, consumer goods marketers and broadcasters in India spend under half-a-percent of sales on market research, and that includes all company-specific qualitative and quantitative research, besides subscribing to industry-wide syndicated research like the one Nielsen Retail Audit and TAM?s peoplemeters throw up periodically. A predominant view with market researchers is that penny-pinching marketers have driven the industry against a wall, and that sub-standard data quality is its natural corollary coming to haunt them now.

Assuming Nielsen and TAM were to double their samples in the quest for a better representation and take cognisance of genuine complaints by marketers and broadcasters who use its data, the industry too will need to rise up above its self-defeating emphasis on keeping investment in research untenably low, something that has hurt marketers more than research companies in the long run. Templatising large parts of consumer research on global lines by transnational marketers and researchers too is contributing to the current impasse where marketers are getting caught on the wrong foot in sussing out new consumer trends. Fundamental consumer research, so vital for a complex society and culture like ours, also needs to come back big-time on researchers? agenda and the marketers? budget.

It?s a welcome sign that the government has overhauled a host of indices in the recent past?from revising the base year for inflation numbers, a new survey for estimating poverty is about to be kicked off and IIP will get a new and more current base year soon. But tardy implementation of the Collection of Statistical Act 2008, which was passed by Parliament last year, means the industry still openly flouts data reporting norms, as evident in far-from-compete IIP numbers every month. But luckily, think-tanks like National Council for Applied Economic Research?s Centre for Macro Consumer Research have kicked off a massive survey exercise, covering over 5 lakh households across the country, to get primary-level data on income, consumption, ownership, purchasing power, intra-country money transfers, etc, and encouragingly will update this data every year hence on. That, hopefully, will fill some of the void that the government agency?s income and expenditure data is failing to capture, and should be leveraged by policy makers and marketers to better understand the emerging new India.

shailesh.dobhal@expressindia.com