New research models needed in China

10 February 2012

BEIJING: Brand owners must overcome a variety of obstacles to generate accurate and useable consumer and market insights in China, executives from adidas and Procter & Gamble have argued.

Avni Patel Thompson, senior manager, strategy, at adidas China, and Joseph Thompson, Greater China marketing leader for Gillette, Procter & Gamble's shaving brand, said in the Harvard Business Review that "bad data" was a major issue in China.

One reason for this is that organised chains such as Walmart and Carrefour currently take just 20% of local retail sales, with thousands of hard-to-monitor smaller outlets attracting the remainder.

"To add to that, many of these retailers resist systems that limit their ability to participate in the 'informal economy,'" the study said. This makes it much more difficult to gather figures as Nielsen and Symphony IRI do in countries like the US.

As an example, the article reported that research commissioned by adidas into its market share yielded findings showing the sportswear sector doubled the size of in-house estimates.

"Therefore our business ... was twice as large as we knew it to be based on what we shipped," the paper said. "We use a reputable firm, and its numbers were clearly capturing something we were missing."

In response to the fact this type of information is "hard to come by", brand owners will have to commission large-scale research themselves, covering the entire industry and country, it added.

More specifically, it is initially vital to ask "ruthlessly simple questions", prioritising uncomplicated metrics and absolute numbers, and thus delivering tangible results that can be checked against "intuition".

Procter & Gamble used this model to determine how many men would buy a new product in its first year, looking at establishing the size of the target audience and ideal amount of retailers to reach.

Counterfeits are a further problem for firms like Nike and adidas, and some shoppers, especially in lower-tier cities, frequently think they have bought the real thing, leading to distorted research findings.

Detailed demographics must also be considered, as there are seven tiers of city in China, and even the smallest details within a sample of 10,000 people can reveal meaningful trends when extrapolated outwards to a population topping 1.3bn people.

"Culturally, the Chinese are not comfortable with or used to intensely scrutinising or debating what's presented in these cases," the paper continued, meaning coaching is often important.

Data sourced from Harvard Business Review; additional content by Warc staff