We are a motley group of people with a vibrant mix of youth and experience amongst our ranks, possessing a wide range of skill-sets and training. We are researchers, designers, statisticians, economists, ethnographers hailing from diverse backgrounds. We draw on each other’s talents and capabilities to conceptualize solutions to complex problems and deliver them in the simplest, most impactful manner.

Our expertise lies in the field of forward-looking consumer understanding, be it through segmentation, pricing, forecasting, branding, communication, or product testing. We regularly work with the best minds in academia and technology globally to ensure that our solutions are founded on proven principles and are applied creatively to anticipate the future. We apply cutting edge and latest analytical thoughts, processes and models – from picture recognition to social media analytics to neural networks and agent based modelling – to System-1-inspired methods of consumer engagement. We are constantly validating, challenging, and recasting our own work, in order to keep up with the demanding and ever-changing nature of the times we live in.

Our solutions are flexible enough to adapt themselves to various sectors like FMCG, auto, durables, financial services, telecom etc. They have been used to frame and execute successful strategies across companies and sectors, both in India and globally.

New products are hard to forecast- especially in developing countries, where normative databases may be sparse – and the level of over-claim for questions such as purchase intent is so high that it’s hard to extract any ‘meaning’ from responses.

Visualize is Kantar IMRB’s forecasting suite. It is created specifically to address the need for sound, reliable forecasting in the developing world - where politeness and overclaim can have an overwhelming influence on any rating scale such as purchase intent or preference. The problem is so big that sometimes it’s hard to say whether a ‘good’ score is actually good news.

Visualize is predicated on three learnings:

• Easy to intend, uneasy to buy: People, especially those in developing economies, tend to say “I will definitely buy”, resulting in very high top boxing and little variation across concepts. Intuition often became the guiding principle for decoding concept performance
• Competition is crucial: Good estimates of trials are best obtained when brands are presented to the consumer in an unbiased setting that enables comparison.
• Deconstruction of new launch trial and repeat behaviour: Harnessing the richness and depth of Kantar World Panel data, the trial and repeat-buying framework was analysed as well as theoretical models before building the Visualize framework

It can be used for:

• New products
• Line extensions
• Concept only

How Visualize works

It uses our proprietary ‘variable shelf” comparative methodology which foregoes the need for normative databases and presents the product in a changing, comparative context without drawing undue attention to it, thus yielding realistic estimates. The idea itself is simple; the new launch is placed on a virtual shelf with other competing products.

It uses our proprietary ‘variable shelf” comparative methodology which foregoes the need for normative databases and presents the product in a changing, comparative context without drawing undue attention to it, thus yielding realistic estimates. The idea itself is simple; the new launch is placed on a virtual shelf with other competing products.

It uses our proprietary ‘variable shelf” comparative methodology which foregoes the need for normative databases and presents the product in a changing, comparative context without drawing undue attention to it, thus yielding realistic estimates. The idea itself is simple; the new launch is placed on a virtual shelf with other competing products.

In order to achieve greater alignment with market reality, awareness, word-of-mouth spread, and impact of planned distribution is built in.

The volumes are predicted through a judicious mix of choice modelling and NBD (Negative Binomial Distribution) concepts implemented in a customized R based software to provide volume estimates for first two years, maximum achievable trial and repeat. We also estimate the likely source of business and provide diagnostics on concept and product.

Deliverables

• Sales volumes and revenue forecast
• Diagnostics for improvement, using techniques like Penalty Analysis and Bayesian Belief Network