To minimise the risk of disaster, it is vital to obtain real-life reactions from potential customers both to new product/service concepts (initially) and to polish the product in the pre-launch phase.

Qualitative surveys (especially focus groups) are often used in the initial stages to gauge broad reactions to the concept. A large amount of skill is needed to elicit meaningful responses, especially when the concept is totally new or revolutionary (eg e-commerce, and brand extension into fundamentally new markets; both are being championed by Virgin). Specific qualitative predictive techniques exist.

Later on, perhaps when the most strongly demanded product/service features have been identified and/or incorporated, quantitative research has a part to play both to make a choice about which of several attractive options to develop and in pre-launch fine-tuning.

“Hall tests” are often used for consumer and mass-market NPD surveys, especially where food and drink tasting, or where expensive or heavy mock-ups are involved. A large interview room is normally booked in a hotel or near the high street (often above a pub) and a team of interviewers recruit the necessary types of people off the street. Interviews take place in the “hall”.

In addition to assessing the most desirable price and pack sizes (see Pricing Research on page 13), marketeers are often required to assess the optimum combination of features. In this case, a BPTO-derivative called “Conjoint Analysis” is used which asks respondents to make a series of choices between different products/brands with different features at different prices.

Respondents are shown a series of pairs from which to choose (eg Brand A, 125ml pack, at £2.99 vs. Brand B, 250 ml at £4.99). As each preference is recorded, conjoint software presents only those new combinations that are needed to build up a statistically-robust model to establish the importance of each brand/feature and whether it makes the respondent more or less price sensitive.

Once this model has been built, “What if” scenarios can be run to assess the optimum combinations and likely sales volume for a specific brand at a given price. The model can also assess the optimum price for a brand with a pre-determined set of features. In this way, the marketeer can design several different product variations and assess the optimum price (and likely sales volume) for each.