New Product Development Research (NPD Research)

NielsenIQ (NIQ) reviewed 250,000 consumer packaged goods (CPG) innovations across 130 categories and found that "just 50% of innovation launches show growth by the second year" (NIQ, 19 March 2025). For CPG, half of the launches reviewed were not growing by year two.

A product team in India has to decide at each stage whether to go on, and the stage names alone do not say what to test, how many people to ask or which result means stop.

A concept test, for example, needs a number agreed before fieldwork starts, such as the share of participants who choose the concept over the product they use today, below which the concept is killed. Interviews can settle some questions and others need different data, and mixing the two is where most new product research goes wrong.

In the Stage-Gate model, each stage ends at a gate, a management decision with four outcomes: Go, Kill, Hold or Recycle (Cooper, Product Development and Management Association (PDMA), 9 September 2026). Research supplies what that decision rests on.

The same source names the leading cause of failure as a "lack of understanding of customers and users", meaning their problems, needs and purchasing criteria.

Match the method to the decision at the gate, and write the stop rule before fieldwork starts. The stop rules in the sheet are recommendations, not industry standards.

Interview and group sizes come from a systematic review of saturation tests, which bears on sample size in qualitative work. Hennink & Kaiser (2022, Social Science & Medicine) found data saturation, the point where new interviews stop producing new themes, typically at 9 to 17 interviews or 4 to 8 focus groups for similar groups with narrow aims. That measures when themes stop appearing, not whether a product decision is safe. Diary, survey and in-home-use-test sizes are recommendations. The survey size applies to concept testing, the usability checks at the prototype gate are prototype user testing, and the validation sizes apply to an in-home use test.

A new product has failed when it misses a measure set in advance, so "failure" needs a measure before any figure means anything. NIQ's measure is growth by year two for CPG innovations. The finding applies to CPG; do not assume it for software, services or durables.

For a CPG launch, use year-two growth as a benchmark. For any product, write your own measure into the research brief: what to track, over what period and against what plan. Missing a sales target, being delisted and never reaching the shelf are different outcomes, and each calls for a different earlier test.

Stated intent predicts sales only imperfectly. A review of 60 years of research concludes that purchase intentions "are correlated and predict future sales, but do so imperfectly" (Morwitz, 2014).

Interviews and groups are strong on the why: whether people understand the idea, what puts them off, and what they do today. They are weak on how many will buy. In discovery interviews, ask about the last time the person met the problem and what it cost in rupees and hours. That is behaviour, not opinion. See customer discovery interview questions.

Recruit the number who must finish divided by the share who will finish. This is a planning model, with 85% and 65% as example completion rates. For 60 usable households at 85% completion, 60 ÷ 0.85 = 70.6, so recruit 71. At 65% completion, 60 ÷ 0.65 = 92.3, so recruit 93.

Write the pass mark into the brief before recruiting. For example, set the share of households who say they would reuse the product. The result then cannot be argued after the fact.

NPD research is not a separate method. It is market research, usability testing and product testing organised around the gates of one product. Concept testing is one gate, "a market research method in which customers are presented with a description of a product or service" (OpenStax, Principles of Marketing).

We recruit and run qualitative studies for the discovery, concept-reasons, prototype and home-use stages. Our formats are in-depth interviews, focus groups, digital diary studies, shop-alongs, ethnography, usability and app testing, and product testing (IHUTs).

We recruit by city, check participant identity before invitation, target by household and demographic profile, and run sessions in languages other than English. Moderators can be ours or yours. Deliverables such as recordings, transcripts, data exports and reports are agreed in the scope.

We have no fixed rate card; each project is quoted after we read your brief, and we offer a free pilot. Start from our guide to writing a market research brief, then share your brief with us (support@qualfacto.io) at support@qualfacto.io. Use WhatsApp for quick queries. We reply Monday to Friday, 10:00 to 18:00 Indian Standard Time (IST), and aim to answer within one business day.