Generative AI Accelerates Product Innovation

The first step to product innovation is knowing your customer requirements. For consumer products this means interviewing people and conducting consumer surveys. Designing a well-structured question set with probing questions to gather focused consumer insights was the first step. The second step was to find willing participants and interview them, recording what they said, generating a transcript. The third step is reviewing the transcripts, pulling together the insights that will end up being the key consumer insights.

If the innovation is occurring with an existing product, additional data should be gathered from online website scraping and quality database review. These additional insights highlight both deficiencies and highlights versus competitive offerings. This allows the innovation team to avoid over engineering or eliminating key features of a product that consumers love, while eliminating headaches or deficiencies. 

My experience is that the consumer insights steps are many times skipped by companies because it is difficult, time consuming, and expensive. The insights instead come from sales, marketing, and technical people that have a pulse on the consumer needs, but may be biased. With the help of Retrieval Augmented Generative (RAG) AI, both the internal expert insights and the consumer insights can be quickly distilled down into meaningful summaries. Gen AI can even create the key functions of the product based on analysis of the transcripts.

Once the consumer insights have been gathered the product innovation team can begin to brainstorm and develop concepts that resolve current issues and delight customers with new features. Again, Gen AI can help organize ideas, ranking each idea as they pertain to the consumer insights.

Gen AI does not replace the human factor but it does speed up the analyze and can remove bias. RAG is a tool that humans use to quickly distill down large amounts of information. Activities that would have taken days or weeks can happen in seconds. AI is not perfect, however instead of skipping or reducing down the consumer interviews, RAG can help get at least 80% of the way there in minutes versus days of listening to audio and then distilling down the information.