Take The Customers With You - AI product innovation

Jul 4, 2024

3 min read

In a recent collaboration with Kathleen Rouse, I contributed some thought around - why the pace of technology innovation and the advanced nature of the technology requires a significant lead time to ensure customer onboarding matches the expected timelines of product release. For more on this please check the blog "The Executive Guide to Cutting-Edge Tech Customer Onboarding Strategies".

In this blog I want to go a bit deeper into the progression of "innovation-led" projects. Specifically, I want to focus on innovations that have some component of transformation involving either a new set of AI-assisted features, new data interpretations, or new strategic decisioning frameworks that contribute to the value proposition of the innovation.

AI wears many hats. Some view it as a technology innovation; while others see it as an evolving catalyst of change. AI is about data. As analytical concepts evolved and merged with mathematical, algorithms, machine and computing technology, the characteristics and profile of AI is seen through many lens. This creates considerable chaos in organizations who have been very traditional in their approaches to bringing products into the market. The need to recognize not just the new AI roles but also the new AI engagement models is closely linked to the successful outcomes of business.

AI augmented assets can be a golden road if a business does their due diligence – not just about customer onboarding but also in establishing a customer willingness to adopt the new product.

Take the example of a data-driven model score that is used by customers in making additional decisions as a part of their business process.

1. While they continue to receive a score the innovations in data and AI techniques may change.

2. Certain new data or features in the model may have been changed.

If these product innovations are not communicated to the customer ahead of time it can create distrust in the product or a branding of “inaccurate outcomes”. A software update is a structured and rigorously tested against the functionality that they deliver in a process pipeline. Algorithmic changes or data used in training the algorithms are also tested for improved accuracy or improved insights.

These will still deliver a model score. What has changed is the very meaning of what the score represents. The customer needs to understand and redesign their decisioning strategies to accommodate around these changes.

When Credit reporting agencies, enriched and augmented their models with new data and new AI techniques, an individual who had received a score of 690 now is seen to receive a score of 650. As a renter, how am I to know if I should continue to use the same thresholds for decisioning? What should I do if my rent application systems are coded to reject all applications below 680?

Customer (renter dilemma):

· Should they invest in updating the application acceptance process and do manual review till updates happen (There is a system update cost attached to this option).

· How should the lower scores be interpreted for current contracts and future renters? (Current system thresholds will automatically decline future renters, reducing the prospect pipeline- vacancies may go up)

Customer education and socialization of the forthcoming changings ahead of time would at least have allowed the renter to be better prepared. Explaining the value of the new information in the context of the applicants might help the business be assured that 650 is the old 680 or something else! Keeping that trust in your product is a critical role that customer onboarding, done properly, can achieve.

In summary, planning new product innovations with new AI technologies- Take the customers with you so the final product is relevant, accepted, and valued by your customers.