Generative AI Roadmapping
As companies look to prioritize AI initiatives, a framework is needed to asses which initiatives to pursue first. Three dimensions to consider evaluating AI capabilities based on include:
AI Risk
When thinking about risk, different type of AI initiatives produce more or less critical results depending on the application and potential use.
?
Lower Risk initiatives might be surfacing the right knowledge base document for a particular service ticket. You’re facilitating a more efficient process, but the accuracy of the result doesn’t have financial, regulatory or compliance implications.
Other low risk initiatives include content production - whether creative or research. These are low risk as long as content produced is only used as a more efficient input into other processes, and not an end-product into itself.
Higher Risk initiatives might include using AI as a virtual data aggregator for reporting to replace a physical data warehouse and reporting platform. Pulling the right data together accurately based on the query is vital to the end-user if they are going to make business decisions based on that data. The right data from the right sources for the right time frame needs to be pulled together in the right way to accurately provide a response to the user’s inquiry.
AI Complexity
AI Complexity represents the degree of AI training to get a quality response. Simpler capabilities require less training and fine-tuning that others, if any.? A simpler example would be our knowledge base example; finding the right help document for a service ticket is a binary result; finding the right ticket or not. This requires very little algorithm training versus more of an issues mapping.
A higher complexity capability would be predicting churn risk of existing customers based on their renewal history, NPS score, usage patterns, etc. This requires the fine-tuning and training of a predictive algorithm over time, and assessing the quality of the results.
Customer Impact
Then there is the breadth of the customer impact of the AI capability. How many customers would it impact, and what value would it provide for them in making them more efficient, more productive or helping them make better decisions. Does the capability solve a use case than only a few customer have, or a capability to help a wider breadth of customers.
Ideally, you want to prioritize lower risk, less complex, high impact capabilities over higher risk, more complex, lower impact capabilities. Using this framework can help you prioritize which AI initiatives to pursue first.
Competitor AI Initiatives
The other dimension for evaluating AI initiatives is looking at what your competitors are doing in the AI space. You want to stay competitive, so assessing where competitors are focused in another way to help prioritize which AI initiatives to pursue.
As an example, in the subscription management and billing space, after analyzing a number of vendors, the top three areas where AI solutions are being developed include:
This gives others in the same industry an idea of where competitors are find low-hanging initial value to focus on with AI.
#SaaS_AI #AI_Framework #AI_Prioritization #Zuora #Maxio #Recurly #Stripe