Impact of AI on KM Strategy: A Two-Way Street
In one example of how PolyAnalyst was used, a grocery retailer interested in understanding customer reactions to its blended juices selected “customer satisfaction” as the KPI. The initial output from Megaputer’s PolyAnalyst software was a multilevel hierarchy of categories discovered in feedback, with a graphic visualization of how each category affected the KPI. One of the key causes of dissatisfaction was identified as “labeling and claims.” Customers who were concerned about this topic rated the product with 2.28 stars on average, compared with an overall average of 4.6 stars.
“Some customers found the labeling to be deceptive or misleading,” Ananyan detailed, “because the product label did not indicate all the fruits in the blend.” Negative business impacts included product returns and lack of repeat purchases. The retailer also had concerns about liability issues in the event that customers did not have a full understanding of the ingredients.
The findings were then passed to a report-generating agent, which had access to the customer feedback and could seek outadditional relevant information on the internet. “In the AI-generated report,” continued Ananyan, “specific recommendations were made about label redesign, more visible disclosure of ingredient percentages, possibilities for product line modification, and customer education strategies.” The result enabled the retailer to improve customer satisfaction and find new business opportunities.
The system was able to make sense of 20,000 customer comments and produce a report with actionable items within a matter of hours rather than days or weeks. Compared to previous generations of customer sentiment analysis tools, Megaputer’s process allowed for flexible categories rather than predefined ones, a more nuanced interpretation of the data, and the creation of an informative GenAI report.
Solutions like AI Insight Extractor encourage companies to rethink their KM strategies by turning customer feedback into a new strategic knowledge resource. They help identify key drivers of the selected KPIs, perform in-depth analysis of the discovered gaps, provide reports that trace possible root causes, and offer recommendations for improvement. The reports deliver additional knowledge that did not exist before and can become part of the knowledgebase.
Activated Knowledge Linked to Business Outcomes
AI has revealed the struggle that organizations have in truly activating their knowledge assets, according to Frank Palermo, COO of NewRocket LLC. “With data assets so disparate in legacy systems, a lot needs to happen for real transformation,” he said. “Knowledge needs to be activated and executed, not just retrieved.” This cannot happen without an enterprise-wide system for accessing and leveraging corporate knowledge. NewRocket is a consultancy that works primarily with ServiceNow, a cloud-based platform for automating and managing digital workflows.
NewRocket provides solutions across a broad spectrum of industries, including manufacturing, retail, financial services, and healthcare. Its new FlightPath.AI identifies and applies GenAI and agentic AI to business processes within the ServiceNow environment.
A consistent message from both vendors and practitioners is that data quality is a strong determinant of AI success. “Good data is a prerequisite for AI,” emphasized Palermo. “We do a lot of work around data assets. A big piece of this is to make metadata consistent within the KM space.” ServiceNow brings data into a single system in a hub-and-spoke configuration, and reconciling differing sets of metadata is an important step in this integration.