Cognitive Computing: Planning for Success
In a recent post, we looked at the reasons why so many cognitive computing initiatives fail. And that leads to the next obvious question, “So, how do you avoid failure and plan for success?”
At Attivio, we think of cognitive computing as a set of building blocks with AI capabilities such as machine learning, NLP, text analytics, and so on. Cognitive search uses many of the same building blocks, which makes it a good place to start any cognitive computing project.
Do Your Homework: Establish KPIs
In the case of cognitive search, a good question to ask is, “How can I use cognitive search to raise my enterprise IQ?” Then you can pinpoint important business processes and digital customer experiences that cognitive search could enhance. It could be customer service. Or fraud detection. Or predictive maintenance.
Walk through every step of the process to see whether there’s organizational knowledge that could improve it or lead to better, faster decisions. What are the KPIs you want to use to measure performance? Common examples would be:
- Knowledge worker productivity
- Reduction of false positives
- Faster mean time to repair (MTTR)
If you don’t have KPIs, you can’t set clear goals and gauge ROI. And, you’re a sitting duck for open ended SOWs and license agreements. Cognitive search solutions should be self-funding.
Know Your Data
When we talk to customers, we often find that their analysts have previously reviewed and marked up a large research corpus or set of emails for true and false positives or other types of categories. So, they often have good training data that we can make use of quickly. But they’re not thinking about the organizational data stores from a broad enough perspective.
Companies often discover unexpected benefits once they implement cognitive search — benefits that extend beyond their original KPIs. For example, in lowering MTTR an organization finds that it also has the knowledge to automatically route work to the best resource based on an improved understanding of materials and experts. Or it can modify staffing because lower level employees can more complex tasks when aided by cognitive search.
In fact, many companies find that in addition to meeting specific KPIs, cognitive search raises overall employee satisfaction. Cognitive search simply makes knowledge work more satisfying and less frustrating. So, if you don’t offer your employees a cognitive experience, where will they go to find it?
Ready to get your cognitive computing initiatives rolling with AI-powered search? Contact us to learn more and get a demo.