Ironic isn’t it that technical organizations are sometimes the ones that have the greatest mental blocks against replacing manual processes with automation. And in this case, we actually mean “manual” processes – namely the diagnostics procedures from the user manuals!
We recently spoke to a prospective customer who totally got the benefit of capturing and automating diagnostic workflow as an alternative to the status quo of giving users static PDF-based troubleshooting guides. She briefed her Field Service group on how she could replace their static PDFs and use LogicNets to easily capture and deliver their diagnostics as an interactive process. She very accurately described to the group the advantages of such a system including:
- Guided navigation that would be good enough to give directly to end-users, while flexible enough to support veteran reps who would undoubtedly want to choose from a range of alternative pathways and possibly skip ahead to different steps or end-points.
- Quantitative encoding and management of diagnostics elements, allowing diagnostic processes to be characterized, tracked and managed based on the relationships between symptoms, solution end-points, and resources. This would transform the knowledge-base from a collection of prose (some of it pretty heavy reading), whose usefulness is entirely dependent on Search, into an automated network much easier to navigate, capture, and maintain.
- Heuristic learning where the system itself would keep track of all diagnostics usage to dynamically predict the most probable solutions and marginalize irrelevant pathways over time – i.e. solve the knowledge-base maintenance challenge.
She thought she had everyone on board and then a product manager spoiled the moment. He observed that in customer satisfaction surveys, the company’s diagnostics manuals were “very highly rated” by the users. “That settles it,” declared a manager, “no need to fix what ain’t broke!”
Thankfully, we were able to get a discussion going with the two nay-sayers and we were able to turn them around pretty quickly and effectively. Our argument was simple: “how do you know them user manuals ain’t broke?” We asked how many times their diagnostics manuals had been used in the last year by users and reps. Did they even know if current manuals were being used? Did they have a breakdown by product or diagnostic area? Were they able to identify in complex scenarios which pathways were most successful and did the field have any feedback informing them of evolving improvements? As they tried to answer these questions, they realized that, aside from a SurveyMonkey exercise, they actually knew nothing about how their PDF manuals were used by customers and that with static prose they had no means of closing the interaction loop with a better understanding of product performance and the actual customer experience. We then described how several of our users have intentionally directed reps and customers to use LogicNets Interactive Troubleshooting guides and our Knowledge Center module in place of their traditional manuals. The results have been noteworthy. Not only have users found it easier and more convenient to get to their answers, but every single step they work through is recorded, providing an accurate paper trail and giving reps back the time they used to waste recounting their cases in the CRM system. LogicNets has provided clear and quantitative information about the effort and path taken to solve problems in hundreds and thousands of incidents. Furthermore, with the Predictive Diagnostics module it automatically adjusts to real-world usage, constantly refining the recommend pathways and resources needed to optimally arrive at a solution. This information is also available to Engineering systems providing critical feedback about actual product performance so that proactive maintenance programs can be implemented.
So, we guess the moral is: “if it ain’t broke, don’t mean it ain’t broke!” We actually think that the combination of prose and PDF is the big red flag. Surely if you make your living on technical products, you would understand the limitations of unstructured text and the de facto benefits of automating… especially if it is affordable.