The Productivity Pane it lacks essential contextual awareness, making key features ineffective.
In our scenario, knowledge articles are directly related to accounts via a custom relationship, with each account maintaining 50–100 articles covering unique systems, processes, and configurations. On the case form based knowledge search control, this works flawlessly, the account filter is applied automatically, and the initial search is driven by the case subject.
Yet, in the Productivity Pane’s knowledge search, while the initial search can be automated, there is no filtering or contextual awareness whatsoever. As a result, agents are flooded with irrelevant knowledge articles that have nothing to do with the case/account they are working on.
For example, in testing:
- Form based knowledge search returned 2 relevant results (based on case account as a filter and case subject as the search criteria)
- Productivity Pane knowledge search returned 98 results, most of which were completely unrelated due to the lack of filtering on account.
Despite this, the Productivity Pane knowledge search is now the recommended approach for multi-session apps. Yet:
- Knowledge search is unusable, forcing agents to manually filter through irrelevant results, slowing workflows and increasing the risk of misinformation.
- Copilot is surfacing incorrect information because it cannot distinguish between accounts. This is particularly damaging when similar keywords exist across different accounts but require account-specific solutions.
A logical fix would be to enable contextual filtering, whether at an individual feature level (e.g., KB search or Copilot) or across the entire Productivity Pane, compatible with both custom and OOTB relationships. Currently, there is no automatic filtering at all. The system already knows the case and account being worked on, so why isn’t this context being used to drive intelligent filtering? This should be a given, yet it remains an oversight that cripples productivity.
Other customers also rely on similar relationships to streamline workflows, yet the Productivity Pane completely ignores this need. Unless running a barebones, vanilla Dynamics instance, this design choice makes no sense. The system has the data but can not use it. Is leveraging relationships for filtering really that complex - or is the feature simply built for the most basic, superficial use cases?
We have completely disabled the Productivity Pane now as it delivers nothing but noise.
