Submitted by Denis Patenaude, Information Developer, Business Analytics
We all share a common goal: to keep our customers happy - or at least minimize their displeasure and frustration. Recently, Information Development (ID) has undertaken an initiative to work with Customer Support to see if we could combine our efforts. By working cooperatively, we wanted to demonstrate that the existing product documentation could benefit by incremental improvement, and that the data gathered by Customer Support (CS) had value in the wider realm of IBM Cognos Business Intelligence.
Some concrete examples
Here's how some of our quality goals listed further below were manifested in the documentation to help CS solve customer issues and to help close calls faster.
For example, we updated the Upgrade section of the Cognos BI Installation Guide to align with customer support best practices. By doing this, we reduced the documentation by more than half, but still covered real-life deployments.
Another example: we now include step-by-step instructions for configuring Microsoft Internet Information Services for Cognos content. This has been a significant pain point for our Windows customers.
For troubleshooting, there is now a Troubleshooting and support section in the Information Center, with links to troubleshooting resources that customers can bookmark: the IBM Cognos Resource Center and the Cognos Customer Center.
The value of cooperation
ID set up a series of monthly interlock meetings between Customer Support representatives, knowledge engineers, analysts, and writers. We hoped that teamwork could lead to advantages such as faster closing of calls where CS could point to existing documentation - which saves money. Likewise, CS could bring to light how customers are using the information that we in ID provide with the product.
These meetings helped us to develop strategies for delivering quality information - core documentation, technotes, and proven practices - to the customer when and where the customer needs it. We wanted actionable ideas that we could implement by using measurable processes.
The work of the interlock committee led to the following solutions:
reducing the percentage of PMRs associated with documentation
including high-value content from technotes in the core documentation
improving the installation experience
improving error prevention and recovery
Reducing the percentage of PMRs associated with documentation
How could we reduce the percentage of Problem Management Records (PMRs) associated with documentation (that is, where information cannot be found or is incomplete)? We used the Software Metrics Dashboard (SMD) and the IM Metrics Portal to analyze PMRs for trends and common issues. Writers collaborated with Development and Proven Practices folks to determine a plan of action based on specific problem areas, for example installation and upgrade issues.
We set a target to analyze a minimum of 50 PMRs. Individual PMR analysis may take only minutes, but the resulting action could require hours or days of effort. Some actions required as much procedural work as regular content: meetings with subject matter experts, technical reviews, testing, and so on.
Result - we analyzed 134 PMRs (101 PMRs related to installation and upgrade issues, and 33 PMRs related to other areas), exceeding our original goal -50 PMRs - by 84. We took action on 43 PMRs, exceeding our original goal - 10 PMRs - by 33.
Including high-value content from technotes
Because technotes provide information that is typically discovered post-release or during the final regression testing cycles after the documentation has been finalized and closed for the release, we wanted to capture the expertise and knowledge of post-sales technical support contained in the DCF database. This way, we hoped to improve the quality of the core documentation and possibly divert incoming PMRs by including in the core documentation content that had already been used to address known issues and questions.
Information Development received reports from Customer Support (run against the DCF technote database) that we used to search for likely candidates to incorporate into the core documentation. We filtered these reports using statistics such as the highest number of hits, the highest satisfaction rating, and the technotes related to the highest number of PMRs.
Technotes were examined for accuracy and relevancy. All technote-related content was technically reviewed by subject matter experts and edited by the ID editing team before inclusion in the core docs.
As an added bonus, because the core documentation for BI is translated into 21 languages this work not only meets the needs of our English users, but also the needs of our growth markets.
Result - we evaluated 62 technotes, exceeding our original goal - 50 technotes - by 12. We incorporated 40 technotes into the core documents, exceeding our original goal - 10 technotes - by 30.
Improving the installation experience
Installation has been a longtime issue for our customers. We used multiple methods to reduce several installation and configuration roadblocks.
We examined PMRs where the customer activity was listed as installation or configuration. In particular, we searched for cases we believed we could resolve using an ID-based solution, for instance cases where the information provided was difficult to understand, or cases where the customer didn't access the available documentation.
Result - We analyzed 513 PMRs, (exceeding our original goal -100 PMRs - by 413. We took action on 43 PMRs, exceeding our original goal - 10 PMRs - by 33.
We also wanted to reduce the time it takes to get up and running by improving the 'getting started' experience. The BI Getting Started Installation Guide was updated with steps to install BI on a single machine, based on a popular use case: as a single-server QC, testing, or training environment. We based these changes on feedback from a usability survey targeted at field personnel, Customer Support, and external Beta customers.
Using information from the same survey, we included verification topics for targeted pre-installation, installation, and post-installation steps in the BI Installation and Configuration Guide.
Improving error prevention and recovery
Another aim was to improve customer error prevention and recovery by updating the troubleshooting topics derived from the Troubleshooting Guide Development Kit (TGDK), based on recent feedback from BA Technical Services and Customer Support.
We used internal and external feedback to update the content templates that documentation teams use to deliver troubleshooting topics. These topics are designed to help users determine a course of action when an error is encountered, and help them to solve their own problems, thus reducing the amount of time spent with Customer Support on problem resolution.
A key finding was that the troubleshooting documents lacked Cognos-specific links and information. As a result, a new Troubleshooting resources section was added to the Troubleshooting Guide, with links to the following child topics: Support Portal, Problem determination, Service requests, Cognos Customer Center, Fix Central, and Knowledge bases. Each child topic includes a description of the troubleshooting resource and links to related ibm.com pages.
We refined the new section using the results of a survey submitted to internal representatives from a variety of groups (QC analysts, CS Managers, Service Management, System Quality & Security Competency, Proven Practices Advisors, Accelerated Value Specialists).
We can use our mutual experience to increase customer satisfaction both today and in future releases.