Forrester Groundswell Submission: Cisco Technical Assistance Center
Cisco's Technical Assistance Center (TAC) uses social to reinvent Customer Support for high-tech products. The initiative transformed TAC's traditional escalation model and created a networked learning organization by creating knowledge through collaboration.
- Intelligent Matching: Routes service requests to the right engineer first
- Collaborative Engagement Model: Encourages experts to collaborate on complex issues within the CRM system and Tech Zone a Lithium community
- Integrated Workflow: Aggregates multiple tools into a single interface
- Knowledge Creation & Reuse: Captures crowd-sourced knowledge directly within the workflow
- Reputation and Gamification: Encourages quality and participation with Lithium’s social recognition tools
Adoption has gone viral and the results have been impressive:
- Participation went from 300 to 8,500 in nine months
- Generated $7 million in savings
- Reduced time-to-resolution by 32%
- Reduced case escalations by 42%
About Cisco Technical Assistance Center (TAC)
The Cisco Technical Assistance Center is staffed by a team of 3,000 engineers. This team spans 20 different TAC facilities, supporting over 180 countries and customers speaking over 17 different languages. The Technical Assistance Center covers about 7,000 products and features, and handles over one million service requests yearly. TAC support engineers are highly educated and technical product gurus. The group is dynamic, as is each individual's skill set, which makes it difficult to map engineers to customer issues using traditional mechanisms like a call-routing rules engine.
The TAC had an industry leading C-SAT rating of 4.67. This is an organization that was at the top of its game, but was motivated to change because they saw the world changing. Customers and employees were becoming more social and collaborative. Technical problems were becoming more complex with technical dependencies between product lines and partner companies.
Description of the problem
High-tech products like Cisco's are increasingly complex with problems that touch multiple technologies at once, making it difficult to resolve customer support cases. There is no good mechanism for handling the customer support calls that can't be resolved by a Level 1 support engineer. Traditional protocol for tier escalation and emailing requests for help don't work well. Escalating to another agent burdens customers, requiring them to restate their problem and the attempted fixes thus far. Emailed requests for help have a low success rate. If and when they do succeed, the resulting knowledge ends up siloed in Outlook or an engineer's desktop.
Last year, Cisco followed a traditional escalation model, Level 1 = 80%, Level 2 = 15% and Level 3 = 5%. The management team within Cisco TAC believes the escalation based structure within a support organization is fundamentally flawed for these three reasons:
1. Everyone loses.
- Client gets bounced around while company tries to find the right way to solve the problem. When they finally get it solved at Level 3, feels it took too long.
- Level 1 person feels dumb that he couldn't solve the problem
- Level 3 person is frustrated because even though he solved the problem, he will get a bad c/sat rating from the frustrated client.
2. There is no learning or knowledge transfer. Level 1 didn't learn how to resolve the customer's issue and didn't expand his expertise. It takes a Cisco engineer 6 years to become an expert, and escalating to Level 2 or Level 3 doesn't help engineers learn faster.
3. There is no knowledge capture for re-use by agents or in a self-service portal.
- The next time this customer issue arises, the answer isn't in the KB.
- The escalation isn't likely to route to the appropriate engineer by chance.
What we built
Cisco TAC took advantage of the social networking revolution to move their workforce toward global collaboration. The transformation effort included a five-point initiative to transition from a traditional escalation model to a networked learning organization.
- Intelligent Matching: Routes service requests to the right engineer based on skills and requirements matching
- Collaborative Engagement Model: Encourages experts to collaborate on complex issues within Tech Zone, a Lithium community
- Integrated Workflow: Aggregates multiple tools into a single interface integrated with the case management system
- Knowledge Creation and Re-use: Creates crowd-sourced knowledge articles directly within the resolution workflow
- Reputation and Gamification: Encourages quality and participation with Lithium’s social recognition tools
How it works
First Engineer, Right Engineer
In the new service delivery model, Intelligent Matching of customer to engineer precedes collaborative engagement. Only when the engineer has been matched, can it be determined whether or not collaboration is necessary to the resolution of the case. Using our Business Rules Engine (BRE), we will improve our ability to accurately match customer needs to the most appropriately skilled engineer at first contact – “First Engineer, Right Engineer”. At the same time we need to allow engineers to “swarm” to a customer problem by rapidly bringing together correctly skilled and available individuals when the current owner of the case needs them. In this way, customer problems are resolved by built on-demand solution teams.
To minimize engagements being moved from one engineer to another, our Intelligent Matching Business Rules Engine (BRE) will evaluate the initial engagement and immediately match it to the engineer with the right qualifications, based on what is initially known about the customer problem.
When an agent doesn't know the answer to a customer issue, the first step is to search Knowledge Base (KB) sources including the Tech Zone crowd-sourced knowledge base. The system does a search on the question and displays potential answers including quality indicators like kudos and the number of service requests linked to article. If an article meets their needs, they link it to the case. If they don't find a knowledge resource, they can reach out to others for help. The model provides three channels through which to collaborate.
- Urgent Collaboration: Specific expertise needed immediately.
It is guaranteed you will get someone in a very short timeframe. That person is dedicated to working with you directly to help solve the problem in real time through in-person or phone meetings, chat, email, and/or WebEx.
- Non Urgent Collaboration: Specific expertise not needed immediately.
The request goes out to people who are skilled in the particular area. They pick up the request but may not respond immediately.
- Post a Question to Tech Zone: Guidance requested from Cisco community of experts.
The audience can be anyone in the company willing to help give guidance on a particular question.
Tech Zone is a collaboration channel that allows Cisco engineers to experience real-time collaboration with other Cisco engineers around the world. Using this powerful environment for knowledge sharing and content creation, capture and reuse, engineers can post questions, transfer conversations into other documents for reuse and tap into “crowd sourcing” to get help from experts in other areas.
Participating agents can post a question to Tech Zone directly from the case management system. The CRM case file has already categorized the product and problem type, which is used to route the request to the correct community board. The system auto-populates the case ID into the subject, and includes relevant technical data. People who have the skills to help answer will see the new issue as a work item in their case management tools. All Tech Zone contributions are added chronologically in the case notes. Once an issue gets resolved, anyone can mark the resolution as an "Accepted Solution." This is important because people want to feel empowered and trusted. To prevent gaming or inaccurate markings, anyone can also remove the Accepted Solution indicator. These open permissions enable the community to monitor and govern itself.
Tech Zone enables engineers to build a reputation on the content they contribute and the collaboration they participate in. This reputation will contribute to the larger reputation that accounts for all of their work.
Participation is 100% voluntary and adoption of other new initiatives has historically been low. By contrast, Tech Zone was built by engineers for engineers, and participation within the Tech Zone community has been fantastic. Participation is motivated by the innate gratification of helping according to your interests and skills, plus some awesome gamification and social recognition on public leaderboards and in management reports.
Supporting quotes from TAC engineers:
“Having Marcin as an Expert here is really a blessing, doing labs with him made me learn new things - not just fix a one-time problem.”
“This [service request] was an example of our new tools working in the customer’s best interest. Kudos to Raymond for following the new process and pulling in the appropriate resources...”
The initial support engineer is responsible for capturing new information and adding it to the crowd-sourced knowledge base either by contributing to an existing article or authoring a new article. In this scenario, the support engineer doesn't need to start with a blank page, instead they drag and drop from the discussion thread to create a new article and the system automatically captures and recognizes those whose work contributed to the article. By capturing knowledge in the workflow and linking resolved service requests to a specific knowledge collaboration, the support engineer can maximize their time and as a bonus resolve similar customer problems faster in the future.
The initial articles are created for internal use within the TAC. Each time an article is used to help resolve a customer's service request, the engineer links the article to the service request. As multiple cases get linked to an article, the article bubbles up for conversion into an external article that gets posted to the self-service portal on www.cisco.com.
Keys to Success
- Timely and accurate matching of engineer to customer needs
- Social knowledge capture. Traditional knowledge bases are content repositories, we wanted a content creation engine that is living and dynamically updates content with new service requests.
- Gamification and social reputation. Traditional customer support departments measure performance for individual agents, we needed to encourage and recognize contributions to other people's cases.
- Ideas Exchange and Case Console. Many traditional apps are forced upon business users, while Tech Zone and the Case Console are built by engineers for engineers. Participation in Tech Zone is voluntary so we involve users in continuously improving the Tech Zone community so that it meets their needs and makes them more productive. The Lithium platform is flexible enough that we are able to take ideas and implement requests quickly.
The business impact for Cisco's Technical Assistance Center has been remarkable. Estimated cost savings for the first year are $7 million with an anticipated return of $32 million over the next 3 years. The knowledge capture and sharing drove down Time-to-Resolution, as did the collaborative problem solving. Together these drove Time-to-Resolution down by a whopping 32%. As the social knowledge base grows, improvements in level-1 resolutions and time-to-resolution are expected. Case escalations are costly both in agent labor and customer effort. Case escalations were down by 42% and case transfers were down by 11%.
The service engineers have seen enormous value in the system and have readily adopted the new capabilities which is particularly meaningful given the dramatic change in workflow. Active participation in Tech Zone is 75% and growing. The satisfaction rating for the process and tools is 4.6 out of 5 and savings in technical training is expected around $2 million per year.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.