Agent: Hello, thank you for calling Longbow Telecom. How may I help you today?
Customer: I think I need to reset my router but I can’t work out how.
Agent: One moment please, while I access that information.
Agent: Thank you for holding. I’m pulling it up for you right now.
Agent: Thanks for your patience. I’m just locating the right section.
Agent: Apologies for the delay, please continue to hold while I escalate this query to Tier 2 support.
Customer: Never mind, I’ll figure it out myself.
The average agent spends “nearly 20% of their time looking for internal information or tracking down colleagues who can help with specific tasks.” This lack of efficiency has not escaped the notice of customers. Over a quarter say that getting their questions answered by customer service is becoming harder.
Core challenges for contact centers
Although 91% of companies have invested heavily in contact center knowledge bases, very few are seeing real ROI. Most knowledge bases are hindered by poor usability. They contain thousands of articles and to access the right one, an agent must answer Yes or No to dozens of questions related to product and issue symptoms. Even with intelligent NLP-powered search, contact center agents often find that the system returns too many articles, causing confusion and significant delays to the customer episode.
What’s worse is that most of these articles are written by and for experts. Even when the agent eventually finds the correct article that can explain the desired issue resolution, they must often wade through long tracts of technical text to find the nuggets of information that will help the customer.
Proactive knowledge suggestions
Incorporating proactive search functions into contact center knowledge bases can enable them to deliver the right information faster, by cross-referencing individual customer histories. For example, the knowledge base can check the customer’s recent purchases as soon as they provide a reference or ID number and present information on the device that the call is likely to be about.
Some knowledge bases can also proactively present issue resolution suggestions to customer service representatives in real time. These agent assist systems provide timely pop-ups, giving the agent the information to help the customer faster without requiring manual searches. But the rise of proactive knowledge solutions is only the beginning. Leading providers are now tapping the potential of images to fundamentally transform their knowledge bases.
The visual gap in service delivery
Moving from text search to automated visual search is as significant as the switch from paper to digital. Image-oriented knowledge bases represent a massive conceptual change in the field. With a text-based knowledge base, the success of the outcome is highly dependent on the customer’s ability to accurately describe the issue, or for the agent to come up with the correct keywords. The agent then needs to understand, interpret and convey lengthy text-based instructions.
With a Computer Vision-powered agent assist system, issue identification and the knowledge base search for the correct resolution happen automatically. Image-sharing creates instant, clear understanding, builds agent confidence and reduces frustration on both sides. But how can companies get there?
Creating a visual contact center knowledge base
First, a Visual Assistance provider converts a company’s existing contact center knowledge base into a visual knowledge base. The provider takes each article and extract the visual symptoms of every device and issue described, including cables, components and LED error messages.
Then the Computer Vision AI model is trained, using several different approaches: synthetic visual data (gathered in the lab), existing visual resources and images supplied by customers.
How it works in practice
During a live video interaction with a customer, the visual agent assist system recognizes the device and the symptoms of the issue and searches the visual knowledge base in real time, presenting the agent with a proven resolution. The agent confirms the proposed resolution and sends the customer images showing the steps they need to take, using Augmented Reality annotations to pinpoint each action. Aided by the system, the agent then confirms that each step has been completed successfully and that the issue is resolved.
Additional benefits of a visual contact center knowledge base
Visually connecting teams
The trend of outsourcing contact center operations has been a defining theme of the field for decades, while distributed workforces are an increasingly popular way of tackling the perennial recruitment crisis by allowing agents to work on their own terms. However, this raises the major issue of aligning agents in multiple locations. Implementing a visual knowledge base is a straightforward way to achieve this, since in most cases, there are no linguistic elements that could be misinterpreted. A visual knowledge base enables any agent to instantly become an expert on any issue.
Dynamic visual knowledge bases also incorporate statistical models, enabling contact center managers to identify the most common issues, and the average time it takes to complete each step of a resolution.
Building an effective visual knowledge base relies on Computer Vision and Deep Learning models that must be trained with a huge volume of labeled data. A company-wide crowdsourcing effort to tag images of interactions enables the system to get smarter faster. Through automated recognition of similarities between devices and issues, the visual knowledge base becomes richer and more efficient. In other words, it applies its own logic to create new solutions independently.
A visual contact center knowledge base is the key to delivering superior customer service through Computer Vision-powered agent assistance. The ability to automatically recognize devices, parts and issues symptoms in real time transforms the process of identifying the right resolutions. With 31% of customers stating that a knowledgeable customer service representative is the most important aspect of any interaction, implementing a visual contact center knowledge base can be just what the contact center needs to succeed.