Many articles have addressed online customer service and the associated best practices to optimize the experience, reduce costs, and increase loyalty. This post will focus on how technology can improve customer service for online retailers.
1. Be Proactive: Notify Customers of Issues
Streaming data can be analyzed from multiple sources, in real-time, to enable proactive customer service. This could include data from the website, such as a customer’s clicks, as well as mobile app clicks and shipment information.
Specialized tools called “complex event processing” systems are often used to correlate all the data in real-time, analyze it based on the rules defined in the tool, and take actions to notify the customer, another application, or someone internal to the business. Storm, Samza, Apama, and InfoStreams are some of the complex event processing tools. Storm and Samza are open source offerings. My company, Software AG, makes Apama.
Retailers can use these systems to reduce call volumes and resolve issues resulting in happier shoppers and higher revenue.
A good example is Dell, which uses a system that gives its reps visibility into the operational environment that covers logistics and inventory management, resulting in proactively resolving problems or notifying customers of any issues or delays before they are impacted.
2. 360 Degree View of Customers
The concept of obtaining visibility to all the customer interactions and taking action based on that has been around for several years, but still it has not been widely adopted. The goal is not only to align customer interactions in different systems like campaigns, recommendations, and social networks, but also to correlate this data to determine the next best action to serve the customer. If a retailer knows these activities about a customer, the benefits (such as product recommendations) are huge.
Nordstrom is serious about implementing this and has reportedly allocated roughly $1 billion over the next five years on its ecommerce platform to improve upon its legendary customer service. For small retailers, this can be accomplished by first integrating the different systems in their environment and gradually building the correlations to get a full view of the customer.
3. Improve Speed of Response
Most of us have heard a customer service representative say that his systems are running slow, resulting in increased wait time. If this problem is so common, why are retailers not taking action to fix it? It requires an investment, for sure. But the money saved by serving more customers in the same amount of time, or increased revenue because the customers get answers faster, could make it worthwhile.
The easiest technical option to improve the speed of response is to move data that is accessed most frequently — such as customer profile information, product catalog, and return policies — into an in-memory system to reduce the data access time. Additionally, if a rep has to assemble information from multiple systems (i.e., billing, fulfillment, order data) before answering a customer’s question, consider assembling and loading that information into the in-memory system to make processing of queries faster.
4. Up-to-date Knowledge Base
Many retailers track customer queries by creating tickets. These tickets are sometimes handled manually (bad), sometimes in an Excel spreadsheet (still bad), and sometimes in an online ticketing system (good). The objective is not only to resolve the customer issue, but also to avoid spending time responding to the same issue later. This can be achieved if the resolutions are updated in the online ticketing system and made available to the customer in a knowledge base. Customers scan search for the issue and if they do not find the resolution, they can contact customer service.
Ebay has implemented a system where all the information is maintained in a knowledge base and customers are provided with suggestions to resolve their issues. This knowledge base has a lot of information and works well, but it is very difficult to speak to someone at Ebay, if required. This should not be the case.
5. Enhanced Search
Customer service representatives usually search a customer’s history before working on the issue. In most cases, the search is limited to past interactions with service reps. It is important, however, to search through all customer communication, including audio files for phone calls, chat interactions, and any email communications. Without this enhanced search, the reps are getting an incomplete view of the issue, leaving the customer frustrated.
Speech recognition tools enable voice files to be searched or they can be tagged with the relevant keywords manually when the rep is working on the issue over the phone. Software solutions (such as Nuance and Mavis) are available for searching through audio files and some come bundled with customer-relationship and ticketing solutions.
6. Create Service-level Targets
For the retailer’s customer service to be stellar, targets need to be defined for all key operations — such as wait times, same customer calling within a specific period of time, and time for one call. These targets can also define service level agreements, where appropriate. These targets should be linked with automated alerts that get triggered when they are not met and used to notify the appropriate staff to resolve. This will help maintain the desired service levels.
Most customer service solutions, like Zendesk, have a feature to create automated alerts, though not every retailer enables this feature. It is not that difficult if a step-by-step approach is used: (a) Define the appropriate targets and (b) create alerts to resolve the issues quickly.
7. Empower Reps to Resolve Issues
The customer should not be transferred from one rep to the next to resolve the issue. The reps need to be empowered to resolve the issue. This typically requires providing access to the reps to different parts of the system. The reps need to have the power to refund money (up to a certain limit), give away free products or services, or issue a one-time savings, such as free shipping or 25 percent off on the next order.
8. Monitor Social Networks
A negative post on a social network can do serious damage to a retailer’s brand. Consumers know that. Hence customer service teams cannot ignore social networks. They must maintain a live feed dashboard that shows the sentiments on different social networks and what is trending. Based on this, action needs to be taken.
Retailers can track this using sentiment analysis that is available in several products. Salesforce offers this feature as part of its Radian6 offering. Shopify integrates with a product called Rignite that offers sentiment analysis, among other features.
9. Keep Reps Aware of Updates
Retailers should always ensure that customer service personnel know about important changes to the business, such as site design updates, new features, new products, and new campaigns, as well as temporary issues, such as shipping delays. The reps need to see the release notes and understand what is being changed in the environment before they start receiving questions from shoppers.
This sounds simple to do. But I have not encountered a single retailer, in my 15-year career in ecommerce, that does this well.
10. Facilitate Feedback
Great customer service can only be improved if it incorporates improvements based on shopper feedback. It is therefore important to make it as easy as possible to gather feedback (positive and negative) from customers and visitors. If a retailer uses a support ticket system, then the customer should be able to provide feedback on the ticket, just like they would review a product after purchase.
For example, Sephora, the cosmetics retailer, uses a product from UserVoice to engage with customers and solicit suggestions. For a few dollars every month, a small retailer can also implement a similar solution.