Contrary to what one might think, the use of email to reach customer service is still on the rise. There are many reasons for this:
Discretion: calling a bank or insurance company contact centre while being at work would probably not be appreciated by our colleagues (and even less so by our boss!).
A 24/7 channel: we can send emails whenever we need to and we can read the answer in detail whenever we want, which is not the case for phone calls.
Confidence: unlike the phone, emails offer a formal way of communication with a person or company, which gives the user confidence.
In the past, contact center infrastructures have been equipped with several tools to efficiently manage phone calls, but often one aspect has been completely neglected: a way to properly manage incoming emails and leads.
The consequences of an unsatisfactory interaction with a customer via email are much worse than in the case of a phone call, since the customer has a written copy of the answer that he or she can quickly spread on social networks.
Inbenta has helped many businesses to improve customer service using advanced automatic Natural Language Processing (NLP) technologies. Thanks to this experience, we have learned the important points to keep in mind in any customer service environment. These lessons can be summarized in 5 simple rules:
1. Communication, the key to managing incoming emails
Let your customers know as soon as you receive their email. Users will appreciate getting an acknowledgement of receipt and being informed of when to expect a reply. Ideally, this email notification should contain a case number or unique identifier so that your customer can refer to it in the event of a new email (or phone call).
2. An interaction history for better follow-up
Your agents must have constant access to all customers’ emails.
“I was told this information had already been updated.”
“I’m sorry sir, but I can’t see anything about this.”
Similar conversations often take place in call centres, resulting in frustrated and angry customers. Written documents are an excellent opportunity to avoid this kind of situation in the future. To do so, agents need to be able to easily access the email (and phone call) logs for all customers.
3. Predefined but qualitative answers
Ready-made responses can be good resources, but only if they are of excellent quality. One of our clients wanted to create and store a selection of standard replies based on the answers used by their call center agents. The idea was to create a knowledge base using those answers that were already written. Using a search engine, agents could then find these ‘solutions’ and supposedly reuse their content.
Unfortunately, this ‘knowledge base’ had more than 15,000 entries within a few months and most of them were duplicates, which made it difficult for the engine to find the most adequate answer. As a consequence, the search engine became obsolete, as did the knowledge base, and agents swent back to writing the answers again, even knowing that 80% of users’ questions were repetitive.
Standard answers (also called ready-made answers or FAQs, Frequently Asked Questions) are an excellent tool in terms of productivity, but only if the following conditions are taken into account:
There should be an editorial committee, who is in charge of reviewing and editing the responses on a regular basis so that they don’t become obsolete or (even worse) inconsistent with each other.
An intelligent search engine should be used so that agents are able to find the right standard answers easily and quickly. Using automatic Natural Language Processing (NLP) and semantic search tools has proven to be very effective with this kind of use.
The extent of the knowledge base needs to be closely monitored. Based on our experience with many of our clients, a knowledge base containing between 500 and 2500 FAQs is strong enough to meet 80% of the requests while remaining small enough to be easily managed.
4. Customers enjoy self-service when it’s performing perfectly
If rule #3 has been correctly followed, this one can be easily applied: let users search by themselves in the same knowledge base used by the company’s agents.
With several of our clients, the amount of incoming emails and phone calls has been reduced by about 90% just by allowing their web users to search for answers via online solutions such as question/answer systems, virtual assistants and semantic search engines.
In the event that a customer can’t find the answer to his question, he will have the option to write an email or use an online form in order to get in touch with customer service.
Tip: do not ask your customers to repeat their question or complaint when they transition from web self-service to email or phone as there’s nothing more irritating than having to write this twice, (it’s a common mistake in call centres, so here’s a good opportunity to avoid it).
5. Measuring the level of service
Any email management tool in a contact center should be able to deliver the following statistics: number of emails received and sent, average minimum and maximum response time, SLA (Service Level Agreement), performance level, productivity and performance of each contact center agent, complete traceability of incoming and outgoing messages and so on.
Far from being used only as a tracking system for agents, these statistics should be used to identify aspects of the user experience that need to be redesigned and the type of communication held with customers, allowing you to improve the performance of the tool over time.
Based on our experience in various industries, there are other aspects that should not be overlooked, such as:
Received emails represent a good opportunity for cross-selling
Answers should be written using the same writing style and tone of voice
The system should be compatible with your company’s CRM
It can be integrated with your CTI system to avoid receiving phone calls while answering emails, etc.
While the amount of customer service emails will keep on growing in the upcoming years, we firmly believe that more and more new tools will take these and many other features into account.
If you would like to go further and learn how your omnichannel strategy can be served by bots, then our guide is made for you.
Check out our similar articles
- 8 Must-Have Chatbot Features to Improve Customer Experience
- Why Self-Serve Support Is Only as Good as Content
- The Most Popular NLP Use Cases