For live chat to be really effective, it is important to understand the key information that users of the service generate. What should you be measuring and how can it be used?
The last blog on testing your live chat service detailed the process of defining test coverage ie ‘what’ you will test. To keep things detailed enough for in-depth learning and continuous improvement, yet simple enough not to stifle action we test 10 areas here at Yomdel.
- Location of chat button/ image/ text
- Design of chat button/ image/ text
- Chat interface end-user experience
- Proactive chat invite effectiveness
- Operator efficiency
- Operator language skills
- Operator client/ product knowledge
- Operator professionalism
- End-user common questions
- End-user satisfaction
The next step in the design of a test plan is to identify your information requirements. This is the specific information that you will need to gather during testing in order to ensure that you achieve your desired test coverage.
At Yomdel we know we’ll need to gather the following information:
- Conversion to chat ie what percentage of those who chatted went on to perform a specific action such as buying a product, requesting a call back or signing up for a newsletter?
- Average time spent on chat ie how long did the average chat last?
- Chats per operator per hour – how efficient were the operators at handling multiple chats and how quickly did they give customers what they needed?
- Ease of use for operators – do operators find the system easy to use or are there improvements that can be made (or additional training requirements)?
- Ease of use for end-users – does the technology offer a good customer experience?
- Top questions asked by end users – what’s important to end users? Is there content missing from the website? Are these areas covered in the operator knowledge base? Can we develop canned answers for some of these questions to increase operator efficiency?
- Answers given by operators – and importantly, how end users responded to them
- Outcomes of chat – positive and negative
- Average order values – keep an eye on this to make sure it goes up. If it doesn’t you have a problem. If it does why? And what more can be done to increase it further?
- Customer satisfaction scores – our operators ask end users to complete a post chat survey. This way we hope to monitor the effectiveness of every chat and zoom in on the reasons behind lower scores.
Time to dig deeper
Reading through the above you should get a feel for how the information that we capture can help pin-point specific aspects of the service that are doing well and not so well. Once you have this information you can dig deeper, so as to find reasons for outcomes. This is when the continuous improvement process really kicks in – as you develop strategies for repeating excellence, capabilities to better serve customers and find ways to address problems.
In a nutshell, once you understand how well your service is performing you can begin making improvements.
However, capturing information is never an easy process. Whether it’s the information that I have outlined above or some other information you require – resources and processes will need to be utilised in the gathering of that information. Remember that the aim of this blog is not to tell you what to do, but to advise you on how to do it. To see how I got to the information requirements above – if you haven’t done so yet –read part one and part two of this series.
The areas of resources and processes will be addressed respectively in parts four and five.
Until part four I wish you all the best in your endeavours to build the perfect live chat solution for your website. If you have any comments or thoughts on information requirements for live chat testing or would like more information about any aspect of live chat please do comment below.