SSE Project – Tracking new ‘Live Chat’ feature ● ● ● ● ●

Client Name: SSE Client Contact Information (phone/email): Vince Keay, 0203 9227 7311 Client URL: www.sse.co.uk, my.sse.co.uk Client GA Account Number: UA-61342901-1 Background: SSE are one of the top 6 energy providers in the UK. In addition to Energy, they sell home services such as boiler repairs, and telephone and broadband services from their website. The website also provides online services to existing customers such as viewing and paying bills, giving meter readings and assessing energy usage.

Technical Ability (3-4 Paragraphs) SSE were installing Oracle RightNow Chat Service to provide another point of engagement with their prospective and existing customers in addition to their website and call centre staff. The live chat has rules which fire two journeys, one a service based chat and the other a sales based chat, depending on user behaviour on the site. That functionality is all managed within the Oracle tool. SSE wanted to understand the entire customer journey through the chat tool and be able to report that using Google Analytics. Much of the journey was being reported through the Oracle tool, but key elements, such as what percentage of users were being shown a chat invite, were not available via Oracle reporting. The invites appear and disappear on certain pages depending on agent availability in the contact centre, so the first challenge was to find a suitable trigger for when an invite was being shown. We set up a JavaScript listener in a custom HTML tag to push an event into the data layer when one of the 3 different chat invitations were presented to users:

The chat window itself is served up on a different domain: sse.custhelp.com. We used the same UA ID and GTM container for the chat window and amended the auto link domains in the page view tracking to ensure smooth cross domain tracking. Analytical Ability (3-8 Paragraphs) There were a number of factors that SSE wanted to understand about their chat facility. The percentage of users were presented with an opportunity to chat. Of those, how many responded. Of those, how many entered the required personal details. Of those, how many went straight through to a chat and how many queued. Of those, how many persisted through the queue and how many abandoned. How many agents started chats, and how many users actually responded and started chatting. Finally, how many users successfully completed and closed the chat window. We designed the tagging solution to be able to capture users at each of these points in the journey and sent them as events to GA. We then used the results to construct funnel reports showing the success rates of different types of invites, the impact of queueing on willingness to chat, and the impact of having had a chat on ultimate conversion rates. Business Impact (2-3 Paragraphs) The chat feature has proved to be very popular with users. We have demonstrated that

showing the chat invite to around 8 percent of visits is optimal based on the availability of customer service agents. We have been able to show that minimising the amount of time queueing optimises outcomes from chat from a conversion perspective. The full business value has not been realised as yet, but SSE have all the information they need at their fingertips to make decisions around chat.