Improving conversion rate on the website by 10% based on customer sentiment attribution
WatchShop was established back in 1991 and is the biggest independent watch retailer in the UK. WatchShop sell watches from over 150 different brands, including brands like Tissot, Versace, SEIKO, and other exclusively available watches only found on WatchShop.
WatchShop is an online retailer and for them, behavioral data and business data play a big role. Testing and optimizing UX and customer journeys in different channels is common at WatchShop. Callum Delvin-Surrell works as a Data Analyst in the Marketing team at WatchShop. Callum manages and analyzes different types of data to understand:
- How customers find their site
- How customers experience the site
Based on the data, they implement tests to improve and optimize the experience that their customers perceive. They actively use tools like Hotjar and watch session recordings to get a grip of how visitors behave on their website.
Wanted to understand how customer sentiments impact customer satisfaction
But the problem was that behavioral data doesn’t give you the whole truth. They didn’t have a way to take into account the sentiment that certain pain points might cause – they couldn’t get to the bottom of the issues that customers were experiencing.
Whenever they got actual feedback, they didn’t have a way to compile it together, so the feedback wasn’t really used for anything.
They identified a need for more qualitative data to back up the result got from the behavioral data. WatchShop decided that it was time to make a change. Callum took the lead to move the project forward.
”I decided, okay, we need something which can pull everything together in an intelligent way, which is more than just a person could do. A team of people could maybe do it, but any individual wouldn’t. Knowing the size of the scale of our company, it wasn’t an option to have a whole team spend time on this full-time”, states Callum.
Callum knew that they needed a more intelligent solution that was based on machine learning and could manage feedback in several languages since WatchShop also operates outside the UK. They needed a tool that didn’t “just do the math” but could also take sentiments into account and could show what topics following a certain sentiment contributes to improved customer satisfaction.
Callum made an extensive evaluation of tools that could help them solve their challenge. He found that many tools were a bit too broad and wouldn’t, therefore, be intelligent enough.
What drew Callum’s interest to Lumoa, was the analytical approach Lumoa has towards analyzing customer feedback.
”It seemed like Lumoa has been built from a very analytical perspective, which was what I wanted”, comments Callum.
Measure customer sentiment in multiple touchpoints
Ultimately what WatchShop wanted to measure was the customer sentiment in different touchpoints along the customer journey. Ideally, they want to run all their feedback sources through Lumoa, such as:
- Pop-up surveys on the website
- Live chat transcripts
- Email correspondence with Customer service
The idea was that WatchShop would let Lumoa analyze the sentiment of all the feedbacks, which would result in a top-level sentiment score. WatchShop wants to understand what the current overall customer sentiment score is, and what are the things that contribute to positive and negative sentiment in specific journeys.
“Our primary focus with Lumoa will be to improve customer satisfaction, but rather than looking at a certain CX score, we’re specifically going to focus on the customer sentiment”, comments Callum.
The idea behind creating the top-level sentiment score is to make business reporting more relevant for different stakeholders within WatchShop. For top-level management, they are able to show the current status and the development of the overall customer sentiment.
Then they can dig deeper into certain areas which are contributing negatively to the score and ensure that the team who is responsible for that specific area can fix the issues.
“ It just builds a very good structure in regards to how you can actually project manage going forwards to improve the overall sentiment. Each team just want to know – what where are the problem points, where the good points, and what are we doing to even leverage the good stuff and, and to address the bad”, says Callum.
When everything has been implemented the idea is that there will be two levels of key stakeholders that will have access to Lumoa.
“From Lumoa, reports will be sent to the senior leadership team. Then underneath that, it will be a delegate from all the departments involved in which areas have been highlighted by this.
With that I mean, development, finances, commercial marketing, operations and customer services. The key delegate from each one of those areas, will be responsible for anything which is fed out from Lumoa, which will go directly to each of those”, asserts Callum.
The delegate from the specific department is then free to choose how they want to delegate within the team, for example, if they want, the whole team can access the insights in Lumoa.
Analyzing customer reviews and comparing performance against competitors
So far WatchShop has integrated Trustpilot with Lumoa. This enables them to automatically fetch and analyze customer reviews that customers write on Trustpilot. In addition, they also analyze competitor reviews from Trustpilot with Lumoa. This allows them to compare their own performance and create leverage against their competitors. They will soon also integrate Feefo with Lumoa so that they can automatically analyze the reviews from there as well.
When all feedback has been gathered, they expect to analyze several hundreds of thousands of feedback with Lumoa. WatchShop is using Lumoa not only to analyze survey feedback but also to analyze feedback that is automatically fetched from different platforms via various integrations.
Next Watchshop will integrate Zendesk with Lumoa. This allows them to automatically analyze customer emails and other correspondence that customer support has with their customers.
”Over the scope of a year, you’re looking at potentially hundreds of thousands of feedbacks that is going to analyze with Lumoa, so there is going to be a lot of data”, states Callum
The results from Lumoa made Watchshop change the list of priorities
Experienced in mathematics and statistics, Callum had before taking Lumoa into use done his own analysis of some feedback related to their website’s user experience.
When they took Lumoa into use, Callum decided to reanalyze the data in Lumoa, to see if the insights from Lumoa would be any different.
In the comparison with his own analysis and the insights found in Lumoa, Lumoa had highlighted another set of pain points that weren’t being prioritized. This resulted in that they reordered their list of priorities for their user experience project.
“The sentiment scores are very valuable, in the way that Lumoa was able to show – here are the things customers commonly talk about. And here are the things which they associated most positively and negatively”, states Callum.
Improved conversion rates with almost 10%
In the user experience project, they aimed to improve their Product Listing Pages, where they based on the insights from Lumoa, tested: filter options, change product attributes, product terminology, etc.
Based on the results from the first round of tests where they tested different filter options, they were able to increase the conversion rate on their Product Listing Pages by almost 4%.
In the second test round, where they have done several UX changes, the increase in conversion rates was close to 10%.
Understanding how the overall score is impacted by certain topics
According to Callum, one of the most valuable things with Lumoa, is the fact that you can bring different forms of feedback together. See how much of the feedback relates to a certain topic and know how much each topic impacts the overall sentiment score, which allows you to prioritize your initiatives in an efficient manner.
In addition, when you start implementing changes, you can measure the change in your scores over time in Lumoa which is also very valuable.
” The most valuable thing is the sentiment attribution, and the value that is related to the topics -it’s by far the most valuable thing! Because with that, you can then prioritize the actions you got from the insights”, states Callum.
Callum concludes, “with those three things in combination, you have a very robust system. Which is really making something empirical, which in my experience, within marketing, until now, I’ve never seen anywhere else, apart from Lumoa.”