Online reviews form a major part of the modern consumer’s journey when selecting goods and services. Recent research from the US claims 85% of consumers trust online reviews at least as much as personal recommendations and more than 95% look at online reviews regularly.
As an example, trustpilot.com receives more than five million UK visits per month and is just one of many sites aggregating user feedback. As consumers increase their use of and trust in these kinds of sources, they become a fascinating new competitive arena for companies seeking to differentiate and excel.
However, with thousands of scores and long form reviews written by consumers, there is also a wealth of information that can be used both to provide insight to support improvements in customer experience. In addition, the fact that these sites are public and carry reviews of multiple providers in the same space allows for an assessment of a company’s ‘experience landscape’ including benchmarks and the opportunity for some simple competitive intelligence gathering.
KAE has gathered over six and a half thousand reviews of UK high street and challenger digital banks. We used several natural language processing techniques to try to understand how words and phrases lead to positive and negative reviews.
Looking at overall review scores over the last three years, we have found some results should be of significant concern to the high street.
Figure 1: Average review scores and trends for UK financial services providers (scores are recorded as a number between one and five)
A simple review of average scores, and trends in scores reveals some startling results:
- UK challengers (Monzo, Revolut etc.) outscore the high street banks (HSBC, Lloyds etc.) to an extraordinary degree. In addition, their users are much more digitally engaged, publishing twice as many reviews on average, despite lower numbers of accounts held
- Despite the amazing beginning for the challengers, familiarity seems to breed contempt in banking and it seems like at least some of the gap can be accounted for that the new players haven’t had time to annoy their customers yet. The challengers’ scores have fallen by an average of more than 13% in the last year. Revolut’s scores are showing the most staying power, but Monzo and Starling have seen rapid falls (15% and 16% respectively)
Further insights can be derived from a year on year analysis. Since 2016, Barclays has been improving its scores, and is the only consistently positive story in the entire marketplace. Other players are either consistently negative or erratic, perhaps indicating a lack of defined and actionable CX strategy.
Figure 2: Average review scores for high street and challenger banks 2016-2018
Language processing to extract the ‘why’ behind the data
While the scores and timing of reviews are powerful, the reviews contain another highly valuable data set, the long form reviews themselves. The downside of all this is the volume. For this research we were able to accumulate over 600 pages (size 11 font) and more than 500,000 words of reviews. This, to provide some extra context, is slightly longer than J. R. R. Tolkein’s complete Lord of the Rings trilogy, and with all due respect to the UK consumer, is not nearly as well written; simply reading through it all is not an option.
To process this volume of information, we turn to natural language processing to try and find out what’s going on. In this case, we use a simple classification system to try to understand which words and phrases correspond to high and low review scores for each bank. The motivation behind all this is to help us understand why the scores (which seem to correspond quite well to published NPS figures) and trends are what they are. For example:
The disruptive challenger banks are having an impact on the UK banking market. Their reception is vastly more positive than the traditional high street banks, and despite a fall in their popularity, they are still some way ahead. A look at the terms positively correlated to their most favourable reviews reveals additional detail
- Monzo – users are happy with the simplicity of Monzo’s service, comparing it favourably to banks. The exchange rates offered by Monzo are also cited as being much better than the traditional alternative
- Starling – the app has received a lot of praise, as have the fee structure and the interest rates
- Revolut – the speed of service is consistently mentioned, as is the noticeable ‘improvement’ on the service received compared with high street banks
All three of these challenger banks, and more so within the last six months, have however received more criticism as more users are exposed to their services. Failures to process transactions, freezing user interfaces and blocked accounts are all reported as being issues by those who are unhappy with them.
It seems reasonable to expect that the gloss might come off the challenger banks as their systems are exposed to higher user numbers, but this kind of analysis can quickly and efficiently provide the kinds of insights that could help them maintain their considerable advantage.
Using this kind of analysis, consumer facing organisations gain access to a potentially significant source of insight. Online feedback is by no means representative, but provides both rich, text-based data and benchmarking that is both timely and highly efficient. KAE believes it should form part of a broader strategy to build a comprehensive view of the experience an organisation provides.