How TfW used a quant/qual cocktail to recommend changes to teams with data-backed confidence
Passenger safety and satisfaction are TfW's watchwords, so finding new ways to keep their travellers safe in the midst of rapid change was essential.
Safety in numbers
"Is there more that can be done to help passengers making essential journeys feel safe?"
When passenger safety is at the core of your brand’s values, this is a question you're going to ask yourself again and again.
But Passengers don’t always complain in plain sight, or even in the same way. Especially when talking about feeling safe.
So how do you analyse siloed data to get the clearest picture of how customers actually feel about a topic you care about?
This was Transport for Wales (TfW)’s main objective when they first started using Wordnerds.
Keeping Wales safe
TfW is an enterprising and dynamic train operating company that put passenger experience, happiness and safety at the forefront of everything they do.
With hoards of social media, survey data and customer complaints on their hands, TfW needed an accurate and efficient way to analyse the data.
Getting to the root cause of issues quicker would mean they could take action on what mattered when it mattered most.
So finding a solution that meant they could confidently ditch manual management of the data was imperative, to save time and effort needed to identify and act on emerging trends.
Kicking things off, social media mentions and public untagged Tweets were set up for tracking on the platform, then customer complaints and survey data were imported for analysis.
This meant they could analyse all of their data on a large scale in one place, and begin building a single, holistic view of their customers’ voice.
TfW trained and fine-tuned their own artificial intelligence in the platform* meaning they were able to find and surface common themes within all of their data.
*This isn’t as technical or time-consuming as it sounds. In fact, it’s so easy, even an eight-year-old can do it.
This flexible text analytics solution enabled TfW to:
- See the issues impacting sentiment or being talked about most - and trend how it was changing
- Group and analyse their data by the subjects that they cared about with keyword and context searches
- See what their customers were talking about, right out of the box, with our unsupervised text clustering (unknown unknowns)
- Search and discover the conversations that they cared about without having to second guess every keyword that customers might be using
An early finding was how passengers felt about their safety whilst travelling during the third (!!!) lockdown, which did not reflect the high standards that the brand had set for itself.
Taking this insight to the next level, they commissioned us to flex our data chops in a report where we conducted a deep dive analysis, exploring issues that passengers weren’t flagging with TfW directly, in untagged social comments.
Key result — The project highlighted concerns around onboard mask compliance, despite assurances from station teams that they were enforcing the rules.
Key result — TfW Insight & Innovation Manager (and all-around legend) Mike was able to present this finding at his weekly board meeting and request that action be taken.
Key result — TfW used this insight to alert their onboard teams, and increase messaging and enforcement around compliance.
As a result of this proactive approach, TfW saw mask compliance sentiment ascend into positive territory in the following reporting period as TfW were better able to help their passengers feel safe whilst restrictions were in place.
We are now providing TfW with real-time actionable insights from social conversations around performance issues, key stations and monitoring passenger experiences to find areas for improvement.