BrandGPT and the future of CXJanuary 10, 2024
Technology choice, adoption and the paradox of choiceJanuary 18, 2024
Today’s interview is with Tue Martin Berg, the CEO and part of the founding team at Capturi, Scandinavia’s leading software provider for conversation analysis. Tue joins me today to talk about what they are up to, why some Scandinavian organisations who after implementing call/interaction analytics solutions from one of the big US providers, then ripped it out, whether the problem they faced applies to all brands that operate in languages other than English, the impact that has on their ability to offer automated conversational features as part of their services and what many Scandinavian organisations are now doing and achieving with the help of Capturi.
This interview follows on from my recent interview – Are you doing change to people or with people? – Interview with Phil Lewis and Claire Croft of Corporate Punk – and is number 490 in the series of interviews with authors and business leaders who are doing great things, providing valuable insights, helping businesses innovate and delivering great service and experience to both their customers and their employees.
Here are the highlights of my conversation with Tue:
- Our mantra/DNA is that a conversation is just a conversation until someone takes that conversation and uses it to drive a change or improvement. And it’s not until that action is taken that value is actually created.
- A Scandinavian (Danish) company implemented a call/interaction analytics solution from one of the big providers (US-based). But, a short time after implementation, they shut it down because the accuracy of its analysis and insights was not very good at all.
- Many of the big US. players have really cool solutions but they base their technology on what they’ve learned in the US and use English as the foundation for that.
- Danish and the other Nordic languages are minority languages that differ very, very much from English. So, when people were trying out the US solutions, the quality of the technology translating the spoken into text was simply very poor.
- These (US) technologies couldn’t capture the linguistic differences or some of the cultural nuances.
- The biggest problem for minority langauges is available training material.
- There basically isn’t any when it comes to the Nordics. So what we set out to do was, first of all, create the training material, but also focus on building the engine that allowed us to focus on Danish, on Swedish and on Norwegian, so that we could take into account those lingusitic differences.
- Similar situations exist with other languages to varying extents.
- The bigger the language, the bigger the potential for training material to exist.
- In the Nordics, even though we are relatively big from a land perspective, we are relatively small from a population perspective and the languages are often viewed as odd. I think Danish is described as one of the most difficult languages to learn because of all the exceptions and the oddities of the language.
- From a channel perspective, I see voice becoming increasingly important as well as voice bots.
- By analyzing more than 40 million conversations a year, we’ve built an incredible understanding of what customer service is going to look like in the Nordics and how to apply that based on the insights that we’re able to gain.
- That local understanding is what I think will ensure that we will have a relevant role to play.
- Some of the bigger companies that chose not to continue with the US-based vendors was not only because of the technology, but also because of the insights and the advice coming from them that reflected a one-size-fits-all US-based vendor approach, which for many Nordic companies is not how we go about it.
- There is a common perception, but it’s also a misconception that says good customer service or good customer experience is the same wherever you go.
- In reality, it looks different to different people. We can maybe talk about it in the same way, but how it is in actual practice and what people prioritize and what’s important for them is different across the world.
- Taking a language-specific approach can be slower and more expensive. But, taking that approach also impies a keen and sincere interest in what the customer thinks and the customer experience.
- If you’re always looking for the cheapest price but still expecting the best quality, you’re deluded.
- One large customer wanted us to demonstrate value within six weeks. So, we worked with them to look at their call tagging. This was a manual process before we came on board. We worked with that customer to automate the call tagging based on our analysis of the conversation so that we could identify the main topic but also subtopics. We did that within six weeks and that resulted in a cost saving of around 250,000 euros a year. And, the employees were very happy too as they were freed up from doing the tedious task of manually tagging each call.
- Whatever type of flavour of AI that you’re looking for, everything’s built on data and the quality of the data drives everything.
- I think we have reached peak hype in terms of generative AI. What is absolutely certain is that there will come a point where there will be a dip because some people go in and stuff will go wrong. Or they move too fast and stuff will go wrong.
- Tue’s best advice: Start truly listening to your customers and employees, respect what goes on in those conversations, and use it as an asset in terms of making better and more informed decisions. It will improve your operations.
- Tue’s Punk CX word: MTMOWYAG or Make The Most Of What You’ve Already Got
- Tue’s Punk XL brand: whiteaway (https://www.whiteaway.com/).
As co-founder and CEO of the Conversational AI software provider, Capturi, Tue brings years of experience in transforming call centers into the centerpiece of business initiatives and data-driven operations.
With his legal background and entrepreneurial expertise, he excels in helping companies transform their raw conversational data into valuable initiatives based on data analyses and AI.