Skip to navigation Skip to content


Tadas Labudis - Prodsight

Making sense of customer feedback…

Tadas Labudis - Prodsight

Making sense of customer feedback

Successful companies not only make money by keeping their customers happy but also by listening to them, to understand their needs and find out what they think of their product or service – but if they’re faced with zillions of comments, how do they get “actionable insights” from the noise? Based in Edinburgh and hosted by the University of Glasgow, Prodsight has developed a software solution to automate analysis of customer data, so companies can make sense of feedback and make better business decisions...

Since he came to Scotland from his native Lithuania in 2008, Tadas Labudis has never stopped working – or learning. After two years at Glasgow Kelvin College, where he studied for an HNC in Business, Labudis graduated with an MA in Business & Management at the University of Glasgow, and while he was still studying got involved in a couple of projects – establishing an online drum store, creating a social network for students and setting up an award-winning personalised event discovery platform called Eventhread.

For Labudis, this experience was not only useful in learning the basics of business, but also gave him valuable insights into product development which helped him in his first jobs after graduating five years ago – and when he went on to found Prodsight in September 2017.

“When we developed these websites, we were blind to the needs of our users,” Labudis explains. “We did no research and never asked our customers what they thought of what we were doing. We were totally focused on pushing solutions and assumed far too much about user requirements.”

In his first jobs after leaving university, Labudis continued to learn about product development in the commercial environment. His job at Kotikan was helping to develop mobile apps, and this involved conducting interviews with customers. At Yavi, he led a team developing a mobile messaging app for high-street food and retail chains, and also got involved in user research sessions and usability testing. As he processed the data from these conversations, he started thinking more about customer feedback and spotted a gap in the market – and was so inspired he quit his job soon afterwards to start his own business.

Labudis explains: “We were trying to listen to what users were saying, and there were hundreds of support tickets – full of requests and reports about bugs in the software, etc. It was easy to be overwhelmed by the comments but we knew there were valuable insights to gain, so I created a spreadsheet to start analysing the data. At first, it was impossible to make sense of the noise – the anecdotal data based on ‘gut’ feelings – but gradually patterns began to emerge.”

Building in-house tools to analyse the data would have been the logical next step, but would have been a “detour” for Labudis in his day job, developing software. Some big developers do build their own analytical tools, but smaller firms do not have the resources to do so because it is a relatively small part of their core business. And when Labudis saw the opportunities out there, he decided to go it alone by setting up Prodsight, with a mission to develop a product which would “help companies make data-driven product development decisions,” by identifying usability issues, software bugs and feature requests, using Artificial Intelligence to turn the raw customer data into meaningful insights.

Having been a product manager himself, Labudis started asking other people what they thought about his idea, and started building confidence in what he was aiming to do. He also wanted to establish if analysis of customer feedback was a problem for specific types of companies or a much broader issue for all firms. “It’s good to be niche,” says Labudis, “because it helps you focus. But to build a sustainable business, you need a much wider appeal.”

First subscribers

Soon after setting up the company in Edinburgh’s CodeBase technology hot-house, Labudis signed his first subscriber, a mid-sized US company with 200 employees, quickly followed by several others, paying $5 to $100 a month. For Labudis, this was not only paying the rent but a learning experience right from the start, with six months of savings to keep the new enterprise going – and a vision to sell to subscribers.

According to Labudis, there are two major problems in product development and customer service, when it comes to making use of feedback. First, as companies scale, the volume of customer feedback increases and they need to hire more support agents to deal with the traffic. By using Prodsight to analyse the conversations, says Labudis, they can identify common customer queries, create self-help material and thus reduce the volume of the traffic as well as the cost.

The second problem is strategic, says Labudis. In a highly competitive landscape, companies are constantly looking for ways to gain a competitive edge through innovation. Prodsight is designed to help identify strategic and tactical opportunities to improve products or services by highlighting customer “pain points” discovered in support conversations.

In addition, says Labudis, about 80% of customer feedback revolves around two or three issues. And as companies grow and become more successful, the further away their original founders and management layers will be from their customer base – and the more they need to find a way to keep in touch with customers.

Prodsight’s simple vision was to automate analysis of customer data, gathered from ‘live’ chat, sell the service to subscribers, then ask them for feedback to improve the product and sell it to even more clients. In other words, Labudis knew the best way to develop his solution was to practice what he preached and do exactly the same as his clients – a total of 90 subscribers within the first year.

Different clients wanted to do different things. For example, some simply wanted to archive their customer data, while others wanted to know exactly what their customers were saying about them and learn how to improve their service or product and prioritise what they were doing. Initially, Labudis had to analyse the data by himself. This may have been a time-consuming process for the entrepreneur, but it has given him invaluable insights into market requirements. “It also proved that if I could automate what I was doing manually and scale up the product, the market potential was huge,” he explains.

Labudis was also delighted to see that his product was already leading to positive outcomes for clients, as they identified problems and solved them.

After four or five months, the prototype solution was able to “crunch thousands of customer conversations,” and Labudis started looking for investors to match his ambition. “To develop the product, I needed to hire the right people,” he says, “and that meant I needed investment.”

Investment drive

Labudis had never attempted to raise money before, but in May last year, he was successful in applying for an RSE Unlocking Ambition Fellowship (hosted by the University of Glasgow), and managed to attract his first investors, introduced by Gavin Dutch, his former boss at Kotikan. Labudis was able to prove that his idea was already working and demonstrate he also had the confidence of customers, but next he had to persuade his investors that the product could scale and provide the intelligence needed to analyse vast amounts of customer data, taking advantage of recent advances in natural language processing technology to make sense of the noise.

This first round of investment raised a total of £70,000, on top of the £45,000 from the RSE Fellowship programme. His early backers included a number of prominent angel investors such as Alistair Forbes, Rob Dobson, Judy Wilson and Andrew Barrie, as well as Callum Forsysth, Robin Knox and Paul Walton of SeedHaus. This enabled Labudis to hire two employees – a data scientist and someone who could productise the prototype solution, and help him make his vision a reality. “To unlock the investment and build up the team, I needed evidence, but before I knew it,” says Labudis, “the investment had snowballed.”

As a solo founder, Labudis is now getting ready for Phase 2 of Prodsight. “I have cracked the first part of the puzzle,” he says, “but now I also have two employees to care for."

Future plans

Prodsight currently focuses on text-based feedback gathered from customer live chat, analysing all the words to find any relevant patterns – inferring structure from the text. Even small companies have thousands of conversations with customers, so to capture that and turn it into useful conclusions is “spectacular,” according to Labudis, who is also being helped by two natural language processing specialists from the University of Glasgow “to refine the conversation analysis processes.” In the future, similar solutions may also analyse telephone-based conversations, using existing voice transcription techniques, but Labudis says that text and live chat will continue to be a more popular option for customers who grew up in the digital age (please see below).

Armed with Version Two of his software, Labudis will now build his customer base and seek to convert his existing subscribers, in the process proving that Prodsight can handle much larger volumes of data, as “the serious business” begins, with subscribers paying up to $600 per month. One key advantage, he explains, is that subscribers can manage the service themselves without needing direct support.

Labudis knows he has to prove the numbers first and show that his model will work, and scalability will also be a critical factor, as he seeks to unlock more investment – the third round of investment is already “well in advance.” Another key part of the marketing strategy will be case studies, identifying companies who focus on customer service or product development who can act as ambassadors and pilot the software.

The market potential is huge, says Labudis: “The customer feedback analysis space is still a relatively young market but it is growing very fast – there are currently some 1.5 million companies across the world using live chat functionality and generating billions of customer conversations.”

The solo founder may have built up his business the hard way by using his own cash as “a runway” to future success, but his vision has always inspired him. “Our product gives you clarity,” Labudis explains, “and is all about keeping your focus in the midst of the customer noise. If you can learn from your customers faster, and connect with them better, and analyse what they are saying and what they require, you can then make much better decisions and prioritise product improvements because you are better informed.”

And if Labudis uses his own clever software to make it more clever, he should be able to achieve the same for Prodsight that he promises subscribers – completing a virtuous circle of self-powered growth.


Live chat becomes more popular

According to a number of researchers, live chat continues to be the preferred way to contact online companies and usage is expected to grow by almost 90% over the next two years. The reason for this rise in popularity is that about 50% of all customers would rather chat with someone online in real time than call up on the phone for support. Another survey found that 42% of online customers prefer to use live chat compared to just 23% opting for email, and 16% choosing social media or forums. In another study, 73% of customers described live chat as “the most satisfying” method to communicate online, while 63% said they were more likely to return to a website that offers live chat.

Despite the mounting evidence, only 9% of online companies use live chat on their website, even though it enables a single customer support person to deal with up to six customers at once, compared to phone and email support, when you can only respond to one customer at a time. In addition, the average response time for emails is 12 hours, and the average response time for social media is ten hours, compared to minutes for live chat. Having to wait for responses is one of the reasons consumers take their business elsewhere, while provision of live chat has even been cited as a main reason customers buy from a website – and thus a major boost to sales and revenue.



"Tadas Labudis - Prodsight". Science Scotland (Issue Twenty-three)
Printed from on 03/07/20 11:44:02 PM

Science Scotland is a science & technology publication brought to you by The Royal Society of Edinburgh (