EXCLUSIVE: “How to REALLY Profit From Every Transaction” – Kirill Lisitsyn, Torus, Breno Oliveira, payabl. and Maria Komissarova, Raiffeisen Bank International in ‘The Fintech Magazine’

Transactional profitability is a simple concept – which units of sale to which customers are making you the most money? But it gets a whole lot more complicated the bigger you get

We all want our businesses to make money. And for most of us, that means ensuring that every single transaction costs less than someone pays for it. The simple formula is profit = revenue – costs. But how does this work in practice? And is it really that simple over the client lifecycle? Could you find yourself saving a penny now, but losing potential pounds? And at what point do you hit false economies?

We brought together three experts who are passionate about the long-term value of transaction profitability to talk about the changing priorities in that calculation and what firms must do to survive. Breno Oliveira from payabl., which connects payments and business accounts in one platform, Kirill Lisitsyn from Torus, a payments profitability intelligence platform, and Maria Komissarova from Raiffeisen Bank talked about the business-critical nature of data analysis, how to create value from hidden costs and the power of identifying your niche.

What lies beneath transactions

Transactions today come with a multitude of hidden costs. A big one is data. In the past, a licensed corner shop probably didn’t need a large language model to help the merchant figure out if the young professionals buying up their big-ticket wines were making them more than the 50 school kids who came in every day for a packet of crisps.

But when you’ve thousands of product lines and millions of customers, it’s a challenge. And that goes for every part of the transaction chain – whether you’re a big bank, like Raiffeisen, which is both a card issuer and a card acquirer and so needs to know how to price fees in the context of the bigger value a merchant brings, or one of payabl.’s client businesses, making sales every day.

As businesses strive to build loyalty, collecting data becomes critical… and that comes with a price tag.

“It’s costly,” admits Head of Data & AI Delivery in Retail at Raiffeisen Bank International AG, Maria Komissarova. Globally, the data storage market is already valued at $298.54billion, and expected to more than triple by 2034. Average costs per business are murky to calculate. But with a terabyte (tb) of data costing around $400, and with mid-size firms storing between 10 and 50 tbs, storage is $4,000 to $20,000 annually.

What businesses do with that data determines whether the money for storage was well spent. It’s not ‘the profitability of each transaction, but how we utilise this information afterwards and build some value on top of that’, explains Komissarova.

“But if you do not make use of it, you start losing in this pricing game because you already start by having a huge cost that you don’t have a profit value for. If you don’t know those stats, then you can’t provide a better price and be competitive.”

Salesforce research in 2025 found 63 per cent of companies struggle to align business priorities with data. Firms are feeling pressure to buy expensive data capabilities, without using it to drive revenue. It’s a double loss. Data storage isn’t the only drain on profitability lurking within transactions, the round table pointed out, highlighting fees and customer experience. The process of uncovering costs takes some serious investigating, and first impressions can often be deceiving.

“When we start talking to our prospective customers, in many cases portfolio-wise, business looks good,” says CEO at Torus, Kirill Lisitsyn. “But when you drill down at the merchant level, there is usually some margin erosion.”

To uncover these pitfalls, Lisitsyn will examine many factors, including specific transaction types by geography and card, among others.

Factoring in costs

The transaction cost only looks at the transaction. It doesn’t consider the cost of making the goods or service, which comes up as ‘costs of goods sold’ on the balance sheet, nor any of the salaries, nor business operational expenses. So, at first glance, it may feel like the transaction cost starts and ends with the transaction processing fee. But it’s actually much deeper.

CPO at payabl., Breno Oliveira, urges firms to look towards the ‘total cost of ownership, settlement time, what type of user experience you provide, if a transaction can be within 48 hours, and authorisation rates’.

It can be difficult to quantify user experience, for example, but with shopping cart abandonment hitting 72 per cent, it could be a huge loss-maker. If fewer than three in 10 customers follow through, the costs of supporting 10 customers to the point of transaction actually falls on the three who paid. And that can be an eye-watering issue.

As well as tracking its own transaction costs, Raiffeisen Bank International is on a mission to give business customers the same clarity.

“We have a more holistic picture, we have more information on the market, and we can produce insights that might be very useful for the merchants for their own business”, says Komissarova.

As well as giving ‘clear and clean’ calculations, the bank also offers ‘insights on the ways to improve the businesses’. After all, healthy cashflows are in everyone’s best interest. Collecting data to give a true picture of transaction profitability has been challenging because the various departments are siloed and information is fragmented and hidden from each party.

Overcoming inertia

Siloed legacy systems with fragmented data quality not only affect one firm, but all the underlying clients, too.

“Many times, merchants are plugged into financial institutions with static routing. This can result in stagnant insights and even data leakage. To stay competitive, banks must conduct regular health checks not just for themselves, but their clients, too,” says Oliveira. “If a merchant is in hyper-growth, it needs to be revisited on a quarterly basis.”

So banks’ fragmented data systems could be impacting competitiveness as well as profitability. The longer it’s left unresolved, the worse the problem becomes as data stagnates. Today, banks face the triple challenge of needing to upgrade their systems, understanding how to use the data correctly and then bringing the various data types and sets together coherently. But first, they must overcome the inertia of embracing new data processes and ways of working together.

“We have a lot of good technology in place. The problem is not whether you have the right or wrong technology, it’s about how you treat your data,” explains Komissarova. “Data should not be treated by IT people in a silo, but by the product team. Each team needs to provide their data in a format that is readable and then the other parts of the organisation can generate the profit on this data”.

All three agreed this is why removing data siloes is so fundamental, not only for profitability but also collaboration.

“The biggest problem [with a bank] is ownership”, pinpoints Oliveira. “There’s a risk team, a finance team, operations, product… and a single team does not own anything”.

Changing this could create vast new efficiencies and opportunities for value creation.

“Whenever we give the ownership to a specific organisation inside of the company, then things turn around because we have more skin in the game,” Oliveira adds. “We can make more optimisations and hold more people accountable.”

The correct data architecture

Once banks embrace data optimisation and accountability, one of the first – far from clear-cut – decisions is figuring out the new data architecture. Should firms opt for a data mesh, data fabric, data lake or data warehouse or even a data lakehouse? And more concepts are developing all the time.

“That’s what fintechs do very well”, says Oliveira. “They access many datasets and try to transform them into formats that are easier to digest. Because when we think about some of the legacy systems across the board, they all speak a different language. They all have different formats.”

Using AI, fintech has been able to rapidly create meaningful insights from the tangles of inconsistent and unusable data from legacy systems. Yet, this process comes with some potential pitfalls, as Komissarova warns. AI analysis tools, she says ‘are very powerful, but they are non-deterministic. And that leads us to the problem of the hallucination.’

Hallucinations – where AI models deliver nonsensical or false outcomes – can create potential losses for businesses. For example, an Air Canada chatbot promised a customer a fake bereavement discount, resulting in financial and reputational damage.

“Data utilisation is becoming something that used to be nice-to-have, into a must-have,” says Torus’ Lisitsyn.

Processing granular data

It’s estimated one in every five AI queries results in hallucinations. But these mostly come from fragmented data input, rather than flaws in the model.

“If you do not have accurate data under the hood, you increase the likelihood that you will have incorrect results afterwards,” says Komissarova. “Which might cost you a lot of money.”

The experts advocate for a hybrid system. Lisitsyn points out how the balance between a deterministic core AI model can run alongside a professional with the expertise to oversee and interpret the data. Within this hybrid system, the way firms treat data can make a great difference to their overall success.

“How you process data is as important as getting the data itself,” says Oliveira.

Lisitsyn agrees: “It’s not always about getting more data. In most cases, it’s about extracting value from the data you already have.”

Seven in 10 data and analytics leaders report that the most valuable insights are trapped in unstructured data, such as text, images, social media content, website content, audio files, PDFs or PowerPoints. Making sense of all of this requires a combined AI and human effort. Even then, the road ahead could be bumpy.

“It’s still not an easy feat to make sense out of this humongous amount of data,” says Oliveira.

To get started, the experts encourage firms to understand the bigger picture of what the data is telling them. What are the transactions all pointing towards? And how can businesses use this information to boost profitability?

“When it comes to the value of the transaction, it’s much more important to understand the direction of that value, rather than the transaction per se,” says Oliveira. He uses the example of working with an automotive client.

“We’ll be looking at the rates by car type, we’ll be looking at the issuers, we’ll be looking at geography, so we can make the smaller changes that have the biggest impact,” he says.

Once businesses grasp this, they can use these insights to pick the low-hanging opportunities.

“It’s going to be like the Pareto rule, right?” continues Oliveira. “You want to find the 20 per cent change that [brings] 80 per cent impact.”

For Raiffeisen Bank International, constantly enhancing data analysis benefits its clients, too. By providing more insights to merchants, the merchants conduct more transactions through the bank. Raiffeisen goes one further by also offering competitor analysis to help clients make even more informed decisions, which means they use transaction profitability to drive strategy.

“We provided some statistics to the merchant about how the customers behave in their category, and with their competitors on the aggregated level,” explains Komissarova.

This helps to make merchants better at business, which in turn helps both their clients and the bank. It’s a ‘win-win’ she says.

Identifying your customers and your niche, then competing

Establishing the cost of each transaction is everything. But it’s not as simple as it looks. We’ve all been in situations where we felt we overpaid and never returned again. While the business may have enjoyed a small amount of profit on one transaction, it lost a customer and potentially decades of future business. So that single transaction, which may have seemed profitable, never cashed in to the extent it might have. To find the true cost of each transaction, firms must first identify who their customers are, and consider ho they spend across their whole lifecycle.

“Stop thinking about the profitability as the profitability of the product but as the profitability of the client, and you start moving in the right direction,” Komissarova explains. “It’s no longer a case of talking about some basis points on a transaction level. We’re talking about the whole journey,” Oliveira agrees.

Customers who are satisfied with the transactions will come back again and again. And, as Harvard’s famous 2014 breakthrough study found, just a five per cent increase in retention leads to a 25 per cent to 95 per cent boost in profitability. This means that what could seem at first glance to be a slightly less profitable transaction, actually becomes a reliable goldmine over time.

“You need to be competitive, yet profitable at the same time,” explains Lisitsyn. “It’s about finding your right place, your right segment and being able to calculate all the bits and pieces, extract additional data value that’s on top of what you calculate, and then feed this back to your customers in a way that they can absorb it.”

As a business, identifying your lifetime client and unique niche is the sweet spot for transactional profitability. Lisitsyn explains one case study where his clients maximised profits by offering slightly more competitive prices.

“When they started with us, they could see that this specific segment of transactions was the most profitable for them, so they could easily go 50 basis points lower, and still be very profitable, but much more competitive, which would bring in more profitable revenue, more profitable volumes,” he says.

Excellent data analysis is the key to better business outcomes.

“Find your segment and build,” recommends Komissarova. She encourages engineering ‘a good value proposition, which would differentiate you from the others by relying on data technology’.

With great data analysis comes great transactional profit potential.

Once you have that granular data, you can use it to drive much bigger-picture business decisions. It’s no longer a matter of pricing; it’s a matter of using the existing technology to provide you with the means to make better decisions in real time, instead of waiting for those alerts to come up.

This article was published in The Fintech Magazine Issue #38, Page 12-14

The post EXCLUSIVE: “How to REALLY Profit From Every Transaction” – Kirill Lisitsyn, Torus, Breno Oliveira, payabl. and Maria Komissarova, Raiffeisen Bank International in ‘The Fintech Magazine’ appeared first on FF News | Fintech Finance.