Despite its dreamy name, a data lake is a high tech affair: a repository for company data from multiple internal and external sources which can require some heavy lifting by a company’s financial team, IT and sometimes consultants and outside vendors to set up.

Once implemented, treasurers and finance teams can use them to extract more up-to-the-minute and reliable data from across the company to better plan and strategize–improving cash management and liquidity and even enabling the company to better track patterns and catch instances of fraud, according to several experts in a FinLync webinar on data lakes last week. 

“There could be countless use cases,” said Chad Wekelo, principal of Actualize Consulting and a panelist, who noted trends in disbursements and invoices would be easier to see and then act upon. “If we’re using checks or wires, can we be smarter about our mix … I don’t think a lot of treasury groups are thinking that way and the main reason they’re not is because they just don’t have access to all that information to put the pieces together. They’re in little silos.” 

While data lakes have been around for a number of years, they have become increasingly important as tech stacks and finance technology has evolved and new data feeds have become available, according to Mitch Thomas, a panelist and head of solution engineering-North America at FinLync, a fintech company which serves corporate finance and treasury teams.

“Data lake is the newest term in a line of buzzwords meant to indicate centralization of data,” Thomas wrote in an email. “Complicated data extractions and combinations of manual and automated reporting is what is being replaced … To centralize the data across a company and access real-time bank data so they can be analyzed together has been impossible until now.” 

Data lakes, which typically hold unstructured, structured and semi-structured data such as social media feeds and data from bank statements, are slightly different from data warehouses, which hold structured data such as relational databases with tables, columns and records that are used to manage data for business intelligence (BI) and analysis, he said.  

Probing the unknown

“Data lakes are generally used to probe data for deeper insights and for answering questions that are not known in advance,” Thomas wrote, adding that data lakes typically have lower hardware and implementation costs than warehouses but can be more expensive to extract information from. 

Budgeting is a major challenge that treasurers must overcome to set up such systems. Nearly 90% of the respondents polled during the webinar cited lack of IT resources or allocation as the biggest hurdle to getting started with a data repository, followed by 14% who cited understanding the benefits to treasury.  

Panelists said the the cost of such a system varies greatly but declined to estimate the average cost of implementing such a system. Trellance, a data analytics provider for credit unions, estimated that the cost of building a data warehouse or data lake for an analytics platform at an individual credit union can range from around $500,000 to well over $1 million. But some data experts say the cost varies greatly and can be much less. 

The software and storage technology tools themselves have elastic pricing depending on the size and volume that companies need and competition has also brought prices down, panelists said. The cost of building and maintaining are other cost components which differ. For example, if a company has the expertise internally to be able to build the interfaces and the exchanges into the data lake that will cut the price. But the panelists said companies can reduce the cost by starting with a small project and then using that to make a use case for larger investments. 

“There is a sort of crawl-walk-run approach to this,” Wekelo said.“It doesn’t have to start fast. It can be looked at in phases and steps, almost like a prototype thing. That’s what we see with a lot of clients.”

Keeping IT spending in check is one of the many challenges confronting CFOs and financial executives as they fight record inflation while also seeking to retain a competitive edge in technology. 

Still, the companies that aren’t leveraging their data over time are going to suffer because the quantity of data that's being used is growing, one panelist said. “In a sense there’s a cost to not using the data and that cost is only going to grow,” Rob Granger, senior manager of Actualize Consulting, which has worked with treasury departments on data lake projects, said in the webinar. “As data becomes more and more prevalent in daily lives, leveraging that is going to become more of a differentiator than it is now.”