Baird is a full-service financial firm with more than US$200 billion in assets under management and operations across the United States, Europe, and Asia. Despite the company's global reach, its three-person treasury team also handles accounts receivable. That's a critical function for the client-focused business. Unfortunately, in 2016, receivables processes were largely manual, and staff sometimes had trouble keeping up.

Every day, customers send hundreds of ACH payments and wire transfers to fund their Baird investment accounts. “When these payments come in, we need to post those funds to their accounts as soon as possible,” explains Steve Huse, director of client reporting at Baird. “We work in a real-time environment, with wires coming in throughout the day. These payments are very time-sensitive—they might meet a margin call or fund an imminent purchase. Our financial advisers might be waiting for the funds to hit the customer's account so they can put that money to work in the market.”

To keep up, treasury staff would log onto their banks' portals repeatedly throughout the day and pull down BAI files containing information about incoming wires. “We would ask for payment data in a standard format,” Huse says. “Our institutional clients were typically pretty good about complying, but our retail clients would send information in all kinds of formats. The team would get bogged down with trying to figure out how to properly apply payments to customer accounts and book them in the general ledger.” The lean team was spending about half of a full-time equivalent (FTE) position on ensuring incoming electronic payments were booked properly.

It's imperative that Baird book receivables both quickly and accurately. “Through our online platform, customers see funds hit their account same-day,” Huse says. “If something was keyed incorrectly, that mistake would be very visible. There was the potential for a six- or seven-figure error because someone had fat fingers. And if we had to kick funds back to a client because we didn't get proper information to apply the payment, that also would be a client-service nightmare. Our business is predicated on excellent customer service, so any issues around applying funds would create a significant reputational risk for Baird.”

When the company decided to deploy a new treasury management system, automatic posting of electronic payments was front and center in the list of requirements. “We saw that bringing that workflow into the workstation could remove some of those points of risk,” Huse says. The company wanted to go with a software-as-a-service (SaaS) treasury system, both to reduce the burden on the internal IT team and to streamline upgrades to the system. “As a company, we've made a concerted effort to move toward SaaS-based solutions whenever those alternatives are available,” Huse says.

The challenge was that Baird's desire for an automated receivables-posting process would require customization, something many SaaS vendors shy away from. The treasury team launched the project by collaborating with internal IT to determine how the posting process should be configured. “Our treasury analysts worked with a system architect to whiteboard our preliminary thoughts on how the system should work,” Huse says. “We invested a tremendous amount of time in determining what was and wasn't feasible.” They ultimately came up with the idea of using an application programming interface (API) to connect the new treasury workstation to both the company's banking portals and its internal customer database, for reconciling account information.

Baird didn't find any SaaS-based treasury systems that provided these types of connections. However, a thorough due diligence process revealed that GTreasury was interested in accommodating the company's needs. The vendor provided a proof of concept to demonstrate how the system could work, and Baird decided to move forward.

GTreasury heavily customized the automated receivables-booking processes in its system to accommodate the unpredictability of information surrounding Baird's incoming payments. Now, when Baird receives a BAI file from a bank, the system uses a heavily modified transaction assignment engine to parse through the file to find each payment's associated Baird customer account number. “It starts by looking for an identifier like 'account' or 'reference' followed by the appropriate number of digits,” Huse says. “I think this type of functionality is fairly common.”

What's more unique is that when no identifier is clear, the customized workflows enable the system to continue searching for the customer account number. “A BAI string might contain hundreds of number sequences that could possibly be a match,” Huse continues. “An account number search might run right into phone numbers or other strings, and there's no easy way for the system to determine which digits make up the customer account number.”

When the system encounters such a conundrum, it queries a database containing current Baird customer information. “The engine will send all possible account numbers through an API call to the database,” Huse says. “Assuming there's a match, it'll send a message back to GTreasury with the portion of the string that is an accurate account number. It can also do a secondary match with other information from the account record, such as last name.” When the rules engine is confident it has the correct customer account, GTreasury automatically books the funds to the right account in the general ledger (G/L).

See also:

Instant Insights /

APIs Automate Booking of Payments that Defy Standardization

Related Stories

Recommended For You