The balance was $120. Ohad Samet could afford it. What he had failed to do was remember it: a small amount left on a store credit card, followed by calls from unfamiliar numbers and an unpleasant encounter with a collector. For someone who had worked in financial risk, this was an unusually personal introduction to the other end of the credit business.
There is a useful distinction here. A person can have the money and still dislike the process of paying. A person can want to pay and lack the money. A collection script that treats those situations alike starts with a handicap. TrueAccord, founded in 2013, grew from Samet’s conviction that the experience itself needed redesigning.
- Make contact count. Personalize the timing, channel, message, and payment offer.
- Give people room. Let consumers explore repayment online, on their own schedule.
- Keep the boundary clear. Better engagement still depends on lawful accounts and affordable payments.
The decision before the message
TrueAccord is a debt collection agency with a software engine inside it. HeartBeat, its patented decision system, uses machine learning and data-driven rules to choose how to approach an individual account. When should a message arrive? Through which channel? What should it say? Which offer belongs in front of this person?
Those questions sound like marketing questions. That is part of the point. Businesses spend considerable effort making purchases easy, then can make repayment feel like an interrogation. TrueAccord applies experimentation to that neglected stretch of the customer relationship. Its published account of HeartBeat describes pre-approved compliance boundaries, historical data, and tests that refine the approach.
One experiment is wonderfully specific. In a July 2025 post, the company described a client whose debt-notification emails needed improvement. The team tried animated GIFs and reported a 6% increase in click rates. A GIF is an unlikely guest in a collections department. Yet the experiment illustrates the method: change something observable, test it, inspect the result. A click, of course, is an invitation accepted. It is not a debt repaid.
Two customers, one uncomfortable moment
The paying client is the creditor. The person using the payment journey is the consumer. TrueAccord has to make that arrangement work for both. Its market includes financial institutions, fintech lenders, buy-now-pay-later providers, telecommunications businesses, and credit unions. Yelp and LendUp appeared among named clients in its 2017 funding announcement.
For creditors, the offer is outsourced recovery with individualized outreach and less dependence on a queue of outbound calls. Its third-party service works on contingency: compensation depends on collection. For consumers, the practical attraction is a mobile-friendly portal, available around the clock, where payment plans can be arranged without a conversation.
The company’s consumer FAQ also describes account information, hardship options, and disputes. A consumer who needs an explanation can still seek human help. Available payments and arrangements depend on the account. TrueAccord says it does not buy debts or own the balances it collects; it acts for clients who place accounts with it.
That puts it alongside established collection agencies and digital competitors such as January and InDebted. Its distinction is the combination of collection operations, behavioral decisioning, and self-service resolution. The useful comparison is the complete service: who handles the account, which channels reach the consumer, and what happens when automation needs help.
First prove it. Then broaden it.
Why operate a collection agency at all? Why not simply sell software? In a Receivables Podcast interview, Samet described investors encouraging the software route. But an unfamiliar approach needed evidence from actual collections. Operating the service let the company demonstrate what its technology could do.
“We are educating a market.”
Ohad Samet / Receivables Podcast
Samet described a turning point when lenders asked to use the technology under their own brands. Today, the corporate distinction matters: TrueAccord sits within TrueML Technologies, while sister company TrueML Products offers Retain, a software product for managing delinquent customer communications. The agency and the software business solve related problems with different operating responsibilities.
Building the agency required capital. A $22 million Series B, announced in November 2017 and led by Arbor Ventures, was earmarked for product development, compliance capabilities, expansion, and hiring. It is a reminder that an automated service still needs people to build systems, onboard clients, and keep the machinery accountable.
The company’s published values make room for this tension. They call for measurable outcomes, empathy, and calculated risks, including a willingness to learn from mistakes. In 2025, TrueAccord was selected for the Best Places to Work in Collections program, which includes employee surveys. Those are useful signals about its stated culture, rather than proof of every consumer’s experience.
Consumers served since 2013, as reported by TrueAccord in May 2025. A cumulative reach figure, not a count of completed repayments.
In May 2025, TrueAccord acquired Sentry Credit, an agency with more than 30 years of history. The deal brought clients and first-party collection and litigation services. In February 2026, TrueAccord announced dedicated first-party servicing through Sentry, combining agents with AI support. Teams can work inside a client’s systems or use the provider’s CRM. Digital collection had acquired a broader human apparatus.
The account has to belong here
A pleasant interface cannot settle whether a debt is lawful to collect. In June 2022, Connecticut issued a consent order involving TrueAccord’s collection of certain loans from unlicensed lenders, alongside trust-account and marketing allegations. TrueAccord neither admitted nor denied the allegations. It agreed to a $10,000 civil penalty, borrower refunds, and specified changes to its practices.
That episode belongs beside the technology. Good wording cannot repair an ineligible account. Nor can a well-timed message supply money that a household does not have. Digital access, accurate information, affordability, and legal eligibility remain conditions of the model. TrueAccord’s own terms also limit its services to the United States.
For a creditor evaluating the service, that suggests a more demanding scorecard than response rates alone. Compare recoveries after fees, the handling of disputed accounts, and the ease of reaching a person when the portal is insufficient. Ask whether a test covers comparable accounts and comparable time periods. A successful message experiment can justify another experiment; it cannot establish the economics of an entire portfolio. The distinction matters whenever a pleasing percentage is doing the selling.
The lesson a reader can borrow is modest enough to test tomorrow: examine the step immediately after asking for payment. Can someone understand the balance, question it, and choose a workable next action? Measure what happens beyond the click. TrueAccord’s founding insight retains its force because the original problem was so small. The money was there. The cooperation had to be earned.