NewsLink Q&A: Rocktop Technologies’ Ashley Andrews on How Servicers Can Reduce Operating Costs
According to MBA’s most recent Servicing Operations Study and Forum, servicers’ direct costs have been trending upward. We asked long-time servicing professional Ashley Andrews, Head of Transformation with Rocktop Technologies, how to lower expenses and improve margins.
MBA NewsLink: Servicers have been working to reduce operational costs for years, but our research shows the opposite is happening. Why do you believe the numbers are still going up?

Ashley Andrews: Most cost initiatives focus on the most visible components of total spending that are easiest to see. MBA’s own research puts direct servicing costs at $185 per loan in 2025, up from $181 the year before. About three-quarters of that, roughly $136, is non-default work that touches every loan every month.
When servicers set out to reduce these costs, they usually go after headcount within a function or rates within a vendor contract. What they don’t always address is the work that happens between systems, which is where some of the biggest cost-reduction opportunities lie.
I’ve learned throughout my career that a servicing shop doesn’t run on just one system. It relies on multiple systems, including a system of record, a document repository, and other layers supporting default and investor reporting. And where those systems don’t connect, teams create spreadsheets and manual processes to bridge the gaps. The result is that people become the integration layer. Instead of focusing on their borrowers and the work that really requires their attention, they spend their days rekeying information, reconciling data, and checking one system against another. That kind of work often has no line item and no owner, which is precisely why it survives every cost-reduction initiative even though it continues to drive the underlying cost structure higher.
MBA NewsLink: You have said that these are not the only overlooked costs. What do you mean?
Ashley Andrews: There is also operational leakage that occurs when the cost of manually identifying, preventing, and recovering a loss outweighs the economic value of doing so.
I see this most often in three major areas:
- Recoverable advances that aren’t fully recovered because identifying, validating, and pursuing them within the required timelines is too manual;
- Payments and invoices that aren’t systematically validated against the underlying servicing activity before the money goes out; and
- Exceptions, risks, and opportunities buried in unstructured servicing data, particularly servicing comments, because there hasn’t been a practical way to analyze millions of free-text entries at scale.
These costs don’t always show up as discrete line items, but they directly affect the economics of servicing a loan and the value of a portfolio.
MBA NewsLink: Transfers and portfolio boarding come up in almost every cost conversation right now. Why are they also such expensive events for servicers?
Ashley Andrews: A transfer is a cost event long before it is a revenue event, and a significant portion of that cost comes from what servicers discover only after they receive the portfolio.
Let me give you a real-world example. On a recent bulk transfer, we took in a population of roughly 15,000 loans with more than 14 million pages. Within that population, we found over 63,000 duplicate files, manifests that didn’t match the contents of the folders, corrupt source files, and documents that didn’t reconcile with the loan data provided at the time of sale. And this wasn’t unusual. Anyone who has boarded portfolios at scale has encountered some combination of these issues.
When servicers board a portfolio, the real question isn’t whether they received the files. It’s how long it takes before they can trust them. Until then, they’re making advances, managing timelines, and executing servicing decisions based on a portfolio they can’t fully see. That gap creates real cost and risk, and it isn’t adequately accounted for when the transaction is priced.
MBA NewsLink: Once servicers become aware of the scope of these issues, how can they address them?
Ashley Andrews: Fundamentally, servicers need a more streamlined way to access, validate, and act on the information within their portfolios. That’s the common challenge across all the examples I’ve shared, and it becomes even more important as delinquency and default volumes increase. Your organization’s most recent National Delinquency Survey shows that at the end of Q2 2026, the delinquency rate was up 44 basis points year over year. FHA delinquency rates increased by 122 basis points and VA delinquency rates rose by 57 basis points.
Servicers are on the front lines when the market changes. The operational challenge when that occurs is how to absorb increasing volume without adding proportional headcount, while ensuring every action is based on complete, accurate, and audit-ready data. You can’t simply ask people to work harder as volume increases. You must remove the work that shouldn’t require a person in the first place.
MBA NewsLink: What does this strategy look like at the implementation level? Does it involve purchasing a new system of record?
Ashley Andrews:Actually, this is the opposite of re-platforming. In most cases, the data and information a servicer needs already exists somewhere within the servicing ecosystem. The problem is that it isn’t connected in a way that allows servicers to validate it, make decisions from it, and use it to drive automated processes.
The good news is that you don’t have to rebuild the servicing operation to solve that problem. You can access the information while keeping the platforms that already run your servicing shop. And in many cases, implementation can be done in days or weeks rather than years. I think of it as an intelligence layer that unifies information without replacing the disparate systems underneath it.
As servicers weigh whether to build or buy these solutions, I would argue that the hardest part of implementation isn’t the technology. It’s making sure the people behind it know what a servicing document is supposed to say, what the servicing history is actually telling them, and how to distinguish the exceptions that require action from those that are simply noise.
Beyond implementation, another question I would ask is how quickly the solution can start delivering measurable value. Just as important, does it show its work? An analyst should be able to understand how a conclusion was reached, see the information behind it, and override it when appropriate. To me, that matters more than any accuracy percentage someone quotes. The goal isn’t to remove human judgment. It’s to give people better, more consistent information so they can focus their judgment and expertise on the decisions that actually require it.
(Views expressed in this article do not necessarily reflect policies of the Mortgage Bankers Association, nor do they connote an MBA endorsement of a specific company, product or service. MBA NewsLink welcomes submissions from member firms. Inquiries can be sent to Editor Michael Tucker or Editorial Manager Anneliese Mahoney.)
