This feature originally appeared in the March issue of MReport.
At Fannie Mae, Henry Cason is responsible for the design, development, and launch of a digital suite of products and services across the single-family residential mortgage loan’s life cycle. He also works to foster integration of Fannie Mae’s technology and business infrastructure with that of its customers, striving to help Fannie’s customers grow and create value within the marketplace.
With his tenure at Fannie stretching back 26 years, he has his finger on the pulse of industry innovation, working closely with lenders and servicers in order to try to best align the GSE’s products and services with the needs of the customer base.
Cason spoke exclusively to MReport about why technological advancement is much more effective when paired with process change, the value of test-and-learn, and the transformative power of tying innovation to constant communication.
How can the industry best leverage technology in order to adapt to changes in the marketplace?
You have to introduce process changes to best take advantage of new technology. We’ve always felt that you can’t just do it by implementing technology alone. You have to think about everything holistically. The customers who blend emerging tech with process innovation are going to be able to make those pivots and realize the gains and benefits. It could be increasing efficiencies and lowering the cost of origination. It could be doing more with the same amount of staff.
It starts with asking the question, “What are you trying to solve for?” The lenders that have actually blended emerging technology and process innovation will continue to push the envelope around efficiency and cost savings. For example, you have seen wide shifts in loan volume from purchase to refinancing. As a lender you have to create more capacity, and then you are faced with figuring out what to do with that additional capacity when demand for refinancing fades. Lenders who figure out how to maintain a certain level of capacity and take on new volume efficiently are going to win.
Is there a tendency to focus narrowly on the emergent tech at the expense of reevaluating processes?
You’re absolutely right. You’ve got AI, blockchain. I can go on and on about these emerging technology buzzwords. Before Fannie Mae put customers at the center of its efforts to encourage innovation, we probably would have led with those buzzwords. Now, we don’t even talk about it that way. We sit down with a customer and ask, “What are you trying to do? Let’s understand where you’re headed from a strategic perspective. Let’s understand what you’re trying to do within the context of mortgage origination or servicing. And, then let’s solve for that.”
It’s a mindset shift.
I’m not a firm believer in sitting at a table and building for where you think your target state is. We have seen on multiple occasions that you start out with an idea, and as you get into building and solving for that idea or that problem, you make a lot of pivots. It’s fascinating to see this in action, not only with our customers but with Fannie Mae in general. For example, we have a piece of AI machine learning tech called Ask Poli. We originally went into Ask Poli to automate our guides [for sellers and servicers].
We expect our customers to know how to navigate through that guide, so we originally went with this natural language searchand-machine-learning product to create some efficiencies for them. But, as we started seeing our customers use it, and as the product matured, we were able to see patterns in the data and harness it like we could not previously. With the data garnered from queries made on Ask Poli I can see where I may have some discrepancies in the guide because of the way my customers are posing questions on it. Also, I was able to see the [industry] shift to refi because of the types of questions our customer lenders were asking. That wasn’t our initial vision for Ask Poli, but as we got into it and we started working with it, we saw that the data can be used in a different way.
Which areas of technology are poised for transformative innovation within the next few years?
If you leverage AI and machine learning and use it as a tool to help your human staff, that’s where you’re going to get the most efficiency and gains. Every time I add a certain condition to the validation of a loan, I’m going to add it for every loan, regardless of whether it’s intended for that initial loan or not. If you can apply the right machine learning and AI algorithms, your underwriters can pay attention to the issues they really need to focus on.
We don’t think there’s a world where AI is going to make every decision for you, nor should it be used that way. For example, you could put AI in front of a loan officer and that AI engine could tell the loan officer, “We think this borrower matches these loan products, and here’s why.” The loan officer should still have a conversation with that borrower and confirm that the conclusion is accurate and the loan is appropriate for the borrower.
When we talk about digitization of the mortgage finance process, we’re not talking about how there’s not going to be anybody there physically interacting with customers. Digitization is the intersection of how we use third-party data, combine it with calculations, and then combine that with process change.
In today’s world, you have to be a multidimensional thinker as you’re innovating. You can’t have a project management mindset. You have got to have a product management mindset. I’ve really embraced empowerment and test-and-learn. In the early days of my career at Fannie Mae, I was very “command and control” focused. You were told that this is what you were going to build, without inviting your customers in. We teach our people to take a problem and decompose it into smaller problems, and then solve those smaller problems incrementally.
I think of innovation in the same way. Start small, test and learn, and be willing to fail. I don’t believe you can innovate without failing. If you look at major innovations in other industries, you just didn’t create X. You created X because you failed three or four times, and then you got to X. We try to apply those same problem solving techniques.
Sometimes you don’t know where you’re headed until you actually introduce version one. I’ve learned that you hire good people, you let them think, and you empower them. It’s been transformational how we’ve integrated our customers into our development process.