Anyone Can Build It. Who’s Going to Keep It Running?

Jim Paolino is CEO of mortgage closing cost and fee data provider LodeStar Software Solutions, a four-time Inc. 5000 honoree.

Necessity may be the mother of invention, but in mortgage, scarcity is an equally prolific parent. Every market downturn triggers a cycle that eventually produces a fresh crop of technology entrants eager to inject efficiency into the industry’s admittedly antiquated processes, and the current cycle is no exception.

The gap in per-loan production expense between the top and bottom quintiles of IMBs has nearly tripled since 2020, peaking at almost $5,000 in 2023. Volume covered that gap for years. Now that margins have compressed, lenders will look at vendors they would have ignored before.

What’s different about this cycle is that the traditional barriers to entry for technology developers have been obliterated by AI. Building technology for this industry used to require significant capital investment and development time, which is why the biggest waves of entrants arrived during the recovery rather than in the midst of the downturn. With AI, anyone can become a developer and stand up something that demos well in a matter of hours.

So the question lenders need to ask today is not whether these new entrants can build something. It’s whether they’ll still be here to support what they’ve built.

Beware the carnage

Whether it’s a lender building an LOS, an LOS building a pricing engine or industry outsiders looking to build a better mousetrap, mortgage tech development is a land grab right now. Much like land, however, there are only so many mortgage tech buyers to go around.

In reviewing 431 venture-backed company shutdowns since 2023, CB Insights found that fintechs accounted for the second-highest number of failures and were the thinnest capitalized, with a median of $4 million raised compared with $11 million across all sectors. Running out of money was cited as the primary reason for closure in 70% of all cases, not just for fintechs, but this was often the last thing to go wrong, not the first. The overall root causes of failures were poor product/market fit (43%), bad timing (29%) and unsustainable unit economics (19%).

What didn’t appear on this list was an inability to build the product. That is the easier part, and AI has only made writing functional software faster and cheaper than ever. If a product can be built in an afternoon, it can be copied just as quickly, so writing code quickly is not a moat. What determines whether a tech business lasts is servicing customers, selling into a regulated industry and finding a market willing to pay, and those are as hard as they have ever been. You can’t vibe code a support team.

The stakes tied to getting it wrong are also much higher for mortgage tech startups. If a consumer software company ships a product with a flawed feature, it can simply issue a patch. In mortgage tech, nearly every output carries a regulatory consequence. That’s why getting a product into production in this industry takes months of configuration, testing and compliance review.

This is the real barrier to entry for mortgage tech. The difficulty of getting a product into production, keeping it compliant through successive rule changes and supporting it through a full rate cycle remains the same, and the pool of buyers is shrinking, as lenders themselves continue to consolidate.

No free toasters

Regardless of how a business is funded, it is accountable to someone or something during a downturn. For a bootstrapped company, it’s the bottom line. If it isn’t making money, that money comes from somewhere, often an owner’s own pocket. That is accountability to the nth degree, because otherwise you’re out of business. The upside here is control. As long as the lights stay on, nobody outside the company can force a change in direction.

Venture capital and private equity firms bring more resources, money and connections that enable their companies to grow much faster. The trade is that priorities are set elsewhere. A company designed to grow without regard to profitability can find itself expected to produce it the moment the market turns, which is not what it was built to do.

What matters more than who owns a vendor is how central its product is. Acquisition is often the end goal for many startups, and when a company is acquired, the question is whether it becomes the main revenue driver for the buyer or the free toaster the buyer gives away in service of its main offering.

Not all acquisitions carry the same risk. For example, a point-of-sale (POS) provider buying another POS provider is buying market share, and the acquired product is usually sunsetted while its clients are migrated onto the buyer’s platform. While disruptive, serving the acquired client roster is the whole point of the deal. A provider buying a complementary tool in an adjacent category is buying expansion, and that product generally survives with investment behind it. The quietest risk is the third case, when a large acquirer takes on a portfolio of assets to get the one piece it actually wanted. Everything else arrives without a champion, and what nobody is responsible for growing tends to die of neglect.

When conditions tighten, the acquirer cuts costs, whether under a private equity mandate or an independent company’s cash shortage, and the nice-to-haves go by the wayside. Nice-to-haves don’t survive market downturns.

None of this is a reason to avoid a vendor with an exit in its future, but lenders should ask about the horizon before signing and let the answer shape how deeply a lender integrates, what it negotiates for and what it would do if the product it depends on became somebody else’s nice-to-have.

Count the salespeople

Lenders tend to evaluate vendors the way they run their own businesses. Mortgage companies don’t deal in long-term contracts or guaranteed minimums. Instead, they compete on rate, so the instinct is to drive costs down. Applied to a technology partner, that lens measures the wrong thing. What a vendor costs matters less than what it saves, and neither number says anything about whether the company will still be there in three years.

The companies that last focus on operations before sales, and that is visible from the outside. A company under pressure to show top-line growth starts selling just to acquire logos. Eventually, you meet a sales rep who has been there two years and has never seen one of his signed clients go live.

Growth misleads in the same way. A company doubling its revenue while the market grows nearly as fast is being carried, not built. Growth is only secure when revenue expands faster than headcount because that is what provides a buffer when things turn.

Though none of this shows up in a demo, all of it can be asked about. A lender lives with a technology decision for years, whereas a vendor earns its longevity one retained client at a time. Both sides do better when the questions get asked early.

(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.)