Are vertical fintech models the next step beyond generalist lending?

These days it’s pretty hard to find a fintech niche that hasn’t already been chewed through and fought over by a dozen well-funded players. Most mature segments now end up looking roughly the same:

• A couple of legacy giants who’ve accumulated so much capital, brand, and operational fat over the decades and even centuries that they somehow keep rolling forward even while losing market share year after year.

• A handful of serious challengers armed with venture cash, clever product managers, and a deep hatred of slow decision-making. They usually carve out their piece of the market by doing the obvious things: innovating faster and burning money with higher efficiency (sometimes).

• And finally, swarms of followers, each offering the same five features and hoping that maybe, just maybe, this is the month they “go viral”. Every now and then one of them does rise to the top, mostly through talent, grit, and a sizeable portion of luck.

So how do you actually blaze a trail in this red ocean? How do you find a niche inside a mature segment like lending, where everyone from banks to BNPL apps to your neighbour’s startup seems to be offering credit?

One answer (maybe not the only answer, but at least a promising one) sits at the intersection of very specific customer segments and very deliberate technological decisions.

Why picking a tiny niche sometimes works better than trying to serve everyone

Take SMB lending. On the surface, it looks like a commodity business. Same loan products, similar underwriting, endless spreadsheets. But the companies that manage to stand out usually do one thing differently: pick a niche and go deep.

The logic is simple: if you know one narrow customer segment better than anyone else, you can build tools and risk models that are able to allocate larger limits, provide lower rates, and conduct underwriting quicker. And suddenly your “tiny niche” becomes a moat.

But this only works if your tech isn’t just good – it has to be weirdly good for the specific vertical you’re after.

The tech stack that makes niche lending possible

1. An underwriting model that actually understands the business

You can’t lend to restaurants, or dentists, or micro-SaaS founders using the same risk model. A vertical lender needs a model that swallows the combination of data sources which are the most abundant and credible in the industry you are aiming at:

• financial accounts

• reconstructed reports from bank feeds (open banking helps)

• trade data with suppliers and buyers

• POS behaviour

• behavioural, operational and other data

Then you mash it all together into one risk engine that spits out a result in a few days for big clients, or even minutes for smaller ones.

A good example is Cherry, a BNPL player that focuses purely on healthcare, wellness, and aesthetic procedures. They know the industry, so their underwriting makes sense for that industry.

2. Having access to data your competitors don’t

A lot of lenders plug into big corporate databases (D&B, etc.), but those tend to work better for large companies and jurisdictions rich in public records. Everyone else ends up with a ton of blind spots.

So the smartest niche players get creative:

• tap into trade datasets

• integrate with local registries (here you need to work hard - from reconciling local variations of accounting standards to aspects of regulation and reporting)

• or build their own data-producing machines

Some of the strongest players are actually software businesses first, lenders second.

For example:

• payment platforms like PayPal know more about a merchant than a credit bureau ever will

• CFO automation tools like Mimo can reconstruct financial health in real time

• Integrated restaurant platforms like Toast have an almost unfair advantage: they literally see every customer payment at the moment it happens

3. Building collection mechanisms into the business workflow

If you can design your repayment mechanism so it sits right where the customer makes money, underwriting becomes much easier.

Toast is a textbook case: they offer loans to restaurants, but repayments happen automatically every time a diner pays through the POS. Essentially, the business pays the loan back in tiny bites. The loans themselves are from a partner bank, but Toast controls the data and the rails.

This reduces risk so much that you don’t need to be hyper-strict with underwriting – you just need to know roughly how many pies, tacos, or cocktails this restaurant sells each week.

When niche lending becomes... a consumer app?

And then there are players that go full galaxy-brain. Take Shuffle, a young startup in London. They lend money to restaurants, but (plot twist) they also touch the consumer side.

How?

When diners pay at participating restaurants, Shuffle quietly takes a small portion of each payment directly towards the restaurant’s loan.

And to attract consumers, they throw in discounts for future meals – but only in restaurants that are Shuffle customers.

It’s a closed ecosystem: restaurant gets financing → diners get discounts → repayments happen invisibly → Shuffle gets data + loyalty + risk control

A bit goofy, a bit genius.

So where does this all lead?

What’s interesting is that some of these niche plays – Toast, Cherry, others – grew from a tiny idea inside a huge market into businesses managing billions in transaction volume.

Their trick is always the same:

• focus on a narrowly defined customer

• build a tailored risk model

• plug into the customer’s workflow so tightly that repayment becomes automatic

• and collect data that competitors simply cannot access

In a mature market like lending, being a generalist without proper scale is almost a guarantee of mediocrity.

But being oddly specific? That might just be the next big fintech trend.

Even if it looks a bit goofy from the outside.