Transcript

0:00 — The number you already report

If you run consumer lending at a bank or credit union, you know your application abandonment rate. If you run marketing, it might have been handed to you as a problem to solve.

That abandonment number is real. It's also likely the smaller half, because for every borrower who starts an application and quits, there are more who never started at all. They don't show up anywhere.

And both groups have the same thing in common: they hadn't decided anything yet. And there is likely nothing on your website that was going to help them. You see, there's a gap between getting interested and being ready to apply. And on most sites, it's empty.

I call it the decision readiness gap.

That gap is what the next twenty minutes or so are about. By the end, I'll give you five questions you can use to find how wide your gap is.

I'm Bill Pollock, president and founder of Fintactix. I've spent a little over two decades building the software people use to make financial decisions. Long enough to have a pretty good view of what those tools do well and where they stop.

Let me start with what's already working because I think it matters.

Most institutions have spent real money on digital lending in the last five years. Origination platforms, digital account opening, decisioning.

And those platforms have largely delivered. Applications are shorter, save and return works, documentation is deferred until it's needed.

That's serious software and business engineering against a real problem.

But here's what happens with a platform investment of that size. Once it's in, the conversation shifts. It stops being about whether the technology works, and it starts being about whether it's returning on its investment.

And what an origination platform earns back depends on throughput. It converts borrowers who show up ready to apply for a loan or new account. That's the job. That's what it's excellent at.

So let's look at both halves.

Start with the ones who do reach your application. Some of them click the Get Started button because it's the only thing on your site that might give them real numbers. They're investigating.

And partway through, they get asked for a VIN or a pay stub, and they stop.

They came to get answers and information, not to commit.

That abandonment shows up in your platform numbers, and it gets read as a platform problem. But it isn't one.

The other half never gets that far. They read your loan pages, checked your rates, and left without identifying themselves. No form, no submission, nothing to count.

Your reports can't tell them apart from somebody who landed on the wrong search result.

One group you can count, the other you can't, and the one you can't is bigger.

So that's where we'll spend the next twenty minutes or so, where these groups come from, why a shorter form doesn't reach them, and neither does a better one, and what it takes to have more decision-ready people start an application.

You're not going to be asked to buy anything.

3:18 — Two types of abandonment

Loan application abandonment is really two problems that share the same name. Borrowers who start and don't finish, and borrowers who never start at all.

The difference isn't about causes. It's that one of those numbers shows up in your report and the other one doesn't.

Mid-application abandonment is measurable. You know who started, you know who stopped, and most origination platforms will show you the step where it happened. There's a record.

Pre-application abandonment leaves nothing behind. No form, no submission, no identity. In your analytics, it's just a page view that went nowhere. It looks exactly like someone who landed on the wrong search result and left.

So one of these problems shows up in a report every month, the other one never shows up at all.

And there's nothing there to measure. Somebody reads your home equity page, can't tell a HELOC from a fixed-rate loan, and closes the tab.

You get nothing.

So why do they leave? Seven reasons come up consistently.

• They don't know which product fits their situation.
• They can't tell whether they qualify.
• They're overwhelmed by the number of variables.
• They're comparison shopping and got pulled to a competitor.
• The application assumes a decision they haven't made.
• They wanted to talk to a person and couldn't find an easy way to do that.
• Or they simply got distracted. A phone call, a meeting, life.

Now watch where these get measured.

The distraction issues are mostly a mid-application problem, and that's what save and return exists for.

Most of the rest are pre-application problems. They exist before anyone touches your origination platform.

And one of them sits right in the middle.

The application assumes the borrower has made a decision they haven't made yet.

That one causes mid-application abandonment. It shows up in your platform numbers, but the cause is upstream. That's the gap I mentioned earlier.

There's one more thing here.

Four of those are different shades of the same problem. The product confusion, the qualifying question, they're overwhelmed, and the application that assumes a decision.

Every one of those is a borrower who hasn't figured out what they're doing yet. They ran into confusion, not friction.

This also explains where the industry puts its money.

Nearly all of it went into the application itself for a straightforward reason: that's where the data was. You can't build a business case against a number you can't produce.

So the smaller visible problem got the attention and the infrastructure. The larger one got nothing because it never shows up on a report for anyone to argue about.

And it just might be the bigger of the two. Most of the reasons on our list happen before anyone opens an application, so most of the leaving happens when nobody is watching.

In reality, the pool of potential borrowers is far bigger than the one you're tracking. They came with a real borrowing interest, looked at what you offer, and left. They showed up as a page view, not a lead, and that's how they got treated.

Nothing in your origination stack is built to reach them, and it shouldn't have to be. It starts when the borrower is ready to apply, and it's excellent at that.

The question is what happens before?

7:03 — What's on the page today

I want to spend a few minutes on the abandonment reason in the middle, the borrower who starts an application just to get answers to the questions they had when they clicked Get Started.

Let's call it a false start. The application assumes the borrower made a decision they hadn't made yet.

I think it's the most underestimated thing happening on bank and credit union websites right now.

Some of the people opening your application aren't applying, they're doing research, but the application is the only resource they think might help.

Put yourself on your home equity page for a second. Think of yourself as the potential borrower. You're carrying about twenty-two thousand across three credit cards, the rates are north of twenty percent, and you're wondering whether tapping your home equity is the smart move.

What's actually available to you there?

A rate table? Maybe, but static. It doesn't know anything about you. Product descriptions that explain what a HELOC is and what a fixed-rate loan is, but not which one is better for you. Maybe a calculator which gives you a payment if you already know the amount and the term, a phone number.

And a button that says get started.

The application is the only thing on that page that asks about you, the borrower, your income, your balances, your property, what you're trying to do. It's the only thing on the site that takes your real situation and gives you something specific back.

So people use it.

It's the one thing on the page that meets them where they are.

Your calculator ends up in the same place, by the way.

Somebody uses it, gets a payment in ninety seconds, and the number is correct, but now they've got a question the number can't answer. Is this even the right product for me?

And there's nothing between the calculator and the application, so the application picks up that traffic too. Same question, but a different path in.

Two things happen to these borrowers, and only one of them is visible to you.

The first is that they stop. They get a dozen fields in, and it asks for a pay stub or a payoff statement or two years of returns. That's the moment that stops feeling like research. They came to find out something. Now they're being asked to commit.

So they leave, and that abandonment lands in your platform's numbers.

Which means it gets read as friction. Somebody looks at the drop-off, sees people quitting at the documentation step, and decides that step needs work.

It might.

But this borrower didn't quit because it was hard. They were never going to finish. Somewhere around field twelve, they got what they came for or worked out they couldn't get it here.

The second thing is that they finish. And that costs you more than you think. Now there's a live application from somebody who was still deciding, and a loan officer must run the whole product conversation from scratch.

Sometimes that ends with a good loan. Often it ends with them saying they're not ready or finding out they wanted something else.

Either way, you've spent loan officer time, and you've got an application that must be dispositioned.

So the same root cause produces two different bills. One arrives as an abandonment statistic that gets read as a platform problem. The other arrives as staff time and pipeline noise that never gets attributed to anything at all.

And neither one gets fixed inside the application. You can shorten it, make it beautiful. An undecided borrower is still undecided when they get there.

They needed help making the decision, and they needed it earlier, before the application, not inside it.

11:03 — Decision readiness

So what would reach these borrowers?

Let's define the objective first. A borrower who's decided knows four things:

• Their own situation in numbers.
• Which of your products fit them and why.
• What they're giving up by picking one over another.
• And what happens next.

Somebody who knows those four things is ready for your application. Everyone else is still working it out, and no amount of engineering on that application changes that.

That's what I meant at the start by decision readiness. These four things are the definition. That's a decision-ready borrower.

Decision readiness is about the borrower's mindset, not yours. It says nothing about whether you'd approve them. It says they've worked out what they want, and they're ready to come ask you for it.

And it isn't a score you put on a person, it's a description of where your website got them to.

So then the question gets practical. What must happen on your site to move somebody from not knowing those four things to knowing them?

The first is asking the right questions in the right order.

Every institution asks borrowers questions. The application asks plenty. What's different is what those questions are, who those questions are for, and when they show up.

Your application asks this question as if the decision is already made. It needs the details so it can process the loan.

Guided questions do the opposite job. They help somebody work out what they want in the first place, so they start in a different place.

Your product page asks people to pick. Here's the HELOC, here's the fixed-rate loan, now choose one. Guided questions start by asking about the borrower.

What are you trying to do? How much do you owe? What's it costing you? What are you hoping to achieve?

The rule is simple: identify the situation before the product. Ask somebody to pick before they understand what they're picking between, and you're just asking them to guess. Most people won't guess on something like this. They leave. That's the first reason on our list.

So once we ask the right questions in the right order, the next piece is what to do with the answers. They must run against your real numbers.

This is where a lot of decision support tools fail. It's the difference between something useful and something that spits out a plausible number nobody can act on.

A generic estimate uses generic assumptions, national average rates, standard terms, typical fees. You get a number that's mathematically right, but wrong for the person reading it.

Running real inputs against real rules means their numbers meet your products. What they're trying to do and what their finances look like on one side, your terms, your rate structure, and minimums and maximums on the other. What comes back isn't a national average, it's what you would offer them.

So once you've run their answers against your real numbers, the next piece is what comes back to them. A short list with the options that match their situation.

Show a borrower every product you offer, and they're back where they started.

That's the product page again, and the product page is what sent them to your application. Show them two or three that fit their situation, and they're somewhere completely different. The narrowing is the work.

But the narrowing is only half of it. Once somebody is looking at two or three options, the question changes. They'll need to understand how those options differ and what that means for them.

Go back to the HELOC and the fixed-rate loan, the actual trade-offs. One has a lower payment to start, but the rate can move. But by how much? The other costs more every month and never changes. But how much more? And given what they just told you, here's which way that actually cuts.

The ability to compare options is the part that tends to get engineered out. Without it, you've handed somebody a recommendation and asked them to trust it. Few people extend that kind of trust to a website on a decision this size. They'll go get a second opinion, and that usually means getting it somewhere else.

For the borrower, the explanation is the product. It's the thing they came for and couldn't find. It's also what turns somebody who got a suggestion into somebody who made a decision.

So once the borrower has a short list they understand, the last piece is the handoff. What you get is somebody who has made a decision, and what they get is an application with no surprises in it.

And if the systems talk to each other, what they already entered carries forward. They're not answering the same six questions twice. That one comes down to how things are wired, so ask it of any solution you're evaluating, because a borrower who must start over has just been handed a reason to reconsider at exactly the wrong moment.

One more thing, and it's the important one.

None of this is an approval. Nothing I've described underwrites anybody, makes a credit decision, or qualifies the borrower for an offer. It doesn't pull credit. It doesn't commit your institution to anything.

What comes out the other end is a decision made by the borrower. They understand their situation, they know which of your products fits it, and they're ready to formally ask.

Everything after that point is your origination platform's job, and it's very good at that job. This just changes who's standing in front of it.

Okay, the four pieces again.

• Ask the right questions.
• Run the answers against your real numbers.
• Come back with a short list they can compare and understand.
• And hand it off in both directions.

None of this is new. It's what a good loan officer does in a fifteen-minute phone call. We're just moving it to your website because that's where these borrowers are.

Now, how do you know if any of it's working? I'll warn you now, if you measure it the way most institutions would, it's going to look like not much happened.

17:42 — The funnel model

Here's the measurement trap, and it catches almost everybody.

You should see two changes. More of your loan page visitors start an application, and then more of the ones who start finish.

Each one may move the needle a bit, and in most institutions, those movements belong to two different teams. The start rate sits with marketing. The completion rate sits with lending operations. Each one looks at their own number, sees some modest movement, and moves on.

But nobody multiplies them, and that's where the primary benefit lives.

So if you're on the marketing side of that split, this is the number worth walking down the hall with. It only exists when both teams are looking at it together.

Let's walk through it. Let's consider a mid-size institution's mortgage funnel. Swap in your own numbers as we go.

4,000 monthly sessions on loan product pages. That doesn't change.

Application starts today, 100, or two and a half percent. Put guidance in front of the application, we now get 120, or three percent.

That's a lift of half a percentage point. If that's the only number you report, you're missing something important.

Now let's look at the application side. Completion rate today, 60 percent. With decision-ready applicants coming in, 70 percent. A ten-point swing.

Now, here's what gets missed. You need to multiply the two together. The 60 previous submitted applications becomes 84.

Two rows in this table are doing the credibility work.

Applications abandoned before submission, 40 today, 36 with guidance. That row is measured. Your origination platform counts it, and notice how little it moves. Most of the gain here is people who show up already decided. Very few of them are rescues.

Submitted to funded, 65 percent in both columns, held constant deliberately. Underwriting is unaffected by anything in this video. If a model showed that number improving, I'd want to know why, and so should you.

Funded loans go from 39 to 55. Sixteen more from the same website traffic.

One more thing before you take this anywhere, because this is where a lot of vendor math falls apart in front of a CFO.

Every row in that table is labeled, either measured, modeled, or derived. The sessions and the abandonment counts are measured. Those are in your systems right now. The two lift numbers are modeled. Those are the only two things I'm asking you to take on faith.

Keep the labels on when you present it. A CFO who can see which two numbers are assumptions will argue about those two numbers. But move the modeled values up or down to whatever you're comfortable defending.

Run it at half the lift, and it still clears. That's the version to bring, the conservative one, and the assumptions marked.

Which leaves one question: how do you find out whether any of this applies to you?

21:13 — The audit

When we started, I promised I'd leave you with something you can do without buying anything. So here's the five questions. Call it homework. You can answer most of them from systems you already have.

First question

For one loan product, can you put your application start rate and your completion rate side by side for the same month?

The test here is whether you can get them, not what they say, this week without somebody building your report.

In most institutions, those two numbers live in completely different systems owned by different people or departments, and they've never been looked at together.

If that's true where you are, then whatever is happening in your funnel, you won't see it, good or bad.

Question two

Where do most people often drop out of your application, and is there any evidence they decided anything before they started?

The first half you probably know. The platform reports it. The second half, nobody asks.

Pull your drop-off step, then look at what those borrowers did before they opened the application. If they landed on the product page and started applying inside ninety seconds, that's somebody doing research.

Question three

Have you ever asked an abandoner why they left, or have you only assumed?

Most institutions assume friction, and there's a reason. Friction is the explanation that comes with a fix attached. Twenty phone calls to people who started and didn't finish will tell you more than another round of form optimization. Ask what they were trying to find out. The answers tend to be uncomfortable and useful.

Question four

What share of your inbound lending calls are questions your website could have answered?

This one costs nothing. Have your call center or your loan officers tag calls for two weeks. One bucket for questions the site could have handled, one for everything else. It's usually the most persuasive number an institution puts in front of a CFO because it's their own data about their own borrowers.

And question five

Open your highest volume loan product page as if you were a borrower. What's the first thing on it that responds to your actual situation?

Something that responds, takes what you tell it and gives you back something specific to you. A rate table doesn't do that and neither does a product page.

Count the clicks to it. For most institutions, the honest answer is the application or a phone number. If that's your answer, you've just found the gap and you found it on your own site in about four minutes.

And notice what you're grading, the page, not the person on it. That's true of all five. You're scoring your own funnel, not your borrower's.

Now put those five together

• If you can't get the two numbers.
• If you've never asked an abandoner why they left.
• If your call center is answering questions your site could have handled.
• And if the first thing on your loan page that responds to a borrower is the application itself.

Then what you've got is people showing up at a decision they haven't made yet and nothing on your site to help them make it.

That's the decision readiness gap I named at the start, and it won't appear in any report you run today, which is why it's lasted this long.

And if you came out clean on all five, you're ahead of most institutions I talk to, and you probably don't need what I do.

Where Fintactix fits

Forty seconds on us, and then I'll let you go. I mentioned at the top that I run Fintactix. We've been at this since 2011, and we work with more than 200 banks and credit unions on the digital layer in front of lending. Financial Navigators are our version of what I just walked through. Guided decision experiences for home affordability, mortgage, home equity, and vehicle loans. More to come. They ask about a borrower's situation, run it against your actual products and rates, show the options with the ability to compare and hand off somebody who has made a decision.

They're not a replacement for your origination platform. They sit in front of it.

If you want to see one work, there's a Schedule an Online Meeting link on our site. Thirty minutes is all we ask.

Two last things. Share this with whoever is on the other side from you. If you're in lending, send it to marketing. If you're in marketing, send it to lending.

And start with the five questions. The worksheet below this video has all of them, plus what a bad answer looks like and where to pull each number.

Run them on your own funnel first. Whatever you conclude after that, you'll be working from your own data instead of mine.

Do you have a decision readiness gap? We'd be happy to help. I'm at bill@fintactix.com. Thank you for your time.