Most of the people a bank or credit union serves now do their banking on its website. They open accounts there, shop for loans there, and decide there whether to go further. For a place that important, few people ever ask what the site is for. The question comes up most naturally during a redesign, the rare stretch when the institution rebuilds the site. Even then, the people weighing it talk about how the new site should look. They argue about navigation, about mobile, about whether the homepage works on a phone. They settle those questions first, and they move on before anyone asks the one that decides whether the money was worth spending. What is the new site supposed to do?
That question sounds obvious, and about the website it rarely gets asked. The same institution asks it all the time when it plans a branch. It studies far more than the building. It counts the cars going past and the people on the sidewalk. It checks whether the spot sits near a shopping center with real visibility, across from a large employer, in what the real estate people call a Class A location. The look of the branch matters, and nobody signs a long lease on looks alone. The institution studies who comes by, how often, and why, then puts real money against the answer.
The website gets none of that study. The institution plans it on looks and rebuilds it on looks when the time comes. A redesign is the rare chance to change what the site can do, and most institutions spend it making the same site look better. A better-looking homepage is a thin return on the money.
The person who wants that question answered usually doesn’t own the site or sign off on the vendors who touch it. What they can do is shape what the institution asks for. The clearest chance is a redesign, when the requirements get written down and a designer must answer for them. That is harder than it sounds, because every designer writes “personalized” and “data-driven” into the pitch, whether the site does those things or not. The words are free. The capability is not.
Three questions separate the designers who can do the work from the ones who can only say they can. A designer with the capability answers plainly. A designer without it reaches for adjectives.
“A year later, someone asks whether the redesign paid for itself, and the honest answer is that nobody built the site to know.”
Personalization or segmentation
The first question is about personalization. Most designers deliver something simpler and call it by the same name, so the question must be exact. Ask whether the new site will change what a returning visitor sees based on what that person did the last time they visited. Someone who looked at auto loans twice last week should land on a different homepage than someone who has been reading about certificates. The site should recognize the difference and act on it, without anyone building a campaign to make it happen. A lot of designers answer this by offering to sort visitors into a handful of groups and show each group a fixed page. New visitors see one version, small businesses see another, everyone inside a group sees the same thing. That is segmentation, and they are calling it personalization. The designer sets up those groups once and they hold still. A personalized site keeps adjusting as the person does, visit after visit. Ask the question directly, and a designer who only does segmentation can’t answer it without admitting that’s what it is.
Reporting that reaches a funded loan
The second question is about analytics. Ask whether the designer can trace a visit from the first click all the way to a funded loan or an opened account. Most reporting stops well short of that. It counts visits, time on the page, how many people left. Those numbers describe traffic, and traffic is easy to grow and easy to feel good about. They can’t tell the institution whether the new site moved the deposit or lending goal behind the whole project. The designer must build that connection during the redesign. Connecting a site’s behavior to a funded outcome after launch means tying together systems that were never meant to talk, the website on one side and the platforms where loans and accounts get opened on the other, and that costs more than doing it once at the start. Leave it out, and the institution gets a better-looking site and the same blind spot it had before, unable to say what the site brings in. A year later, someone asks whether the redesign paid for itself, and the honest answer is that nobody built the site to know.
What the site does after launch
The third question is about what happens after the site launches. Ask what it will do six months in that it can’t do on opening day, and who at the institution owns making that happen. A new site is built around what the team believed on the day it shipped. The people who use it don’t hold still, though. Their habits change, new products start to matter, rates move, a branch closes and the people who used it go looking online for what they did in person. The questions people show up with keep moving. A site that can’t read any of that stays frozen at its launch-day guesses and slips further behind every month. It looks current the day it opens and less current every day after. Nobody notices at first, because the site still works. It just works for the people it was built for, and every month there are fewer of them.
A site that reads behavior does the reverse. It learns from what people do and gets better. One Southeast credit union rebuilt its site with personalization that learns, and over time its click rate climbed from 2.7% to a sustained 27.2%, close to ten times where it started. Nobody launched a campaign to do that. The site watched what worked and kept more of it, month after month. It was doing more for the people who used it well after launch than it did the day it opened.
That number belongs to one institution, and a redesign changes several things at once, so another site would land somewhere else. Across the institutions we work with, personalization that learns has raised engagement every time we’ve measured it.
Which redesign costs more
Asking for all three will sound, to whoever owns the budget, like piling extra cost onto a number that is already big. It usually runs the other way. A redesign that only changes how the site looks is the expensive one. The institution gets nothing back against a growth goal between launches. Then, in a few years, it must do the whole thing over because a site that can’t keep up ages out. The money buys a few good-looking years, and then the institution spends it again. A site that reads behavior earns against the goal it was funded to move, month after month, and delays the next rebuild, because it keeps up on its own. The extra capability is what turns the redesign from a cost the institution absorbs into an investment that pays back against the goals the institution set for the year. A budget owner understands that difference, because it’s written in the terms they already work in.
None of these three questions requires the person asking to own the redesign or to know how any of it is built. They work from wherever that person sits. If the redesign is already in the new year’s budget draft, the questions shape what the money is required to buy. If an outside firm is running it, the questions belong in the next review. If the site is staying put for now, the questions are the case for why the next rebuild must do more than freshen the look.
An institution lives with a new website for years, until it builds the next one. These three questions decide what it lives with. They turn a large, rare budget into a site that reads the people who use it and gets better while they do. That is what a redesign budget should be required to buy
