Confessions
‘Banks can’t attract the talent they want or need’: Confessions of a banking startup analyst
- "The origination number alone is not enough information to say if they're doing a good job."
- "For an engineer, the pay is generally competitive, but the working environment and the types of problems they get to solve absolutely aren’t."

Banks have always proclaimed themselves as technology companies with banking licenses. But culturally, banks are still banks: conservative. Innovation teams can only innovate so much before someone in legal or compliance tells them no; they can only move so fast before someone tells them to slow down.
In this installment of Confessions, in which we trade anonymity in exchange for honesty, we spoke with an analyst at a startup attached to a large bank about internal innovation, attracting strong talent and why alternative bank service companies should get serious about becoming or partnering with a bank.
When you work at a startup within a bank, does it feel like you work at a startup or a bank?
I didn’t interact very frequently with people outside of the startup, but culturally if people or my family asked where I worked, I’d say the bank. For people that transferred from other parts of the bank to the startup unit, the culture was markedly different, but coming from the outside, it felt like I worked at a large, corporate bank.
What kinds of people work there?
A lot of engineers and software developers that came from other parts of the bank; a lot from traditional banking — places like Amex and Citi. I’d say 15 to 20 percent came from other startups, including Lending Club and OnDeck but also companies like Blue Apron.
What fraction of your staff transferred internally from other parts of the bank?
Close to half, maybe 40 percent, but not necessarily because they were highly knowledgeable or really great at what they did. It’s quicker to transfer people than it is to hire from outside. It could take three or four months to hire externally for the bank; making an internal transfer, immediately.
Are they knowledgeable and good at what they do?
Some on the development team were really good engineers, some weren’t. Regardless, if they came from other parts of the bank they weren’t experienced in building products for retail customers; their focus was much more on the back-end system execution and not necessarily the user interface (which came out really well because the company used a third-party agency to build it). As we transitioned that stuff in house it became clear the talent wasn’t used to building a product for retail customers, they didn't have that experience of caring about the user experience or having to know the right questions to ask because they didn't have to do it before.
It’s far from the only bank to call itself a tech company. Do you agree?
Technological change, though slower than in other industries because of this regulatory moat around banks, is happening. Things are slowly moving in that direction. To drag these systems out of the 1980s and build something more accessible and interesting does require talent, but established, institutional banks might be overselling themselves on the ability to attract the kind of talent they want or need.
How so?
They're aware they compete with Silicon Valley startups, so they let people wear jeans now — for the company that is a big deal. But for an engineer, the pay is generally competitive, but the working environment and the types of problems they get to solve absolutely aren’t. You think of Google and Facebook as these places with their private chefs and onsite barista and the thing the banks throw you is that now you can wear jeans.
Do people like working there though?
The most talented developers I know want to work on something that is an interesting problem. So much of what I've heard from the technical personnel at major banks is that the stuff they deal with — especially the legacy systems — is not interesting or fun to work on.
Products seem to be doing well, or at least hitting their targets.
I would posit that the origination number alone is not enough information to say if they're doing a good job; you also need to know how much they paid to originate that. Anyone can go and churn out millions of pieces of direct mail and originate those balances — the question whether or not they have done it profitably.
Who’s doing well in the space?
The competitor that’s done the best job is probably SoFi because they have an ecosystem of products where they can cross-sell and upsell. You can refinance your student loan debt but also get wealth management and a personal loan and whatever else they offer next.
So more focused startups expand their product offerings?
If you're going to spend millions and millions of dollars on traditional brand marketing — assessing success by measuring brand awareness, brand favorability, purchase intent — you need to amortize that cost over a number of products. If a single brand can do loans, small business loans, savings accounts, CDs, and wealth management and people broadly know that brand, that boosts your ability to acquire customers across a portfolio of products.
What are you following as you move into another year?
I'm looking at what happens as defaults go up. If you're someone like Prosper or Lending Club -- they only make money when they write the loans. They get a 0.9 percent or 1 percent servicing fee but the majority of their revenue comes from the origination fee when writing loans. If defaults go up and people tighten credit policy, that’s going to hammer Lending Club's revenue immediately and their share price is already in the toilet. The second thing is cost of acquisition; it's expensive to acquire these customers whether you do it via direct mail or via Credit Karma or LendingTree. Everyone’s going after the same people and it gets very expensive. And the more mail everyone churns out, the less responsive those customers get and those campaigns become less effective.
How is marketing different between large financial brands and small financial startups?
The way banks think about it tends to be different than how startup people think about it: they're just going to spend buckets of money building a brand. The startups are 100 percent focused on the unit economics of the next customer they're acquiring which is why they favor channels like Credit Karma and the affiliate ecosystem where you can at a very detailed level say, yes I made money from this account, no I did not, and leave the brand to the side.
Are fintech companies — old or new — starting to look the same?
Any of these companies who are not banks but offer loans, to do most of the other stuff they would want to do, have to partner with or become a bank. You saw it at Sofi, where the message for so long was “Dont bank, SoFi.” Goldman already has plans to fold its deposit platform into the Marcus brand. Barclaycard currently exists but they want to pivot to just be Barclays. No matter what the entry point is — personal loans, payments processing, small business loans — the direction seems to be the same: offering a suite of financial services products. And to do that, at least right now in the U.S., you have to have a bank charter or partner with a bank.