Is it really the employees, or the managers? There could be dilapidated business processes still in place that even prevent the company from healing itself. An interesting take on this is the book How The Mighty Fall.
Some relevant things to know might be if they have restructured a lot, and what their churn has looked like both in the executive suite and in engineering.
- Fractured product line; they have a bunch of products that are similar but different and they themselves are confused about which is which. Product management is in chaos.
- Tech debt from the 90s that has never been paid off.
- Acquisitions that were poorly integrated.
- Some weird attempt to mash their platform together with ebay that never quite fully baked, then more debt incurred hastily ripping them apart again.
- Some attempts at outsourcing development overseas? Can't be certain about this, just an impression.
- A major attempt to rewrite their API, but instead of simplifying they got an architecture astronaut to design it. He wrote a lot of annoying blog entries about HATEOAS but AFAICT no longer works at the company. The modern API was published but it's broken and crippled in so many different ways that they would be better off putting a big "DO NOT USE" at the top of their documentation.
- They've effectively abandoned the new API and apparently are relying on a sort of reverse-takeover from Braintree to save them. Maybe there is progress but it's been a while and I don't see any.
All most merchants wanted was "let me accept credit cards with quick and easy setup, even if it's expensive." (Hence the success of Square, Stripe, etc.). All they needed to do was to offer a few APIs that acted like every other credit card gateway, and they'd be fine.
But instead, they went whole hog on telling consumers "you never need to give the retailer a card number." Now, I'm skeptical this was ever a huge selling point, but it means that any API they can ever offer will always be a nonstandard mess to work with because it means you're going to have to bounce users off your site to finish checkout.
In a business where "the hard technical problem" is how to stop fraud while allowing legitimate transactions to occur, you'd expect bad engineering to mean that the fraudsters win and the whole thing blows up. But what seems to have happened is that they've simply overcompensated on the other side. It's easy to stop fraud if you just flag everything as fraud. You end up with pissed off annoyed users, but you don't go bankrupt.
It's too bad. I really wish PayPal was a vibrant alternative to Visa & Mastercard; they fundamentally have the trust model right. It's just that they've been failing at execution for decades now and even with the braintree acquisition it looks like they're dead in the water, milking the last of their fading reputation.
Stripe is really the class act here, but they're playing a very conservative game. I wish they "owned" the payment process more instead of just being a front for credit cards. Maybe someday.