Back to all posts

The ACTUAL Reasons Why Fast Failed 🔻 image

The ACTUAL Reasons Why Fast Failed 🔻

Apr 9, 2022 | Important PSAs

This originally was posted on Twitter @theryanking

SO, I feel for the Fast team. I know several. Earnest, passionate people.

I also think @domm is creative and capable. I’d likely back him if he started a new company.

The culprit: Stripe pushing them to scale before they knew how.

This forced major compromises.

Yes there’s a lot of gossip about unhealthy internal culture.

That’s unfortunately common at most early startups and is not what killed them.

Their unnaturally forced business strategy was what did them in. Let’s break it down 🔻🔻🔻

~~ Going after micro merchants ~~

Unlike payment processing, ecommerce checkout is NOT self-serve. Each implementation is tough.

Thus why Stripe decided not to do this on their own!

Different tax, shipping, coupon, discount, gift card, shopping cart, ERP, and OMS systems.

Complex, HAIRY integrations.

It’s just as expensive to implement & service a merchant doing $1m in sales as $100m.

Tiny doesn’t scale here!

… Stripe is the opposite: self serve, get lots of devs, some blowup

At Bolt we learned that SMBs were a unit economics death-march and pivoted in ~2019 to enterprise.

Fast was able to list a higher merchant count, but with 1/100 of the underlying GMV or revenue.

~~ Payment processor lock-in ~~

A checkout business fundamentally requires payments flexibility.

Merchants have different payment processors, oftentimes even multiple.

Stripe wanted to be the exclusive payment processor.

This “deal with the devil” gave Fast no chance from day one.

This is yet another reason why Stripe or other payments companies cannot be checkout companies.

~~ A button vs the full checkout ~~

Taking over a checkout experience end-to-end is excruciatingly hard given all the integrations required.

It took us nearly 8 yrs of building do this reliably!

Fast quickly was like “oh sh*t” this is hard, and came up with a workaround.

Product page checkout!

The pros:

  • Much easier integration.
  • For any feature not supported, they could kick a user to the core checkout.
  • This avoided integrations with gift cards, coupons, alternative payment methods, etc.

The cons:

  • Tiny share of checkout, even less revenue and GMV per merchant
  • Clunky shopper experience
  • Messed up reporting

… Sophisticated merchants would never adopt

~~ Focusing on consumer vs merchant ~~

We always knew that to build the consumer network we FIRST had to make merchants love us.

Fast didn’t realize this and started building consumer features way too early.

At Bolt we spent our time on CheckoutOS, the checkout operating system for merchants.

And building every integration under the sun.

They spent their time on consumer portals.

~~ Missing on platform partnerships ~~

Bolt quickly realized that deep embedded partnerships with the ecom platforms would be paramount.

As we went to go sign deals with the platforms we were shocked:

Fast was nowhere to be found!

Not even putting up a fight.

~~ Being second-mover in a network race ~~

This isn’t as much a strategic decision as it is a reality.

The checkout company with the largest network has a huge advantage.

Bolt now has ~14m shoppers and is expected to grow to ~100m in as soon as 12 mo.

In short, Bolt did the HARD long-term thing.

  • Build CheckoutOS, the checkout operating system
  • Take over checkout end-to-end
  • Write all the integrations
  • Focus on merchants
  • Focus on enterprises and platforms

Fast, driven by Stripe’s irresponsible funding, took the opposite approach.

  • Build a button, not a full checkout
  • Sell to micro merchants
  • Focus on consumers
  • Support minimal integrations

This led to unsustainable growth and ultimately failure.

Mob Boss Stripe thought it could pour a pile of money into a company with zero revenue to try to kill Bolt.

Hype + cash = engineers = success, right? It worked for Stripe.

The problem: checkout doesn’t work like that.

It’s much more Enterprise Software than Developer API.

They made a HUGE mistake.


My thoughts on the whole situation?

Read More:

Follow me: