SBF Profile picture
4 Nov, 20 tweets, 4 min read
1) On Hubris
2) "Two vast and trunkless legs of stone
Stand in the desert. . . . Near them, on the sand,
Half sunk a shattered visage lies...

And on the pedestal, these words appear:
My name is Ozymandias, King of Kings;
Look on my Works, ye Mighty, and despair!
Nothing beside remains."
3) Each year, the world stands witness to a giant company, felled by its own arrogance.

And each year the world averts its eyes as many more enterprises dream too small and believe too little.
4) I guess I want to focus on one of @FTX_Official's larger failings, historically: user signup flow.

The core of our issues can be summarized by the following fact:

Until recently, signing up required a user to manually enter their address 2-3 times. Once, really, is enough.
5) This probably slowed down our user growth a lot: entering an address takes a while, and no one wants to bother doing that multiple times! (Our internal numbers, FWIW, back this up.)

So why did we end up there?
6) It's not, to be clear, that no one had ever mentioned to us that our registration flow could stand to be tighter.

We were quite aware of this, and agreed!

But for about a year, we made very little progress on addressing it.
7) We knew that our signup UX could be streamlined, and that doing so would be valuable.

People would bring it up frequently as a reason that we couldn't compete for retail.

And my response, generally, was:

1) Yeah it sucks right now
2) Don't worry, I'm 98% sure we can fix it
8) While we were paying a price for not fixing it, I wasn't super worried strategically.

We have built a number of complex, sophisticated, constrained systems to function at industrial scale.

I had no doubt that we could consolidate pages in our registration flow.
9) In fact, it was _so_ much easier than the trickiest things we've built, that from a medium term strategic perspective I didn't bother thinking about it.

There are a number of difficult tasks ahead of us if we want to get FTX to where it could be; this wasn't one of them.
10) If worst came to worst, at the last minute we could divert all dev and designer time for a few weeks and get it in decent shape.

Instead, I focused on the things that would take everything we had to execute on well, and were also strategically crucial.
11) We were working on complex licensing; matching engine overhauls; risk engine maintenance; product design; and NFT partnerships.

Those were hard challenges that would require months of careful planning and attention, and which we wouldn't get right unless we focused on them.
12) The last few weeks we've finally started overhauling our signup experience.

And, as predicted, we've made very quick progress on it, and I expect we will continue to do so.
13) I'm glad we started planning for LedgerX/FTX US Derivatives; the NFT platform; and the matching engine first, and the UX overhaul second, rather than the other way around.

We were confident that we could do a good job at registration--so confident we took it for granted.
14) And this made strategic sense: trusting in our abilities, and understanding that the progress we've made on the hardest parts of the product wasn't a reflection of some weird risk-engine-specific aptitude, but something more general.
15) But at the same time: for months, we stunted the growth of FTX's userbase by not doing something relatively easy and extremely valuable.

We paid a real price for this, and will continue to do so until we further refine things.
16) Our confidence that we _could_ execute well stopped us from bothering to actually do it.

Our confidence started to veer into hubris, and we risked failing to become what we could be by just not bothering to follow through.
17) And I say 'our', but I take the lion's share of blame here; my job, ultimately, is to make sure we do the important things, and this was important and didn't get done. I was dismissive of it, and so it was dismissed.
18) And, of course, while I remain confident we will quickly pick up the low hanging UX fruit, there's a looooong tail of things that can help make a user's experience beautiful, and we have our work cut out on it.
19) One of the curses of achieving quick, early success is that you can be lulled into complacency, forgetting that the ability to do well is only worth anything if you keep pushing forward.

One of the blessings, though, is understanding how much potential lies ahead.
20) Anyway, my main takeaways are:

a) Aim high, but don't forget to do everything you can to actually reach your lofty goals.

b) Watchmen: episode 3 ("She Was Killed by Space Junk") is beautifully scripted and executed.

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with SBF

SBF Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @SBF_FTX

4 Nov
1) As the 2021 @MLB season comes to a close, some reflections from 21 years as a fan of the @SFGiants.

I suck at baseball.
3) First of all, and unrelatedly -- _huge_ props to @tackettzane, Lou, and the MLB team for the partnership over the past year. We're extremely greatful for being able to work together!
Read 25 tweets
23 Oct
1) Filing taxes isn't fun. As it turns out, for many people, it's also a bit of an odd exercise.…
2) See, the government actually already has a lot of the relevant information.

But rather than make that available to the taxpayer, it sends a form asking you to enter your income, investments, etc.

Asking you to tell the IRS what it already knows.
3) The consequences of messing it up can be years of lawsuits and frustration.

Filing taxes is basically the government issuing you a test, judging how well you do, as it is actually collecting new information.
Read 8 tweets
16 Oct
1) Don't let the drive for a narrative overcome your priors; aka the slow, sad decay of fivethirtyeight.…
2) This article is dumb.

It is centered around the observation that, this season, Lamar Jackson has been better on early downs while Justin Herbert has been better on late downs.

Why is this????
3) Sample sizes, mostly.

The sample sizes so far are pretty small -- around 60 3rd down attempts per team. The difference between the Ravens and Chargers completion % is about 14% (35% vs 49%); the standard deviation due to raw luck here is ~40%/sqrt(60)*sqrt(2) ~ 7%
Read 13 tweets
15 Oct
1) another lazy, incoherent take on COVID:…
2) as with many, this one doesn’t bother attempting a cost benefit analysis.

If it did, it would maybe notice that the claimed annual damages from the flu are around $50b.

How about it’s proposed countermeasures?
3) well, US GDP is $20T, so we have 0.2% to play with—0.4% if you double it to include endemic COVID.

Maybe his proposed interventions would halve the risk—back to 0.2%.

That’s the equivalent of missing 1 day per year of work.
Read 8 tweets
13 Oct
1) A while ago I wrote a thread on VCs:

I've now been through a LOT more raising and investing. Some updated thoughts on the process!
3) At its heart, I've been trying to understand the following question:

Let's say a company had $100m of revenue and $90m of expenses in the past year.

It goes out to raise $50m. What valuation does the raise happen at?
Read 26 tweets
7 Oct
1) Ok come on TIME, this article is pretty pathetic.

How did no editors catch that it was logically impossible?…
2) so, if all relationships were monogamous and heterosexual, and there were as many women in the world as men, then the effect of “economic pressures” on the ratio of single men to single women…
3) …would be null, because there would by definition be exactly the same number of each. There would be some number of couples, and that would be exactly the number of men and of women in a relationship.
Read 7 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!