Hi all! Charley here again.
There've been some great Marqeta S-1 breakdowns published already that I’ll be highlighting through my post, but I’ve written a fair amount about Marqeta in the past and couldn’t miss my chance at diving in deeper myself.
Here we go:
Pivoting into infrastructure by scratching your own itch:
Jason Gardner wrote a letter that touched upon how the company got started - I’ve written about the origin story previously for FTT, but it’s pretty incredible to see where Marqeta is now vs the grocery loyalty card product it initially started as.
Companies get started with problems that are fun to solve. Over time the problem that gets solved launches a product that finds a market. As the story goes, I was eating sushi with my friend Sukhi at a restaurant in San Francisco in December of 2009. I had recently left MoneyGram International, which acquired a company I co-founded. I wanted to start something new. Sukhi had a bunch of Groupon coupons in his pocket, and he said, “You’re a payment nerd; find a way to put these on a card.” I remember looking at the payment terminal in the restaurant and thinking that there was a world beyond that machine that I didn’t fully understand. That intellectual curiosity led us on a path to where we are today.
We’ve seen this story repeat itself for many other critical infrastructure companies, particularly in fintech. The Collison brothers first started working on Stripe in 2010 after being frustrated with how difficult it was to accept payments on the web. Zach and William started Plaid after trying to build a consumer facing application that required bank account data.
Competitive landscape, or rather, where’s the real one:
Marqeta laid out three primary categories of competitors, and specifically called out TSYS, First Data, FIS, Wex, Comdata, Adyen, and Stripe.
I (obviously) think that the list of competitors is much larger, but you have to look a bit more earlier stage. I would place companies like Apto Payments, Bay1, Extend, Privacy in the potential “disruptor” category - with many of these providers having a particular focus on the developer experience. I also found it interesting that Marqeta didn’t include companies like I2C or Galileo, considering that Marqeta has lost some large logos to program management providers like these (including notably Brex who switched off Marqeta in the past couple of years).
John Street Capital on Twitter highlighted the comparison in valuations and multiples of SoFi’s acquisition of Galileo for $1.2B vs where Marqeta is expected to land at $15B+ - a 6X multiple looks like an absolute steal now for SoFi...
Interestingly enough, Marqeta itself claims that its developer experience is a selling point
If the basis of a company’s success has become its ability to create relevant and compelling user experiences, it is the software developer who leads this process. It is now developers who influence some of the most important business decisions, and they, in turn, demand modern platforms that are most likely to keep up with the pace of their imaginations—with tools and services that are of the highest configurability, flexibility, and speed.
But the game that Marqeta plays vs some of these other providers is quite different…
The name of the game is whale hunting:
I wouldn’t describe Marqeta’s GTM strategy as particularly startup friendly or efficient - and it starts from the developer buying experience. It does a solid job on having public documentation, updated sdks, etc - but the main call to action is to get in contact with a sales team and get qualified before getting access to any sort of live environment.
Marqeta’s game is really whale hunting - prioritizing resourcing and sales efforts on landing large, high growth accounts that will drive significant processing volume. Competitors in the market such as the ones I mentioned above are attacking the developer wedge, making it as easy as possible for a developer to spin up cards with the hope of onboarding the next Doordash, Square, or Ramp right at (or even before) official company formation! Marqeta has also increasingly prioritized enterprise revenue - announcing initial deals with JPMorgan to power tokenized card issuance for commercial cards.
Everyone has already highlighted that Square represents 73% of net revenue, which is obviously a very significant concentration considering the next customer doesn’t hit 10%. NRR measured across the revenue base is fantastic, but annual cohort growth does look a bit weaker -
Interchange is a weirdly amazing way to make money:
James at Altimeter did a great job laying out the take rate across the various providers involved in a transaction- the issuer (Square, Doordash, etc), issuer processor (Marqeta), network (Visa, Mastercard, Discovery, etc) and Issuing Bank (Sutton).
Tanay goes a bit further in his breakdown to lay out Marqeta’s net take rate - which has actually decreased from 30bps to 22 bps from 2019 to 2021.
This is where we can start to see some interesting considerations around Marqeta’s whale chasing strategy.
Most client contracts with Marqeta have built in revenue share splits dependent upon transaction volume. As clients bring on more processing volumes, the contracts are structured such as Marqeta gives up more rev share to the client. My assumption is that as some of Marqeta’s largest clients hit higher tiers, the net take rate retained by Marqeta decreased (e.g. Square and others started to really hit some numbers and increased their incremental take rate)
On Marqeta’s side, it does get some leverage to negotiate higher rebates with the network as TPV increase as well as push down processing fees - but those do tend to be much more fixed.
Interchange is also highly suspect to regulation - one could argue that the US fintech industry probably wouldn’t exist if it wasn't for the Durbin amendment. TLDR - the Durbin amendment caps the maximum amount of interchange that financial institution’s above $10B in assets can collect for debit transactions to just $0.21 + 0.05%. As a result, pretty much every fintech company that issues cards does so through a Durbin-exempt financial institution. In Marqeta’s case, Sutton Bank is its primary issuing bank, processing 94% of Marqeta’s payment transactions. Marqeta has also lined up three other US issuing banks in the event that Sutton Bank ever hits the Durbin limit or if something else ever occurs.
In 2019, the average debit card interchange rate for dual-message network transactions (e.g. Visa, Mastercard, Discovery) was around 1.4%. There have also been recent proposed changes by the Fed which Shamir (co-founder of Simple and Sila) highlighted. Basically, any financial institution that issues debit will need to also support domestic EFT debit to meet requirements of being able to route between two unaffiliated debit networks for Card Not Present transactions. As a result, any issuer that has a significant debit portfolio will see its overall interchange drop as debit transactions processed over unaffiliated networks offer less interchange.
The other driver around interchange take rate is the mix of card products that Marqeta’s customers utilize. Square cash is a debit product whereas something like Ramp would fall under commercial credit or prepaid. Commercial credit interchange as an example goes up to 2.5% (Bill.com’s acquisition of Divvy highlighted this, with Divvy at $100M in annualized run rate / $4B in annualized TPV = 2.5% gross interchange). Marqeta’s take rate could potentially increase if it's able to negotiate harder on its take rate for other products, but it’s generally known in the market what a competitive issuer rate is and hard to move up from the 0.30% range.
Fun cap table tidbits:
I also spent some time going through the cap table, and there were some fun tidbits.
For example, as an incentive for select customers to continue to use its platform and not churn, Marqeta has occasionally issued out stock warrants that vest upon attaining specific milestones. Square, Uber, SVB, and Ramp all have such warrants.
There was also an interesting termination agreement that I’ll just leave here as food for thought. Omri Dahan, Marqeta’s CRO, announced in March 2021 that he would be leaving the company, which was interesting given the imminent IPO. He was also one of the longest turned executives having been at the company for 10 years.
In March 2021, we entered into a separation agreement with Mr. Dahan whereby we mutually agreed that his employment with us as our Chief Revenue Officer ended on February 1, 2021, and that thereafter he would serve as a non-executive employee at his current annual salary and benefits until the first to occur of (i) April 1, 2021, or (ii) a certain project renewal by us (the Separation Date). In connection with the separation agreement, Mr. Dahan received (1) a one-time lump sum payment of $262,500; (2) accelerated vesting of his outstanding option awards equal to that number of shares that, absent his separation, were scheduled to become vested through April 1, 2021; and (3) continuation of health insurance under COBRA or Cal-COBRA for nine months following the Separation Date.
I was also surprised at the amount of secondary that was sold through some of the later rounds. It’s become much more common to give founders secondary in later rounds (especially if they’ve been illiquid for a while). The general advice tends to be that investors should allow founders to sell enough such that they don’t have to worry about housing, family costs, etc and incentivize the team on going big vs selling early. You also want to balance against giving full on retirement sums of money.
In total, Jason Gardner has sold $57.3M worth of shares through secondaries and Omri Dahan sold $31.8M worth - although vast majority was in the later stages as the company was gearing up for an IPO.
- 2018 - ICONIQ acquired shares from Jason Gardner and Omri Dahan - $3.5M and $2M
- 2019 w/ series E - Jason Gardner sold $26.9M worth of shares, Omri Dahan sold $14.9M
- 2020 - Jason Gardner sold $26.9M and Omri Dahan sold $14.9M
It's also super fascinating to see who’s going to have a big payout at the IPO. Granite still owns 11.2% of the company since leading the series A and investing into every subsequent round. DFS Services (aka Discovery) also looks like it's going to score a big win with 5.4% of the cap table. Meanwhile, Jason has been diluted down to 14% of the company.
The later investors in the Series E-1, E, D-1, and D also had liquidation preferences which had higher preferences vs other holders of series C, B, and A stock. Through my reading, all the investors had a 1X liquidation preference but not all the prices were disclosed :)
In the event of any Liquidation Event (as defined above) of the Company, the holders of Series E-1, Series E, Series D-1, and Series D redeemable convertible preferred stock shall be entitled to receive a payout of $8.337, $3.891, $1.205, and $1.205 per share, respectively, plus any declared and unpaid dividends, prior and in preference to any distributions made to the holders of Series C, Series B, and Series A redeemable convertible preferred stock and to the holders of common stock.
After the payment of the full liquidation preference of the Series E-1, Series E, Series D-1, and Series D redeemable convertible preferred stock, the holders of Series C, Series B and Series A redeemable convertible preferred stock are entitled to receive an amount equal to $0.353, $0.295, and $0.057 per share, respectively, plus any declared but unpaid dividends, prior and in preference to any distributions made to the holders of common stock.
I’m excited to see how the market responds to Marqeta’s story. Will investors be excited at the rapid growth rate, a seemingly huge greenfield market, sizeable international opportunity (international companies represent just 2% of Marqeta’s revenue to date), and a hot market aka fintech? Or will they be scared away by the revenue concentration, potential upcoming regulation, and increasing competitive pressure in the space over time? We’ll find out soon!!