The Founder's Series: Loop Payments
or how to build meaningful vertical fintech and fix freight payments
Fintech has long promised the complete transformation of all sorts of traditional payments and industries and it’s still early days. Up until now, much of the attention has focused on digitizing consumer payments and infrastructure. Now, much of the space is abuzz with the opportunity in B2B fintech.
And that attention is well warranted! The TAM of all B2B financing, payments, and more is absolutely massive.
But there is a catch: It’s really, really hard to build a fintech platform capable of serving all these disparate stakeholders.
Why? Well, there’s a couple key reasons. But there’s one really important overarching reason.
Fintech isn’t really about the actual money movement.
You can really boil the fintech space into the two different eras: Before Stripe and After Stripe.
Before Stripe, digitizing basic money movement problems was really hard. Stripe had to essentially figure it out from scratch, both from a regulatory perspective and a technical one.
But 12 years later, the fintech space knows how to move money both from a regulatory standpoint and from a technical one. There’s great infrastructure to do so.
What’s challenging about fintech isn’t so much the core problem of moving money, but the data surrounding why money should be moved.
That problem, we are still relatively in the first inning. And it’s why B2B fintech is going vertical.
To add some color, here’s why Loop’s CEO, Matt McKinney thinks the future of fintech is vertical:
The nuances of billing is highly complex and vertically-specific.
The foremost problem in all of fintech is understanding why or why not money should be moved from one business to another. In other words, fintech is most concerned with the context surrounding a payment.
And this context is typically dependent upon the nuances of the vertical.
You can see this most clearly in healthcare, where revenue cycle management (RCM) has become an art form.
Care providers and insurance companies both need to know exactly what procedures were performed on a patient in minute detail.
Bad billing not in accordance with what’s covered? Rejected instantly.
And so RCMs emerged as a tool to ensure that billing was correctly performed to reduce the number of rejected claims.
Many verticals have this sort of complex industry specific data that needs to be contextualized and used to empower participants to act swiftly and accurately around payments.
But revenue lifecycle management isn’t solely concerned with making sure procedures are billed according to spec, they also serve as vendor management systems that allow hospitals, patients, and more to know upfront who they are transacting with and what the provider has agreed to cover.
And that’s another the other huge reason that fintech is going vertical:
Verticals have their own vendors that form important parts of the ecosystem with pre-negotiated contracts for what is billable.
This point prima facie seems somewhat obvious, but it creates huge issues in building meaningful fintech that really does automate away a lot of the pain surrounding industry payments.
To solve that, you need to ingest and contextualize a whole set of documents that don’t originate within your system and understand how the various vendors interact.
Within freight that means dealing with tons of data stemming from the network of shippers, carriers, and 3PLs.
All of this is why I have been extremely excited about what Loop has been up to and why I think they are enacting a playbook that all vertical fintechs will study.
A while back, Packy asked a great question:1
What I find most impressive about Matt, Shaosu, and the rest of the Loop team is that they have called their shot on how they plan to completely change freight payments.
There’s no haphazardly finding product-market fit. They know their path to transforming freight payments, and are focused on executing that vision.
So here’s a little bit about their shot, the team, and how to fix freight payments for good.
Founding
Matt and Shaosu first met at Uber Freight,
But here’s the big problem. Uber Freight was growing incredibly fast. Generating revenue was not the problem. What was a problem was that even as the top-line continued to grow, the bottom line didn’t.
And one of the large culprits was billing and payments – figuring out what to pay, what they were owed, and doing it all in a timely manner.
That seems kinda mindblowing, that is until every freight player also raises their hand and confesses that they have the exact same issue.
And that’s pretty much the experience that Matt and Shaosu ran into. Didn’t matter if it was a shipper in Little Rock, a small carrier in the middle of Indiana, or an advanced digital freight broker. Everyone had the same set of issues around billing and payment.
Logistics and the Billing Regression
The issue in part stems from the same gnarly interoperability and data issues found throughout the rest of freight and partly from how logistics companies contract shipments.
Much like the rest of freight and data, simply understanding what is in an invoice is problem #1. There’s vast differences in how your invoice might show up. There’s different reasons for this complexity, but most of it boils down to a lack of standardized data. Some of the same EDI peculiarities show up here, but the vast majority stems from the underlying contracts and rates.
An invoice is really an attempt to structure a financial understanding of what services were performed. But when that structuring attempt is done by a human from unstructured data - things like fuel surcharge policies and mileage data (that depending on the provider will show up differently), you’re left with a bunch of things that can go wrong. And that means that our financial manager receiving the invoice also has to go through all this unstructured data and look at it themselves.
Not only do we have redundancy amongst financial departments; at a fundamental level, the lack of standardized data and aggregation means our financial managers always have to start from the same point - data assembly. And since that takes up vast amounts of time, our financial manager might not ever get to auditing what’s in the invoice itself.
And that second level after assembly is even more complex! Each invoice can have up to 300 different accessorial charges, and up until now that has meant a poor financial manager has to go through each one.
Said differently, billing is really about contracts. Once you understand what an invoice says, you still have to verify that each of these 300 accessorial charges is in line with what was negotiated.
But wait there’s more! Because not only will our poor manager have to go back into the contract to figure out the payment terms and roughly verify the amount invoiced is correct, but the contract itself will say stuff like:
“Fuel surcharges will be applied to all shipments based on the market rate for fuel at the time of shipment. The current fuel surcharge rate will be made available to the shipper upon request and will be included in the total freight charges for each shipment."
So our financial operator gets stuck in a regression, always going one layer deeper to try to understand what should be paid and why; attempting to assemble more data in order to audit it. Performing an audit, only to realize they need to assemble more data.
It’s Audits All the Way Down
But the regression can be solved. It just takes great technologists and industry experts coming together to take a crack.
But it’s also about timing and that’s partly why right now is the perfect time to solve this for good.
You can read more about the paradigm shifts in my freight deep dive, but perhaps the most important one is the general movement to system interoperability.
Historically, the thing that has killed industry innovation is the integration issue. Everyone dreaded the question often asked: How much of a pain will this be to integrate?
The Convoys, Uber Freights, and Flexports have caused the whole industry to graduate to APIs and live data, or as Matt puts it, “we finally have running water.”
Now that we have that, we can start to build higher order platforms.
So what do you do when you have running water?
Well, you get cracking on turning the data running through the pipes into useful insights.
Phase 1: The Audit Layer
In Phase 1, Loop is saving our financial manager from the need to go through every single invoice, contract, and comb through every line item.
To do that, Loop first focused on building core infrastructure to structure and standardize all these disparate documents. That unlocks things that are revolutionary for the industry; even things that seem basic: like standardizing facility data.2
This is really only possible because of the team and the approach that Loop has taken. First, they are using some of the latest innovations in NLP and ML practices to make sense of all these documents.3
Second, it’s possible because the Loop team is a huge proponent of domain-driven design.
Building a model with the right primitives that can handle all sorts of industry specific edge cases is crucial.4 It's the only way to ensure you can actually solve all the complexity within a vertical. Loop's Max Heinritz wrote a great piece about the process that you can read here.
Now equipped with the right model and platform, Loop moved into deriving second-order insights from these documents. No need for our financial manager to go in and out of different documents to make sense of what needs to be paid and when. Loop brings that knowledge to the forefront.
In short, Loop transforms the role itself. Financial management morphs from mere audit assembly to financial analysis. That’s infinitely more valuable and our newly named analyst can focus far more on strategic questions and insights unlocked by Loop.
That’s incredibly powerful in its own right. But Loop hasn’t stopped there.
There’s another thing that Loop has been highly focused on since the early days, and that’s ensuring that every part of the freight value chain can wield these tools. To really fix logistics payments, every single participant has to be able to access the same tooling and operate off of strategic insights.
In short, that means Loop spends a ton of energy ensuring that their platform is neutral and unbiased and available for every part of the ecosystem.
And the best part of any great tool extensible to all parts of the ecosystem? Users come for the tool and stay for the network.
So once you build the network, what comes next?
That gets us to Phase 2.
Phase 2: The Clearinghouse
To reiterate what I said at the outset, the number one thing you should understand about all of fintech:
None of the problems in industry-specific payments are actually about money movement.
One of the biggest issues for all industries is the speed of payments with resulting cash flow issues for businesses.
But here too, the problem isn’t the actual money movement.
To wax a bit philosophical, speed of payments is artificially bound not by the speed of the ACH network, but by contracts and by the speed that invoices can clear.
I’m talking here of course about the net-30, net-45, or longer term that exists in most contracts in every industry.
The term mostly started as a value add term, extending short term credit in order to close a deal. But that’s not really what it is now. It’s a de facto standard.
Structurally it now exists to build in artificial resiliency and time for businesses to audit the transaction and guarantee the contract has been fulfilled. But the result is that in every SME, we have large cash flow issues stemming from the ubiquity of the term.
Does auditing invoices take 30+ days in the 21st century? Only if the tech stack is lacking!
And that brings us back to Loop. Loop’s tackling the payment issue head on with the goal of building the industry clearing house.
This takes two forms. First, Loop’s platform already houses the payment terms from the contract and they’ve gone even further and help their customers orchestrate the payment itself.
That’s already really powerful. There’s no need to log into a bank portal, a financial manager can automate all the payments directly inside Loop. Ready to pay an invoice? Boom. You’re done. That alone is already cutting the time for customers from when an invoice is received and paid from 50 days to 3.
And building this sort of intuitive payments orchestration allows Loop to build the sort of trust to take on the grander component: the clearing house.
Remember how Loop is building a network? This is where that will come to fruition as Loop emerges as the standard for all logistics payments to be settled and cleared.5
And there’s precedent for this, even within logistics, the British Railway Clearing-House (RCH).
The RCH has long been deprecated, but emerged in the mid-19th century as the settlement network for rail companies who needed consensus on what they were owed for certain legs of the transit of goods and passengers.
The basic setup was this: the RCH got all the receipts from goods and services transported over the rail system, apportioned funds to the various operators, and then settled the actual transactions.
Sound similar? That’s because it is. Loop’s end state is a modern logistics clearing house.
As shippers, carriers, and 3PLs start to rely upon Loop more and more, clearing more and more transactions and invoices through Loop, certain synergies will develop.
Perhaps the biggest network effect will be the invoice itself. See an invoice generated by Loop? You can know that invoice comes with some rigor. Ultimately as the network expands, it will serve the same purpose stamps served in the railway days: a testament that this is a true and accurate billing and revenue should be apportioned accordingly.6
In the process, you could imagine if two parties stipulate to use Loop as the source of truth for all invoicing, audits, and fund distribution, we could end up with quite a different payment term in the future: “Payment will be distributed upon an invoice audit within the Loop platform or after 30 days of the invoice being received, whichever comes faster.”
Lastly, if all invoices are flowing through your platform, you can also have plenty of confidence in eventual payment of these invoices, even the ones that take a bit longer to settle. And that means Loop can extend financing to their network accordingly to help alleviate additional cash flow issues when they crop up.
What’s perhaps most fun to think about is the end state of a Loop clearing house. Since clearing houses are prized for their neutrality, they often end up as trusted industry partners to help industries modernize.
The modern form of the RCH, the Rail Settlement Plan, doesn’t merely settle funds, it does much more.
The company now:
Collects retail sales data from 8,500 ticket issuing systems
Carries out the correct allocation of ticket revenue to train operators
Settles that revenue to the operators
Maintains the central industry fares database and provides tools for train operators to set fares
Distributes fares, timetable, station and other industry data to ticket issuing and information systems
Provides the National Reservations Service enabling retailers to book reservations on all trains with reservable seats
Clearing houses end up building all sorts of adjacent products for their network. And that’s probably what’s most exciting about Loop as they continue to execute on the first two phases, phases 3-10 are somewhat unpredictable, yet mean deeper industry support.
But here’s what that could look like.
Phase 3 and On: Expansions and Applications
There’s at least two expansions that make sense down the road for Loop.
Modality Expansion
Freight is a many legged journey across air, ocean, and land. Each has quirks in invoicing and thus as Loop continues to serve more and more of the value chain, they will also serve more and more types of freight movement and the ecosystem. Currently, Loop is fully focused on American trucking. But because they spent the time up front on the domain model, they can absolutely expand into additional freight workflows.
Geographic Expansion
And of course, there’s always opportunities to expand globally. But what’s fascinating about Loop is that none of their business thesis fully rests on capturing all this global activity.
Matt’s thought a lot about the future of freight and what supply chains will look like post-Covid.
To way oversimplify, Covid exposed how vulnerable our supply chains are to global events with corresponding economic and social havoc when its core materials like chips, pharmaceuticals, and PPE supplies.
And so we are already seeing large parts of the economy reshore to prevent global shocks that may cripple their business.
That means that American trucking is going to change quite a bit as well.
With a large reshoring movement, American trucking will morph into an ecosystem with far more middle and last mile routes than long haul routes. Less journeys back and forth to ports and more stops between various suppliers, customers, and more.
That will mean we probably end up with an increase in the total number of hauls, and correspondingly a lot more invoices to process.
So even if Loop doesn’t go global for quite a while, global manufacturing is coming back home and bringing along with it an American trucking renaissance that Loop will benefit from as well.
So what could the other phases look like?
The Application Layer
One distinct possibility is that Loop ends up building other useful applications for different aspects of the value chain. I have in mind here Carta, who had the key insight that a bunch of equities data was locked in spreadsheets. By centralizing that data, they could offer all sorts of additional products targeted to different parts of the private company ecosystem. You’re a stock admin? Here’s your cap table management suite. Investor? Here’s your fund administration product.
The key insight was that once Carta became the source of truth for core pieces of financial data needed by all sides of the industry, they could build additional product lines more tailored for niche workflows.
That playbook could definitely be wielded by Loop or down the road by a startup that uses Loop’s core platform as a base layer.
As Matt puts it: the best part about building key infrastructure is that there’s all sorts of non-obvious applications that others might build on top.
Vertical Visionaries
So there’s a look at how Loop has called their shot and how they are executing their vision.
When Mark Leonard and Constellation launched their venture fund for their employees, Mark made this quip:
If you have a unique insight into a particular customer problem through deep knowledge and experience in a vertical, and this insight is not obvious to agile and deep pocketed competitors, then you have an earned secret.
I think we need a term for these sorts and I prefer “vertical visionaries.”
Vertical visionaries arise from a deep understanding of vertical specific pain, a deep understanding of how technology can impact the vertical, and then savagely execute and build alongside industry experts.
That’s Loop in a nutshell and why I’m eager to watch them continue to build the business and write the playbook for vertical fintechs.
Logistics continues to be a sneaky pick for which industry will have transformed the most by the end of the decade. Loop is going to play a massive role.
One carrier might bill to 5600 Address Avenue, another might bill to 5600 Address Ave. Even something that minor can create all sorts of complications for matching to pricing agreements.
What I love about Loop is that you won’t see them marketing themselves as an AI company - even though that’s exactly what phase 1 relies upon. They’re focused on what the technology unlocks, not on the technology itself.
I love the concept of domain-driven design. It rhymes with a lot of what I covered around the Tenets of Vertical Disruption and primitive selection.
To go back to healthcare: clearing houses are everywhere, sitting between providers and payers and helping clear the claims.
Perhaps most importantly, Loop monetizes through a consumption model tied to payments volume. Funny enough, that’s how every single clearing house works.
As always, great stuff! One interesting conflict I see here is between the notion of a "called shot" per Packy and an "earned secret" per Mark Leonard. This is something I've thought about frequently and my current view is you have to pick one or the other. Thoughts?