Michael Garrity
Executive Chair
Derek Szeto
Founder & CEO
Scott Elliot
Co-Founder & CEO
Simon Bourgeois
Founder & CEO
Banking, Disguised
Abstract: In this panel on embedded finance, embedded financial services, and the evolving fight for distribution, Michael Garrity (Chairperson, Financeit) moderated a discussion with Derek Szeto (CEO, Walnut), Scott Elliot (CEO, Slate), and Simon Bourgeois (CEO, Nmbr) on what happens when financial products are placed directly inside the platforms, workflows, and customer journeys where people already operate. The panel explored embedded insurance, embedded lending, embedded payroll, partner economics, data access, customer experience, implementation risk, investor perceptions, and the trade-offs between customization and scalability. Across insurance, payroll, lending, and platform perspectives, the message was clear: embedded finance is not just about cheaper acquisition—it is about context, timing, relevance, data, and building financial products that feel native to the customer journey.
👉 Check out the full VIDEO here.
Michael Garrity: This whole panel is going to be focused on the concept of embedding.
If you are a financial technology company, you look for an alternative universe to plug into for some benefit. We will talk about both the benefits and the trade-offs. Is it symbiotic? Is it parasitic? I do not know. That is going to be up to you to figure out as we go through today’s discussion.
Everybody up here, including your moderator, does something in and around embedded finance and embedded financial services. Because we only have a short period of time, I am going to ask each panelist to say the name of their company and how their company qualifies to be up here talking about embedded financial services.
Derek, let’s start with you.
Derek Szeto: I am Derek Szeto, co-founder and CEO of Walnut. We are embedded insurance.
What we do is help partners offer the right products at the right time. We love to leverage data because insurance is sometimes not the most exciting product. So how do we make it fit within the flow, within the customer journey, and help our partners get to the destination?
We work with two of the Big Six Canadian banks, three of the big neobanks, including Wealthsimple and Neo Financial, and two of the big three telcos. Everybody is a little bit different, and it is all about what they know about the customer that is relevant for insurance, and what the right products and services are that we can help them deliver in a seamless fashion.
Simon Bourgeois: Along the lines of boring, our business is an embedded payroll business. We help other companies commercialize their own payroll software.
That ranges from banks, including ATB Financial on ATB Payroll, all the way down to startups that are trying to revolutionize how payroll works in Canada.
We simplify everything on the back end of processing payroll so our partners can focus on building great experiences.
Scott Elliot: I am Scott, co-founder of Slate, and we do embedded lending.
Think fintechs, vertical SaaS platforms, and even banks, empowering them to offer lending to their customers.
Michael Garrity: We are going to double click on what you get from embedding.
The typical way this is framed, even in the fight for distribution, suggests that the benefit of plugging into a larger ecosystem is distribution or maybe lower cost of acquisition. But I think that is just the tip of the iceberg.
What is the payoff? What do you get from being embedded?
Scott Elliot: Besides distribution, we get data around the borrowers.
That can be risk and qualification data to understand whether we can give them capital, or it can be as powerful as understanding whether this person needs capital.
If we have data from your platform, we may know your payroll is due on Friday, that you have invoices coming up next week, and what your cash balances look like. We can understand where that cash flow gap could be and target you with lending offers that actually make sense for you.
It is not just a distribution spray-and-pray play. There is a lot of context around the SMB borrower that we get through a platform.
Simon Bourgeois: In embedded payroll, the benefit is similar.
Most payroll platforms historically were standalone solutions that lived outside the workflow of a typical business. By being an embedded platform, we integrate directly into the workflows that business owners are already using, which simplifies the process of running payroll.
Often, that means access to the data that powers payroll. We work with scheduling software providers, for example, and they already have the hours required to pay employees. For the business owner using that scheduling software, running payroll suddenly becomes easier.
For us, it is partly about distribution, but it is also about delivering a better experience for the people using the products we help build.
Derek Szeto: It is the experience.
Insurance is not the most loved product. So how do you help people get what they need in a low-friction manner?
If you know someone is in a property journey, buying a home and needing home insurance, or if they are a tenant and need tenant insurance, the fact that you know they need it and can deliver it with low friction means the customer is happier.
It is also better from a matching perspective. For the insurance partners we work with, we are always thinking about how to match. If you are in a consistent flow, that is actually better risk than being anti-selected in an open marketplace.
Michael Garrity: I want to stay on this topic.
What do you get out of this beyond cost of acquisition being lowered? Are you getting data elements? Are you becoming a better service provider as a result of doing these integrations?
And what is in it for the partner? You have to convince partner platforms that it is in their best interest not to build something themselves, but to work with a third party that sits inside their system in some way and delivers services they do not fully control. That is a big ask.
Scott, what is in it for the partner?
Scott Elliot: Obviously revenue, but that is the obvious answer.
What we have learned is that lending creates stickiness and adoption for products they already have on their platform.
A lot of our early partners used lending to drive customers toward another feature. It could be payroll, which activates a payroll-linked lending product. Ultimately, they use lending as a tool to create stickiness, drive customers where they want them to go on the platform, and then make revenue through a revenue share.
Simon Bourgeois: My guess is that our answers will all be similar around stickiness.
Building any of these products is extremely hard. Whether it is the regulatory moat or the complexity of embedded financial products, building them is tough.
Being able to add a sticky product like lending, payroll, or insurance in a way that is low friction for the partner is hugely valuable.
In our case, many of the companies we work with already have customers asking them to sell payroll. They say, “You already have all our hours. Why do you not sell me payroll?” Or, “You are my bank. You have all my money. Why can you not help me pay my employees more efficiently?”
We are taking the friction out of offering a product that customers already want.
That matters because product roadmaps are tight, and teams are running leaner than they were five years ago. Being able to launch a high-value product without massive investment is huge.
Derek Szeto: It is some combination of generating revenue and generating stickiness.
Insurance can be valuable from both angles, and you have levers to pull depending on where you want to play.
The other thing is that because we work with many different partners, from big banks to smaller institutions and startups, we can bring expertise to the table. We can help partners move faster.
It is not that they cannot do embedded insurance themselves. We actually encourage people, if they want to do embedded insurance themselves, to go for it. We want to see more embedded insurance in the world. But it is a lot of lift, and there is a lot of regulatory complexity.
We think of ourselves as a partner. Our customer is the distribution partner first. If they find a space where they want to build, we are happy to support them. We do not need to do all the things. It is an ecosystem play.
Michael Garrity: A lot of times on panels, we talk about everything that worked. I want to ask what has not worked.
No names, but when we talk about embedding, the perfect scenario is that you have a partner, you are both symbiotically interested, you have long-term compatibility, the data sets work well, the customer benefits, and the partner moves as quickly as you do.
Give me an example where it did not work, so people here can learn from it.
Scott Elliot: One of our early partners was not moving as fast as we were.
We were trying to move quickly, and because this was a very early customer, we took the path of least resistance to launch. One of the issues that came up was privacy.
For platforms serving customers, if they are thinking about adding products like payroll, insurance, or lending, data sharing and privacy are very important.
In this case, the partner’s existing terms of service and privacy agreements did not necessarily allow them to share data with us. We quickly learned that this has to be one of the first things Slate looks at in due diligence on a partner.
It did not go terribly, but it was a big obstacle and it slowed the process down.
Simon Bourgeois: When we started in 2023, we were eager to get to market and deliver revenue for ourselves and our investors. At the same time, one of the key value-adds for partners is the promise of waving a magic wand and having a product overnight.
Our first batch of partners was eager too, and we did not know better than to let things move really fast.
What we have learned is that even though an embedded product can get you to market faster than building it yourself, you still need to take a lot of time thinking about the experience, the data flows, the legal terms, and how the product will actually work.
Today, we often have to help our partners slow down their eagerness to get to market with a new payroll product and make sure it is tested before it is not just live, but pushed aggressively to customers.
Derek Szeto: Two things come to mind.
First, customers that are willing to pay upfront and do technology integrations are much more successful. We have had some partners that liked the concept but ultimately just wanted a link-out. They did not want to pay setup fees or SaaS fees. They wanted to dip their toe in and try it.
In the embedded world, we want the logos and we want to prove conversion, but there needs to be alignment that they will do some work. In an embedded insurance model, it is not something you can just hand off entirely. They have to be at least a little committed to doing the work, otherwise it will not succeed.
AI may help with some of that by making integrations easier, but the partner still needs to be committed.
Michael Garrity: Let’s zoom back out.
We have startups in the room, and we have financial institutions in the room. Let’s start with the startup question.
An embedded platform inherently depends on partners for the customer base, data sets, and integration. From an investor perspective, that could mean risk. It could mean you are not in control of your own destiny, you do not have a standalone brand, and you could lose partnerships.
How have you navigated the argument that embedding makes you more valuable, rather than embedding being a risk?
Derek Szeto: From an insurance perspective, if you look at the history of insurtech 1.0, there were companies that raised hundreds of millions of dollars to go direct to consumer.
Our belief is that more insurance will be sold in an embedded manner over time. That is the bet.
Part of the reason is that nobody wakes up and says, “I am really excited to buy insurance today.” If you have an insurance app, you are not going to get people to open it multiple times a day.
From our product lens, embedded insurance makes more sense. We think it is better for the customer and better for risk, because if you blast your name out there with millions of dollars of customer acquisition cost, you can attract adverse risk.
From an insurance perspective, embedded can be very valuable. It allows everyone in the ecosystem to win: the distribution partner, the insurer carrying the risk, and us as a business. It is also more capital efficient and can support strong loss ratios.
Simon Bourgeois: We are biased toward the positives because we are entrepreneurs.
I agree with what Derek said, and in payroll, a lot of that is also true. But there are disadvantages to being an embedded provider.
The biggest drawback tends to be the delay in realizing revenue. There is a significant delay in bringing a product to market.
We have an amazing payroll infrastructure product that could be deployed tomorrow. But every time we sign a new partner, we work with them to build a new workflow from scratch, and that takes time if you are doing it carefully and correctly.
What we ask from investors is patience. That is unusual in a world where everything is moving at the speed of AI, and people talk about companies going from zero to 10 million in ARR in two months.
We are often the opposite of that. It is important to identify investors who understand the model and to be honest with ourselves when planning the business.
That is the real disadvantage, because there is otherwise a lot to love about embedded.
Scott Elliot: These are hard businesses to pull off, and I think that is becoming more attractive to investors.
Hard businesses are becoming more attractive because we cannot be vibe-coded over the weekend. Someone cannot leave the room today and have embedded lending by Monday.
There is a lot of effort that goes into the integrations and partnerships. That difficulty can become part of the moat.
Michael Garrity: Now for the bank question.
There are people in the room from financial institutions who are innovation curious. In a perfect world, they take something meaningful from this panel back to their institution and say, “I have an interesting idea for something we could try, a pilot, or an initiative to prove something.”
If you were talking to a financial institution and giving them a pitch they could take back, what would it sound like?
Derek Szeto: Everybody with a large customer base and data about those customers has customers who need insurance.
The question is: what are your customers spending money on in the insurance bucket? It is probably higher than you think. You can help them with that.
How can we, as partners, accomplish that? It is not one-size-fits-all. It has to make sense for your business. But if you know something about your customer, you are already ahead of the game.
Simon Bourgeois: Everyone who works gets paid, so everyone needs payroll in some way, shape, or form.
Traditional financial institutions and payroll have been accelerating a merge over the last couple of years. That is because there are challengers threatening established financial institutions, and those challengers are innovating around the products they offer. They are trying to offer more and better experiences.
Payroll can be an incredible way to offer customers something they definitely use and that is likely to be offered by challenger banks or challenger financial institutions.
There is a real opportunity.
Scott Elliot: In lending, historically it has been expensive to serve small businesses.
Partnering with someone like Slate opens up the affordability and feasibility of delivering these services to your customers.
A lot of your customers are going elsewhere to find lending, and they are not always having good experiences. Partnering with an embedded player allows you to keep that experience in your ecosystem and deliver value to your business customers.
Michael Garrity: From my own experience working with financial institutions, the bigger you are, the harder it is to innovate and the harder it is to do something small.
Almost every one of these things has to start small and detailed if it is going to work. For larger financial institutions, and really for regulated institutions of almost any size, working with a fintech to identify something small and pilot it with managed risk is an excellent way to test innovation without major expense or heavy lifting.
Michael Garrity: One last question.
Partners often want something unique for their platform. They know you can plug into other platforms, so they say, “Here is what we will do differently. Here is how we will customize this embedded product so it feels different here than it does elsewhere.”
That creates a challenge because now you are managing multiple versions of who you are across channels. Have you wrestled with that, and how have you resolved it?
Derek Szeto: Ultimately, the customer is the boss. If they really want something, we can help them find it.
At the same time, we are a matching process. We have to find the capacity, and even if we want to do it, that does not mean insurers can do it or that the opportunity is large enough.
We think about this in phases. You cannot blow the doors off on day one, recreate a new product, and assume everything will work. There has to be learning, testing, and iteration.
The lowest risk approach is often to start with something off the shelf as version one, learn from that, and then make changes in version two if there is traction.
If there is a special need, we can do something unique out of the gate. But generally, it works better to start smaller and then grow into something custom.
Simon Bourgeois: I do not think we have that same luxury.
In payroll, every single integration is extremely customized. Some platforms already have employee information and an employee experience. Others do not. From day one, every partnership ends up being a net new product that is unique.
One of the things we offer is the ability to support end users through our experienced payroll support team. That means we need to be very involved with every product in market.
That creates a challenge around understanding the variances between the products we are helping serve. We have had to become experts in every product we have in market.
From a technical perspective, everything is standard in the background. We do not build customizations from an API standpoint that only one customer uses. But every partner has a unique experience, and our product, engineering, and customer support teams need to understand those differences.
It is one of the key challenges of our business, but we have gotten better at it over time.
Scott Elliot: When we tried the off-the-shelf model from day one, we did not get as much investment or buy-in from the customer.
We have actually seen the opposite. The more customization they ask for, the better the relationship tends to be in the long term.
Handling that is still difficult, but we thought about it when building the foundations of the technology. On the back end, it is all largely the same. To the partner, it may feel like they are asking for a nuance or change, but it should not take us a month to develop.
I like when partners want customization because that is when I know they are invested.
Michael Garrity: That is a declaration of intent.
The thing about the CLA that we love so much is that it facilitates conversations. Anyone you want to grab time with, go and grab them. Our job is to connect, create conversations, and hopefully do something together.
Thank you for your time and attention. Enjoy your conference.
Here are 10 key insights from the panel:
1. Embedded finance is about more than distribution
Panelists emphasized that embedding is not only about lower customer acquisition costs. It also creates access to context, data, workflow timing, and better customer relevance.
2. Data makes embedded products more precise
For embedded lending, platform data can help identify not only whether a borrower qualifies, but whether they actually need capital at a specific moment.
3. Embedded products can improve the customer experience
Payroll, insurance, and lending become more valuable when they appear naturally inside the workflows customers already use.
4. Partners benefit from stickiness and product depth
Embedded financial products can help platforms increase customer retention, drive adoption of existing features, and create new revenue streams.
5. The best embedded partnerships require real commitment
Partners that are willing to invest time, fees, and technology resources tend to be more successful than those that only want a simple link-out.
6. Privacy and data-sharing terms need to be solved early
One key lesson from the panel was that embedded models can stall if existing terms of service and privacy agreements do not permit the data sharing required.
7. Speed must be balanced with careful implementation
Embedded products can launch faster than products built from scratch, but the user experience, data flows, legal terms, and testing still need thoughtful design.
8. Embedded models can be attractive because they are hard to copy
The complexity of partnerships, integrations, compliance, and workflows can create a moat that cannot be quickly replicated.
9. Financial institutions can use embedded partnerships to test innovation
For banks and regulated institutions, partnering with fintechs can allow small, managed pilots without large internal investment or heavy operational lift.
10. Customization can signal deeper partner commitment
While customization adds complexity, panelists noted that partners asking for tailored experiences may also be more invested in the success of the relationship.
Sign up for the CLA Finance Summit Series
