SMBs now expect the platform they use every day to own their financial workflows - not just the transaction, but everything around it.
Something has shifted in what SMBs expect from the platforms they use. It used to be enough to offer a great core product - lending, payments, accounting, or vertical-specific tooling. That era is ending.
This guide is written for product and commercial leaders at SMB platforms. It covers the landscape honestly - including where building in-house still makes sense - and gives you the frameworks to make the right decision for your platform.
The expansion from single-service tool to full financial OS is not a new category - it is the next chapter of a pattern every successful SMB platform has followed.
To understand where SMB platforms are heading, it helps to trace where they came from. The pattern is consistent across verticals.
The first generation of modern SMB platforms won by doing one thing exceptionally well. Payments processors focused on conversion. Accounting software focused on bookkeeping. Lending platforms focused on underwriting speed. The value proposition was clarity.
As platforms matured, they began expanding horizontally - wallets, cards, FX, payroll, embedded payments. The goal: increase revenue per SMB and make switching more costly.
The frontier is now financial workflows - invoicing, AR, AP, reconciliation. What has changed is that this layer is no longer too complex or too costly to embed. Platforms that move here next have a structural advantage: SMBs embedded in a full financial OS do not switch. They cannot afford to.
Chapter 1 · Accounts Receivable
AR: what platforms need to know - and where they stall
Payment matching - step four - is where every in-house AR build eventually breaks, and where the difference between a production-quality product and a growing manual backlog is decided.
Accounts receivable looks straightforward on paper. In practice, the distance between invoice and payment represents months of engineering, significant compliance exposure, and - if the matching logic is wrong - a reconciliation backlog that grows indefinitely.
The full AR loop
- Invoice creation - natural language or structured input, correct tax application, e-invoicing format generation.
- Delivery - email, PDF, Peppol network, or customer portal. Delivery tracking matters for dispute resolution.
- Chasing and reminders - automated, tone-adapted, sent at the right intervals based on customer payment history.
- Payment matching - partial payments, multi-currency, reference mismatches. This is where every in-house build eventually breaks.
- Reconciliation - matching payments to invoices, posting to ledger, flagging exceptions.
- Edge cases - quotes, credit notes, partial payments, foreign exchange, disputed invoices, tax adjustments, early payment discounts. Each requires its own logic, its own compliance consideration, and its own reconciliation treatment.
Real-world example: neobanks adding invoicing
Platforms like Revolut Business, Tide, and Monzo Business have made invoicing a competitive battleground. The pattern is consistent: the UI gets built well. Then the compliance layer, the matching logic, or the chasing engine breaks.
A realistic in-house build is 6–12 months from scoping to production - and that assumes everything goes to plan. It also requires:
- A team who knows how to build it - invoicing, compliance, and reconciliation are specialist domains.
- Undivided focus - product teams are juggling multiple priorities. Financial workflows rarely sit at the top of the backlog consistently enough to ship on time.
- No material delays - a single compliance change, an edge case that breaks the matching logic, or a payment rail issue can add months.
- Ongoing maintenance - e-invoicing mandates change, payment formats evolve, tax rules update. This is a perpetual engineering commitment.
Most platforms that start this journey either ship late, ship a product that breaks at scale, or quietly deprioritise it twelve months in.
Chapter 2 · Accounts Payable
AP: why lenders and vSaaS platforms are moving here next
Platforms that own the AP layer see borrowing intent before anyone else - a bill arriving for £45,000 with net-30 terms is a real-time signal that the SMB may need a short-term facility.
If AR is about getting paid, AP is about paying - managing supplier invoices, scheduling payments, routing approvals, and ensuring nothing falls through the cracks.
Why lenders are adding AP
When an SMB needs to pay a large supplier invoice - equipment, seasonal stock, a contractor - they often need to borrow to do it. The AP workflow is a lending trigger. But the benefits go further:
- Better underwriting data - knowing which suppliers an SMB pays, on what terms, and how reliably gives a lender a richer picture than bank statements alone.
- SMBs managing costs on platform - when AP is embedded, SMBs no longer need to jump between their banking app, a spreadsheet, and a separate accounts tool.
- More time on platform - every workflow an SMB manages inside your product is time they are not spending in a competitor's.
- Better data on how SMBs operate - payment patterns, supplier relationships, cost structure, and cash flow timing all become visible.
Why vSaaS platforms want both
Vertical SaaS platforms built for specific industries are increasingly positioned to become the full financial OS for their vertical. Adding both AR and AP is what makes the platform genuinely indispensable.
Example - Construction vSaaS: A small construction firm - a 5-person contractor doing £800k a year - issues invoices to developers and main contractors (AR), manages payments to subcontractors, materials suppliers, and equipment hire companies (AP), and juggles both while running jobs with different payment terms, retentions, and staged milestone payments. A construction vSaaS platform that embeds both AR and AP becomes the single place where jobs, costs, and cash flow live together.
Regulation
The compliance clock: e-invoicing mandates, MTD and what's coming
An invoice is no longer just a PDF - in an increasing number of jurisdictions, it is a legally structured data object that must meet format requirements, be routed through specific channels, and submitted to government systems in real time.
Compliance used to be something platforms could treat as a future consideration. That window has closed.
EU e-invoicing landscape
- Italy - B2B e-invoicing via SDI mandatory since 2019. Every invoice must be in FatturaPA format.
- France - phased mandatory B2B e-invoicing underway, cascading to SMBs through registered PDPs.
- Germany - B2B e-invoicing mandatory from January 2025 for receiving structured invoices. Full transmission requirements follow through 2028.
- Wider EU - the ViDA directive mandates real-time digital reporting for intra-EU B2B transactions.
- Peppol - now the default interoperability standard across EU public procurement and increasingly adopted for B2B.
UK: Making Tax Digital
From April 2026, sole traders and landlords above £50,000 income must submit quarterly digital updates to HMRC. The threshold drops to £30,000 from April 2027. Platforms not MTD-ready will create friction for their SMBs at tax filing - and that friction becomes a churn driver.
Chapter 3 · Build vs. Embed
Build, buy or embed: an honest assessment
For most platforms at Series A to Series C, the build timeline and resource commitment represent an opportunity cost that significantly outweighs the control benefits.
There are three paths. We give each a fair hearing.
Path 1: Build in-house
There are genuine cases where building in-house is the right decision - at significant scale, with an existing compliance team, and where financial workflows are the primary thing you are selling.
- 12–24 months to reach a production-quality product - assuming no major mandate changes midway.
- Ongoing engineering cost that does not end at launch. Compliance mandates change. Every update is a sprint.
- The agentic layer requires years of transaction data to train well. You will not have that data on day one.
- Compliance ownership is perpetual.
Path 2: Off-the-shelf software
Buying an existing product solves the build timeline problem, but introduces a different one. Off-the-shelf tools are designed for end-users, not for embedding. The product experience is branded wrong, the data model is misaligned, and the API surface is limited.
Path 3: Embed via API (Dolfin)
The third path is to embed a purpose-built financial ops API designed from the ground up to sit inside another platform - white-labelled, agentic, and compliance-ready on day one.
| Build in-house | Embed with Dolfin |
|---|
| Time to market | 12–24 months | Weeks |
| Engineering cost | High - multiple sprints | Low - we do the heavy lifting |
| Compliance burden | Owned entirely by you | Handled by Dolfin |
| Ongoing maintenance | Perpetual internal cost | Included in partnership |
| Mandate readiness | Requires constant updates | Always-on compliance |
| AI / agentic layer | Years to train | Live on day one |
| Product quality ceiling | Limited by eng bandwidth | Best-in-class, compounding |
For every other platform - which is most platforms - the embed path delivers a better product faster, at lower cost, with compliance handled and an agentic layer that would take years to replicate.
Chapter 4 · Partner Evaluation
What to look for in an AR/AP API partner
These six criteria come from the questions that matter most once a platform is live - not just during the sales process.
Apply these six criteria to any shortlist:
| Criterion | What to look for |
|---|
| Agentic workflow support | Are you just adding more manual tools that SMBs are responsible for managing - or actually delivering on AI expectations by reducing the time and work required? |
| Compliance coverage | Does it handle e-invoicing mandates, Peppol, MTD, and VAT natively - or does compliance sit with you? |
| Payment rail flexibility | Can it route across open banking, card, and SEPA? Multi-rail matters as SMB payment needs vary. |
| Reconciliation depth | Does matching happen automatically, including partial payments, credit notes, and multi-currency? |
| Time to go-live | Weeks, not months. Ask for reference customers who can confirm integration timelines. |
| Commercial model | Is pricing usage-based, fixed, or revenue share? Make sure incentives align. |
Commercial Model
What the revenue model looks like
Financial workflows are not just a feature - they are a revenue layer that compounds over time across three distinct income streams.
Platforms that embed AR and AP successfully unlock three distinct income streams:
| Revenue stream | Range | Platform uplift |
|---|
| Subscription uplift - charge more for financial features | £10–30 per SMB/month | +£6,000/month |
| Payments margin - earn on every payment processed | 0.3–0.8% of GMV | +£20,000/month |
| Retention improvement - sticky product reduces churn | 20–35% reduction | Significant LTV uplift |
An example: a platform with 1,000 SMBs, 30% agent attach rate, £20/month average subscription uplift, and £200k average monthly payment volume generates approximately £312,000 in additional annual platform revenue.
The retention impact compounds further. SMBs deeply embedded in AR and AP workflows have a significantly higher cost of switching. A 20–35% reduction in churn has long-term LTV implications that dwarf the direct revenue numbers.
We build a custom ROI model for every platform we work with. If you want to see what the numbers look like for your specific SMB base, payment volumes, and growth trajectory, book 20 minutes with the team.
Continue reading
Enter your email to unlock the rest of the guide.
Use the form above to continue.
Use the form on the right to continue.