There are roughly 60 million freelancers in Europe and over 70 million in the United States. A growing percentage of them use neobanks or digital banking platforms as their primary business account. They send invoices from the app. They track expenses in real time. They categorise transactions, manage cash flow, and pay suppliers, all from their phone.
Then, once a quarter or once a year, they hit a wall. The banking app that handles everything else cannot help them file their tax return. They export their data, open a different piece of software, or forward everything to an accountant who manually re-enters the information they already had inside their banking app.
This is the last mile problem in freelancer banking. The data is there. The user is there. The infrastructure to compute and file a return is not.
The Workflow That Everyone Accepts (But Shouldn't)
Consider the typical journey of a freelancer using a modern neobank. Throughout the year, every transaction flows through the app. Income is tagged. Expenses are categorised. VAT is previewed or estimated. Some platforms even set aside a percentage of income into a "tax pot" or savings bucket, helping the user provision for their upcoming liability.
But when it is time to actually file, the workflow breaks. The user downloads a CSV or connects to an accounting tool. The accounting tool may or may not correctly map the bank's transaction categories to tax-relevant categories. The user or their accountant then manually reviews everything, makes adjustments, computes the return, and submits it through a government portal.
The data has left the building. The neobank, which had a perfect view of the user's financial life for 12 months, is no longer involved. The value of all that categorisation and tracking is diminished because it feeds into a manual process that the neobank does not control.
This is not just an inconvenience for the user. It is a structural problem for the neobank's product and business model.
Why This Matters for Neobanks
The first issue is churn. Tax season is the moment when a freelancer is most likely to re-evaluate their financial tools. They are forced to engage with their accounting stack, and if that stack is disconnected from their banking app, they start to question whether their banking app is actually saving them time. Competing platforms that offer more integrated financial services become attractive.
Research from various fintech industry analyses consistently shows that the more financial services a user engages with on a single platform, the lower their churn rate. A user who banks, invoices, and manages expenses on the same platform has meaningful switching costs. A user who only banks on the platform, then goes elsewhere for tax and accounting, has very low switching costs.
The second issue is revenue leakage. When the freelancer's transaction data leaves the neobank and enters the accounting ecosystem, the accountant or software vendor captures the filing revenue. The neobank facilitated the entire year of financial activity but monetises none of the compliance event at the end. For a freelancer paying between 200 and 1,500 euros per year for tax preparation and filing, that is revenue the neobank is leaving on the table.
The third issue is data. When an external accountant or software tool processes the user's transactions for tax purposes, they generate enriched financial data. They determine which expenses are deductible, what the effective tax rate is, how income splits between different revenue streams, and what the user's true profit margin looks like. This enriched data never flows back to the neobank. It sits in the accountant's files or in a separate software system.
If the neobank retained this data by owning the tax filing process, it could use it to offer better lending decisions, more relevant insurance products, smarter cash flow forecasting, and more personalised financial advice.
What "Solving the Last Mile" Looks Like
Solving this problem does not mean the neobank becomes an accounting firm. It means the neobank embeds tax computation and filing into its existing product, the same way it embedded invoicing and expense tracking.
From the user's perspective, the experience should be seamless. They open their banking app. They see a prompt that says their tax return is ready for review. They confirm a few details. The return is filed. They get a confirmation. They never leave the app.
From the neobank's perspective, this means connecting to infrastructure that can take the user's transaction data, apply jurisdiction-specific tax rules, route the return to a qualified professional for review, and submit it to the tax authority. The neobank does not need to understand tax law. It does not need to hire accountants. It needs an API that takes data in and produces filed returns.
The professional review component is essential and often misunderstood. Tax filing is not like payment processing, where software can execute the entire workflow autonomously. In almost every jurisdiction, a tax return carries legal liability. A licensed professional must review and approve it before submission. This requirement cannot be engineered away. It is a legal and regulatory constraint, not a technical one.
The Data Advantage
Neobanks are uniquely positioned to solve this problem because they already have the data. A traditional accountant receives a shoebox of receipts, a bank statement PDF, and a vague recollection of what the client earned last year. They spend hours organising, reconciling, and categorising before they can even begin computing the return.
A neobank has 12 months of structured, categorised transaction data. Every payment has a date, amount, currency, counterparty, and often a category. Many neobanks have already applied machine learning to classify transactions into income and expense categories. Some have sub-accounts or pots that separate business from personal spending.
This data, when fed directly into a tax computation engine, eliminates the most time-consuming and error-prone step in the traditional accounting workflow: data entry and reconciliation. The tax engine receives clean, structured, pre-categorised data. The professional reviewer can focus on judgment calls, such as whether a particular expense is genuinely deductible, rather than on data processing.
The result is faster filings, fewer errors, and lower cost per return. All of which translate into a better user experience and higher margins for the neobank.
Why Neobanks Haven't Done This Yet
If the opportunity is so clear, why haven't more neobanks already closed the last mile gap?
The primary reason is complexity. Tax law is not a single system. It is hundreds of systems, one per jurisdiction, each with its own rates, brackets, deductions, exemptions, filing deadlines, and submission formats. A neobank operating in five European countries would need to build and maintain five separate tax engines. One for French IR and TVA. One for German ESt and USt. One for Italian IRPEF and IVA. One for Spanish IRPF and IVA. One for Dutch IB and BTW.
Each of these engines requires deep domain expertise to build correctly. Tax law changes every year, sometimes multiple times per year. Maintaining accuracy requires constant monitoring of legislative changes, regulatory guidance, and government API updates.
The secondary reason is licensing. As noted above, filing a tax return on behalf of a user requires professional credentials in most jurisdictions. A neobank would need to either hire licensed professionals in each country or partner with existing accounting firms. Both approaches involve significant operational complexity that is outside the neobank's core competence.
The tertiary reason is liability. Tax errors can result in penalties for the taxpayer and, in some cases, for the preparer. A neobank taking on filing responsibility also takes on the risk of errors in a domain that is not its area of expertise.
These barriers are real, but they are not barriers to embedding tax filing. They are barriers to building tax filing in-house. The API model exists precisely to let neobanks offer the capability without owning the complexity.
The Market Opportunity
The numbers are compelling. In the EU alone, there are approximately 30 million sole traders and freelancers. In the UK, 4.3 million. In the US, over 70 million. These are people who file tax returns every year. Many of them use neobanks or digital banking platforms. Very few of them file taxes through their banking app.
At an average filing cost of 300 to 500 euros per return, the addressable market for embedded tax filing in Europe alone is measured in billions of euros annually. Even capturing a small fraction of this market through revenue-sharing arrangements with neobank partners represents a significant business.
For neobanks, the opportunity is not just the direct revenue from filing fees. It is the cumulative impact on retention, engagement, and data enrichment. A neobank that owns the full financial lifecycle of a freelancer, from first invoice to filed return, has a fundamentally stronger product and a fundamentally more defensible market position than one that hands the user off to a third party at tax time.
Closing the Gap
The last mile in freelancer banking is not a technology problem. The technology exists. APIs can move transaction data securely between systems. Tax computation engines can apply jurisdiction-specific rules to structured data. Government portals accept electronic submissions.
The missing piece is the connection between these components. Neobanks have the data. Tax infrastructure providers have the computation and filing capability. Licensed professionals have the credentials to review and sign.
The platforms that connect these pieces first will own the freelancer financial experience end to end. Everyone else will continue exporting CSVs.
Michael Cutajar, CPA — Founder of Accora.