Finally Wanted to Finish the Books. The Books Outlasted Finally.

A customer arrived with a year’s worth of unfinished books and the wrong accounting software. In Felix Rodriguez’s telling, his co-founder was comfortable with QuickBooks and wanted no part of learning Xero for this particular account. He took the work himself. Several weeks later, the engineer had acquired a rather expensive education in repetition.

That early cleanup, recounted in a 2024 PayPod interview, offers a useful entrance to Finally. Selling somebody relief from administration is easy. Providing it means meeting the administration in person. The customer’s transactions will not become less numerous because the founder has an interesting pitch.

Rodriguez’s response was to pursue automation. He called the early AI effort Prometheus. Greek mythology is an ambitious place to shop for a bookkeeper’s name, but the job description was decidedly earthly: deal with the repetitive work. Finally’s original name, Back Office, was almost aggressively honest. The later name sold the sensation an owner hoped to experience when the job was finished.

The customer who came with homework

Felix Rodriguez, Glennys Rodriguez and Edwin Mejia built the company around a familiar mismatch. Running a business requires financial records; being good at the business does not automatically make its owner good at producing them. A dentist can know exactly what a patient needs and still dislike reconciling a bank account. A contractor can price a job and still spend Sunday looking for a receipt.

Founded in January 2018, Back Office became Finally as the proposition grew. By March 2022, the company said more than 1,000 businesses in the United States relied on it. Tempesta Media was among the named customers in its funding announcement. The service offered visibility across connected bank and credit accounts, with automation intended to reduce the manual effort behind that visibility.

The distinction mattered. A business owner did not have to wake up with a new enthusiasm for accounting software. Finally was selling the completion of an obligation. There is a large emotional difference between “here is your tool” and “someone is doing your books.”

A ledger with people attached

Finally’s bookkeeping offering connected bank and card feeds, imported transactions and used machine learning to organize them in an accounting file. Its product pages also described an in-house accounting team, catch-up bookkeeping, cash and accrual accounting, and support for QuickBooks and Xero. The human layer was part of the offer.

That made Finally something more specific than a replacement spreadsheet. It could work with a ledger a customer already knew. Its sales page described dedicated bookkeepers and CPAs, monthly financial statements and access to the underlying records. Automation supplied the repetitive work; a service relationship supplied the person responsible for dealing with it.

Its natural audience was the owner who wanted financial information without becoming the operator of every system producing it. The bookkeeping sales page said 90 percent of clients were small businesses or startups, ranging from tiny shops to businesses with more than $10 million in annual revenue. Treat that as the company’s description of its customer base, rather than an independent census.

“When vendor tools are scattered, business owners are too.”

Felix Rodriguez, February 2024 funding announcement

The line was a tidy statement of the problem. Buying several good tools can still leave an owner with the unpaid job of making them agree. Finally wanted to take responsibility for more of that agreement. Expertise in this setting means knowing the accounting system and getting usable records through it, month after month.

The receipt was the opening move

The company extended that logic to corporate spending. Its card offering included virtual and physical cards, individual limits and accounting integration. The Finally Visa Corporate Card was issued by The Bancorp Bank, N.A. Finally supplied the surrounding product and expense workflow; the issuing bank supplied the card relationship.

Its expense-management pitch focused on the nuisance after the purchase. Employees could be prompted for receipts after a swipe. Controls could limit where a card worked and how much it could spend. Transaction coding connected the purchase to the records that finance would later need to close.

This is where the comparison with a traditional bookkeeper becomes interesting. A bookkeeper can reconstruct a purchase afterward. A connected spending product can ask for the evidence while the purchase is still fresh. The design principle is worth copying: collect the explanation near the event, before remembering it becomes a separate project.

Finally occupied the overlap between managed accounting and spend management. A bookkeeper using QuickBooks or Xero addressed one part of the need; platforms such as Ramp and Brex addressed another. Finally’s wager was that a smaller business would prefer to have more of the work handled together. That is a positioning choice, rather than proof that every part of its bundle was better.

The expensive word was “all”

The historical prices reveal how broad the proposition had become. A separate bookkeeping sales page advertised a core plan starting at $299 a month. Standalone invoicing appeared at $12, $19 and $39 a month, with different client limits. The larger accounting and finance bundles required a quote and were billed annually.

Those figures bought different things. Twelve dollars for invoicing was not the price of having someone keep a company’s books. Nor was the starting bookkeeping fee a universal price for payroll, taxes, reporting and every other service. A buyer had to examine the actual scope, reporting frequency and included accounts.

The company’s own cost of expanding deserves the same care. It announced $95 million of combined equity and debt financing in March 2022, followed by $10 million in February 2024. September brought the two buckets shown above. Adding a credit product to a software-and-service business creates a financing requirement beyond paying people to write code or reconcile accounts.

PeakSpan’s 2024 investment thesis described a move toward bundling back-office services and credit. It treated bookkeeping as the opening to a wider customer relationship. The attraction was understandable: once a business trusted Finally with its records, related financial products had a plausible place to go. Each additional product also added another promise to keep.

The company ended. The assignment did not.

On September 8, 2026, Finally’s operator, Backoffice.co, Inc., entered an Assignment for the Benefit of Creditors. Finally’s website now says it is no longer operating. The remaining company affairs are handled by PFS Finally Liquidation LLC, an estate managed by restructuring firm Province.

The bookkeeping business took a different path. Inkle acquired it. Finally’s notice says the customer contracts moved to Inkle and bookkeeping service continues with the same team. For an owner, the distinction is practical: the company’s ending and the next monthly close are separate events.

Read the outcome at that level of precision. The bookkeeping transfer is not a statement that every former Finally product continues. An old card page, pricing page or app listing is a record of what was offered. It cannot turn a closed company into a current purchasing option.

What is worth copying

Finally’s useful idea was to begin with a job the customer needed completed. The receipt workflow followed the same rule. Start where information is created, reduce the repeated typing, and connect the result to somebody’s actual responsibility. A handsome dashboard is welcome; a reconciled account has a more immediate use.

Consolidation also asks something of the buyer. It makes sense when the provider can handle the business’s real accounting requirements and the owner wants to delegate. It becomes less convincing when the bundle adds services the customer does not need, or when a business requires specialist work outside the agreed scope. Convenience should be measured against the assignment, not the length of a feature list.

The portable lesson is to arrange access to the records while the relationship is healthy. Know who controls the ledger, what a completed month includes, how to retrieve the history and who takes over an unfinished task. These are questions for any outsourced bookkeeping relationship. Finally’s ending simply makes them harder to postpone.

A company called Finally promised the pleasure of having something finished. In the end, the most valuable thing to preserve was work that never really finishes: the next transaction, the next reconciliation, the next month’s books.