There is a specific moment in the monthly close that nobody puts in a case study. You have paid for a reporting tool. It is a good tool. It connected to Xero or QuickBooks in about four minutes and the P&L came through perfectly.

And then you open Excel anyway, because the other four files won't go in as they are.

That step — the reformatting you do before the tool that was supposed to remove the work — is worth timing once. For most people it is the majority of what's left.

This is true of every tool in the category

It is easy to assume it is your particular setup. It isn't. Look at what the products actually document:

Fathom accepts Excel, and requires you to restructure the file into its own import template first. Account classification codes are mandatory and specific to Fathom. Non-financial KPIs have to be created inside Fathom before you can upload any results for them. Formulas don't come across at all — values only.

Joiin goes further than most: Xero, QuickBooks, Sage, CSV uploads, an Excel add-in, an API. And it asks you to map the chart of accounts by hand.

Spotlight Reporting, Reach Reporting and LiveFlow all anchor on the ledger connection and treat the spreadsheet as the fallback path — the one where you conform to the shape they expect.

The pattern across the whole shelf is the same: the ledger is connected, and everything else is your problem.

It is not laziness, it is arithmetic

The temptation is to read this as neglect. It isn't. Every integration is engineering that has to be justified by the customers it unlocks, and the maths is brutal.

Xero has millions of subscribers, so everybody integrates Xero. A regional POS with four hundred sites, an industry-specific ERP, a wholesale portal, a client's hand-maintained workbook — those will never clear the bar, because the number of customers they unlock will never pay for the build.

So what do you do about sources you can't afford to integrate? You publish a template and ask the customer to conform. It is the rational answer. It converts an engineering cost you can't justify into a manual cost somebody else absorbs.

Every company in the category reached that conclusion independently, which is usually a sign the conclusion is correct — from their side of it.

Which leaves the work exactly where it was

Here is the consequence, and it is worth being precise about.

Reporting tools removed the assembly work for the sources that were already easy. For the sources that were never going to get a connector, they removed nothing. The client on QuickBooks got faster. The client on the strange system still takes all afternoon.

Which is the wrong way round, because the client on the strange system was always the expensive one.

There is a second cost that shows up later. A conforming step done by hand every month is a step that can be done slightly differently every month — a column pasted one over, a total row included this time and not last time, a date misparsed in one direction. The tool receives your reformatted file and computes it perfectly. It has no way of knowing you handed it something subtly different from last month's version, so the comparison it produces looks clean and is wrong.

How to tell how much this is costing you

Two questions, and they take about a minute.

Of the files that go into your monthly pack, how many can your tool pull in automatically today?
For the rest — how long between downloading the export and having something the tool will accept?

If the answer to the first is "most of them," your tools fit your clients and this article isn't about you. If the answer is "the ledger, and that's it," then multiply the second number by your client count and you have the size of the gap.

What actually removes it

Not another integration. The list of sources nobody will build for is effectively infinite, and a tool that connects to four hundred systems still won't connect to your client's spreadsheet.

The fix has to work on structure rather than on a list of supported apps:

Read the file as it is — headers on row seven, merged cells, totals at the bottom, numbers stored as text, four date conventions in one column.
Work out what it means once, in terms a person can confirm: this column is the net amount, this row is a subtotal and not a transaction, these two files join on invoice number.
Freeze that understanding as code, so next month is computed identically rather than re-interpreted — which is the difference between a comparison and two unrelated answers.
Hand the result wherever it needs to go, including straight into the reporting tool you already pay for.

That last point matters more than it sounds. The conforming step and the report are separable. If something else does the reformatting properly and hands your existing tool clean, mapped, consistent data, you keep your templates, your branding and your process — and the Excel step in front of them stops existing.

The uncomfortable version

If you are paying for a reporting tool and still opening a spreadsheet before you can use it, you are paying for the easy half of the job and doing the hard half yourself.

That is not an argument against the tool. It is an argument about where the remaining work actually lives — and it has been sitting in the same place, in public, in everybody's help documentation, for years.

Sources: Fathom — importing financials from Excel, Fathom — non-financial KPIs, Joiin, Spotlight Reporting, Reach Reporting. Product details accurate as of August 2026.

Why you still open Excel before your reporting tool