IQ Books: Bookkeeping Software Built by a Bookkeeper

In 2024 I sat down to reconcile a client's PayPal account and found six thousand uncoded transactions waiting for me. Not six hundred. Six thousand. Every one needed a category, a VAT treatment, and a decision about whether it was stock, fees, refunds or the client's occasional habit of buying birthday presents on the business account.

I did what every bookkeeper does. I made a very large cup of tea, built a spreadsheet of keyword rules, and lost most of a week. Somewhere around transaction four thousand, a thought that had been forming for years finally landed: the software is not on my side.

Not because it was badly made. Because of how it was assembled. The ledger was one subscription. The receipt capture tool was another. The reporting layer that partners actually wanted to see was a third. The working papers lived in Excel, because nothing else would do them. Each tool was built by a different company, priced by a different finance team, and connected by integrations that broke quietly and often. The industry calls this an "ecosystem". From inside a practice at 11pm, it feels more like a toll road.

That week produced CodeIQ, the transaction-coding engine that became the first IQ Suite product. What it eventually produced — two years, four products and a great deal of double-entry pedantry later — is IQ Books.

The assembly problem: why "ecosystem" means four invoices

Look at what a small UK practice actually runs on top of its core ledger subscription. A receipt capture tool such as Dext or Hubdoc. A reporting and analytics layer such as Fathom or Syft. Something for working papers. Increasingly, an AI assistant of some description. Each solves a real problem — that is precisely the point. Coding, reconciliation, analysis and year-end are not exotic extras; they are the actual job of bookkeeping. Yet the mainstream platforms treat them as an aftermarket.

The market has noticed. In 2024 Xero acquired Syft Analytics, the reporting tool thousands of its users were already paying for separately. Sage had done the same with Futrli two years earlier. When the platform vendors spend hundreds of millions buying the add-ons their own customers depend on, they are conceding the argument: this functionality belongs inside the ledger. The catch is what happens next — acquired tools get folded into higher pricing tiers, and the capability you used to choose freely becomes a lever for moving you up a plan.

The opposite bet: one ledger, everything native

IQ Books is a full double-entry ledger — sales and purchase invoicing, VAT with the awkward codes included (reverse charge, postponed VAT accounting, domestic reverse charge for construction), multi-currency, bank imports, statutory reports, Making Tax Digital for VAT and income tax. On its own, that makes it a competent, honest bookkeeping system with an interface designed around how bookkeeping actually flows: money in, money out, review, reconcile, report.

What makes it unusual is that the four IQ Suite products are not integrations bolted to its side. They are built into the same ledger, the same database, the same screens:

CodeIQ — the coding engine, inside the drafts queue

Import a bank statement and every line lands as a draft. One click sends the lot through CodeIQ's eight-phase pipeline — transfer detection, invoice matching, your own historical patterns, a crowd-sourced merchant database, MCC categories, semantic analysis with a local embedding model, your personal corrections, and VAT classification. Suggestions land back on the drafts with confidence marks; anything the engine is unsure about is flagged for a human decision rather than posted quietly. Because CodeIQ shares the ledger, its suggestions reference your chart of accounts and its invoice matching sees your actual outstanding invoices — not a synced copy that was accurate last Tuesday. (For the full story of how the pipeline works, see how AI transaction coding actually works.)

ReconcileIQ — reconciliation as a tab, not a ritual

Every bank account in IQ Books carries a Reconcile tab. Drop the statement in and the engine compares it line-by-line against the posted ledger, then walks you through the differences with one-click, double-entry-safe fixes. Missing transaction? Added as a draft. Duplicate? Reversed properly, never deleted. The reconciliation is against the real ledger, so there is no export-import-export loop and no "which version is right" question at month end.

LedgerIQ — analytics on the ledger, not on an export

LedgerIQ reads the general ledger and produces the analysis layer practices normally pay a separate subscription for: trial balance through to cash conversion cycle, break-even, DuPont decomposition, credit risk, forecasting, board packs — forty-plus modules with drill-down to the underlying transactions. Inside IQ Books it opens against your live books for a chosen period. This is the category Xero valued highly enough to buy Syft for; here it is a menu item.

PrepIQ — year-end working papers from the ledger you already keep

At year end, PrepIQ takes the general ledger and trial balance straight out of IQ Books — no export, no re-keying — and produces lead-schedule working papers the way a practice would build them. The gap between "the books are done" and "the accounts are filed" is where small practices lose their January. Closing that gap from inside the bookkeeping software is the kind of decision only someone who has lived that January would prioritise.

RiQ — the copilot across all of it

RiQ is the assistant layer that runs through the suite: it reviews low-confidence codings and explains its reasoning, reads receipts and sales invoices into drafts, answers questions about the numbers with figures you can click to verify against the ledger, and drives the analysis modules conversationally. It is deliberately not a chatbot bolted onto the corner of the screen; it operates the same software you do.

Simplicity and capability are not opposites

There is a persistent belief in accounting software that you must choose: simple tools for people who find bookkeeping confusing, or capable tools for professionals, with nothing in between. I think that belief is a by-product of the assembly problem. When capability arrives via seven integrated products, of course the result is complicated — complexity is the integration tax made visible.

Build the capability into one coherent system and the trade-off mostly dissolves. IQ Books keeps the surface honest and small: a dashboard that reads like a briefing, a banking workflow that goes import → review → confirm, customers and suppliers that behave like a ledger should. The depth — the coding pipeline, the analytics, the working papers — stays out of the way until the moment it is needed, then appears as a button in the place you already are. Simplest interface, deepest toolkit, one login. That is the whole pitch.

The pricing is the other half of the argument

The IQ Suite runs on one subscription with a shared credit allowance that every product draws from — coding a transaction costs a few credits, scanning a receipt a few more, a full analytics session a defined block. Plans start at £5 per month, and every plan gets the entire platform: no feature gates, no "available on the Premium tier", no per-user surcharge for expenses. Additional organisations in IQ Books are £2 per month each, which is the sort of number that makes multi-entity owners check they have read it correctly.

Contrast that with the direction of travel elsewhere: Xero's UK prices rose again in September 2025 with a further increase already announced for September 2026, and QuickBooks raised its UK Plus plan by 47% in a single move in January 2026. The full breakdown deserves its own article — we wrote it here — but the summary is: the incumbents are charging more every year for the core, while the capabilities that matter live behind add-ons. We priced IQ Books the way a bookkeeper prices for their own clients: fairly, predictably, and without the annual letter.

Switching is one click — genuinely

The traditional reason nobody leaves their accounting software is not loyalty; it is conversion dread. Opening balances, contact lists, outstanding invoices, the wrong time of year. IQ Books has a built-in migration engine for Xero, QuickBooks, Sage and Pandle: connect, pick the company, pick your start date (it suggests the start of your financial year and explains why), and it brings across the closing trial balance as opening balances, your full customer and supplier lists, and every outstanding invoice at its original date so aged debtors and creditors carry over intact. It checks the numbers tie, shows you the result before and after, and the whole thing is undoable if you misclicked. We wrote about why switching became this hard and how we collapsed it to one click.

Where it is, honestly

IQ Books is in private beta. It is running real businesses' books in pilot — including businesses migrated in from the platforms above — while we finish hardening it in public view. The rest of the suite (CodeIQ, ReconcileIQ, LedgerIQ, PrepIQ) is live today and works with QuickBooks, Xero, Sage and Pandle, so you can put the automation to work on your existing ledger now and move the ledger itself when access opens. If you want in early, request access — pilot users get a real say in what gets built next, which is exactly how the product has been shaped so far.

Software built by the people who use it has a particular texture. It is opinionated about the things that waste your evening and relaxed about the things that don't matter. It says "review these nine, the rest are done" instead of presenting nine hundred rows and a search box. That texture is the whole reason IQ Books exists — because after enough 11pm coding sessions, building the tool you wished existed stops being a business plan and becomes a compulsion.