Composite case study based on patterns across our agency clients. Names, locations and specific figures have been anonymised; the financial mechanics and tax treatment are real and reflect current UK rules.

The situation

A specialist PPC agency based in Manchester’s Northern Quarter had grown fast. Founded in 2019 by two former in-house marketing managers, the firm had built a reputation for high-ROI Google Ads and Microsoft Advertising campaigns for ecommerce and B2B clients. By early 2025, the agency employed 14 people, 10 delivery staff, two account managers, a part-time bookkeeper, and the two founders. Revenue had climbed from £280,000 in 2022 to £620,000 in 2024, with a forecast of £850,000 for 2025.

The founders had used FreeAgent since launch. It was simple, cheap, and handled their basic invoicing, expense tracking, and VAT returns. But by late 2024, the cracks were showing. The agency now ran multiple client ad accounts through Klaviyo and Google Ads, with complex cost structures: media spend billed to clients, retainer fees, performance bonuses, and subcontractor payments for freelance PPC specialists. FreeAgent’s chart of accounts was flat, just 12 categories, and the bookkeeper was spending eight hours a week manually reclassifying transactions and reconciling client prepayments. The founders wanted something with proper project tracking, multi-currency support (they had two UAE clients paying in dirhams), and a chart of accounts that reflected how the business actually worked.

The decision

The founders decided to migrate from FreeAgent to Xero. They’d heard Xero handled multi-currency better, had stronger reporting, and integrated natively with Float for cash flow forecasting. The part-time bookkeeper recommended it, and a friend at another agency had made the switch with no issues. The founders gave the go-ahead in January 2025, aiming for a February go-live.

The problem? They didn’t plan the migration properly. They assumed it was a simple export-import job. They didn’t clean up FreeAgent first. They didn’t map the chart of accounts in advance. And they didn’t set aside time for the bookkeeper to test the new setup before going live.

What we modelled

When the founders came to us in March 2025, they were in a mess. The migration had gone live on 1 February, but within two weeks the bookkeeper found that 47% of transactions from January hadn’t mapped correctly. Client prepayments were showing as income. Media spend was lumped into a single “cost of sales” category. The multi-currency accounts for the UAE clients were showing negative balances because the exchange rates had been entered manually and incorrectly. The founders had lost three weeks of bookkeeping time and were facing a VAT return deadline in April.

We modelled three options:

  • Option A: Roll back to FreeAgent. This would mean reversing the Xero migration, restoring a backup from January, and re-entering February and March data manually. Cost: roughly £2,000 in bookkeeper overtime and a week of lost productivity. The agency would then need to re-plan the migration properly later.
  • Option B: Fix Xero in place. This meant rebuilding the chart of accounts from scratch, re-mapping all historical data, correcting the multi-currency entries, and reconciling the VAT return manually. Estimated cost: £4,500 in additional bookkeeper hours and two weeks of the founders’ time.
  • Option C: Full reset with professional support. We would clean up FreeAgent first (correcting the chart of accounts, removing duplicates, standardising supplier names), then re-export a clean data file, map the new Xero chart of accounts properly, and run a parallel test for one month before going live again. Estimated cost: £6,000 in professional fees and one week of the founders’ time, but with a guarantee of no data loss and a properly structured system going forward.

We also modelled the ongoing time saving. Under the old FreeAgent setup, the bookkeeper spent eight hours a week on manual reclassification and reconciliation. With a properly structured Xero chart of accounts, 35 categories, automated bank feeds, and project tracking for each client, that would drop to two hours a week. At the bookkeeper’s rate of £35 per hour, that was a saving of £210 per week, or £10,920 per year. Add in the time the founders saved not having to review dodgy reports, and the total management time saving was around £14,000 annually.

The outcome

The founders chose Option C. It was the most expensive upfront, but it solved the root cause, not just the symptom. We spent two weeks cleaning up FreeAgent. The chart of accounts had 18 duplicate supplier entries, 23 uncategorised transactions from 2023, and three client prepayments that had been sitting as income for over six months. We standardised everything, created a proper cost-of-sales breakdown (media spend, subcontractors, software tools, platform fees), and set up project-level tracking for each of the 11 active clients.

We then exported a clean data file, mapped it to the new Xero chart of accounts, and ran a parallel test for March. The bookkeeper reconciled both systems side by side. When the numbers matched, we switched over on 1 April, just in time for the VAT return. The return was filed on time, with no errors.

The financial result was immediate. The bookkeeper’s weekly time dropped from eight hours to two. The founders could now run a real-time profit-and-loss report by client, which they’d never been able to do before. They spotted that one client was actually loss-making once media spend and management time were properly allocated, they renegotiated the retainer and turned it into a 12% margin client within three months. The multi-currency accounts for the UAE clients now reconciled automatically, saving another hour a week.

By the end of 2025, the agency had saved £14,000 in management time. The £6,000 professional fee was recouped in under six months. The founders said the biggest lesson was that “a software migration is a business process change, not an IT job.”

What others can learn

  • Clean your data before you move. A migration is only as good as the data you take with you. Spend a week tidying up your chart of accounts, removing duplicates, and reconciling balances before you export. It saves weeks of pain later.
  • Run a parallel test. Don’t go live on day one. Run both systems side by side for at least one full month. Reconcile the numbers. Only switch over when they match perfectly.
  • Map your chart of accounts to your business model. If you’re a PPC agency, your cost of sales isn’t just “cost of sales”. Break it out: media spend, subcontractors, platform fees, software tools. A flat chart of accounts hides problems.
  • Budget for professional support. The £6,000 we charged was less than the cost of the lost bookkeeping time and the founders’ stress. A good accountant or Xero consultant will save you money in the long run.
  • Think about the ongoing saving, not just the upfront cost. The £14,000 annual time saving was five times the migration cost. If a proper setup saves you two hours a week forever, it’s worth investing in getting it right.