Month-end close automation for a multi-entity group in London (2026)

The verdict in three sentences

A group with 14 entities on 3 ERPs that closes in 12 working days reports too late to steer the next month. Automating intercompany reconciliations and recurring journals makes a 5-day close realistic, for a custom build of roughly 50,000 to 100,000 GBP in 2026, or a packaged close platform at about 25,000 to 35,000 GBP a year. The project takes 4 months and usually pays back within the first year.

Where the 12 days go

The CFO of a London group knows the pattern: the finance team spends the first week chasing trial balances from subsidiaries, matching intercompany invoices that never agree and keying the same accruals, prepayments and recharges every month. Consolidation only starts on day 8, and management accounts reach the board in the third week.

The 7 days come from removing low-value tasks while keeping human review on material variances.

Packaged platform or custom automation

In 2026 there are two routes. Close and reconciliation platforms (BlackLine, FloQast, Trintech and others) bring proven features but expect you to adapt your processes. Custom automation plugs into the 3 existing ERPs and mirrors the group's chart of accounts exactly.

For a group with unusual processes and 3 ERPs, custom often wins over 5 years. A group already on a single ERP is often well served by a packaged tool.

At Kolonell, close automation falls into our medium application tier (4,000,000 to 7,000,000 FCFA, or 6,100 to 10,670 EUR) or complex tier (7,000,000 to 15,000,000+ FCFA, or 10,670 to 22,870+ EUR), depending on the number of ERPs and rules. Monthly maintenance is 35,000, 75,000 or 120,000 FCFA (53, 114 or 183 EUR). See automation and AI.

A 4-month plan

The parallel run is essential: close one month with both processes to prove identical results before cutover, which your external auditors will welcome.

Data quality is the hidden cost driver. If entity charts of accounts are not mapped to a group chart, or intercompany counterparties use different codes in each ERP, budget two to three extra weeks for a mapping table owned by group finance. That table becomes the foundation for every later automation, from cash forecasting to statutory reporting, so it is worth doing properly. Ask also for a documented runbook so the team can operate the close if the original developer is not available.

Mini case study

Sarah, CFO of a services group in London, runs a team of 8 accountants tied up at 70 % during close. Saving 7 days a month frees 8 × 7 × 0.7 = 39 person-days a month, about 470 a year. At a loaded 400 GBP a day, that is worth 188,000 GBP a year, even if only part is redeployed to analysis. Against an 80,000 GBP custom build, payback comes in about 5 months, before counting faster board reporting.

FAQ

Do we need a new ERP first?

No, automation connects to the 3 existing ERPs through exports or APIs. An ERP replacement typically costs 10 times more and takes 12 to 24 months.

Will our auditors accept automated journals?

Yes, if every generated entry is logged and controls are documented. A timestamped audit trail and a one-month parallel run make their review easier.

What intercompany auto-match rate is realistic?

Around 80 % to 90 % of lines matched without intervention is a fair 2026 benchmark. The remaining 10 % to 20 % are shown with the variance and likely cause.

Can we start with one entity?

Yes, starting with 3 or 4 entities on the same ERP lowers risk. Extending to all 14 then takes 4 to 6 weeks.

Where does AI fit in?

It suggests matches for ambiguous lines and drafts variance commentary. Journals are still approved by an accountant, keeping 100 % human control.

Let's scope your project. Share your close calendar, ERPs and team structure, and we will scope the priority automations and a 4-month rollout with a parallel run. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.