How a single automation recovers 170 hours a year

The problem

A firm like this has often been running for years, with a workflow that works, just not efficiently. Every new job starts with a message on WhatsApp. Someone then copies the job details into a shared spreadsheet, and when the job is complete, creates an invoice in FreeAgent by copying the same details again.

That takes roughly 15 minutes per job. At 14 jobs a week, that is 3.5 hours every week on pure data entry, rarely questioned because it has always been that way.

What the audit found

A workflow session typically makes clear that all three tools (WhatsApp Business, Google Sheets, and FreeAgent) have API access available. The data entry happens manually only because no one has connected them. There is no technical barrier; it simply has not been set up.

An audit report identifies this as the highest-priority item: high frequency, low technical complexity, with the manual step fully removable.

What gets built

A Make.com scenario that watches the WhatsApp Business inbox for messages containing job keywords, parses the job details, creates a new row in the spreadsheet, and drafts an invoice in FreeAgent ready to review and send. The whole scenario runs in under 30 seconds from message receipt.

You review each invoice before it goes out: the automation drafts, a human approves. That keeps you in control of what every client sees.

What changes

3.5 hours a week becomes roughly 20 minutes of review time: around 170 hours a year back, the equivalent of four full working weeks. At a conservative admin rate of £20/hour, that is about £3,400 in recovered time a year, against a £500 audit.

Once it is live, a scenario like this runs without manual intervention. When a new job type comes up, you update the keyword list yourself.