Blog Details
Your Team Is Spending 40% of Its Week on Work a Machine Should Do
Here’s an exercise we run with almost every client in our first workshop. We ask each team member to write down everything they did yesterday. Not the job description version — the honest version.
The lists are always the same. Copying data from emails into the CRM. Chasing someone for a document. Formatting a report that looks exactly like last month’s report. Answering the same customer question for the eleventh time. Writing a follow-up email that’s 90% identical to every other follow-up email.
Then we ask one question: which of these tasks actually needed you?
The silence is usually the answer.
Industry research has put the number on repetitive, automatable work at anywhere between 20% and 40% of the average knowledge worker’s week. In our own client engagements, we’d say that’s conservative for admin-heavy businesses. A five-person operations team doing 30% repetitive work is effectively paying 1.5 salaries for tasks that software handles better, faster, and without sick days.
The shift that changed everything is that automation no longer requires rigid rules. Old-school automation broke the moment an invoice arrived in a slightly different format. Modern AI — models like Claude and GPT — can read. They handle messy emails, scanned documents, half-complete forms, and ambiguous requests the way a competent junior employee would, except instantly and around the clock.
What does this look like in practice? An incoming enquiry gets read, classified, logged in the CRM, and answered with a personalised draft before anyone opens their inbox. A supplier invoice gets extracted, matched against the purchase order, and queued for one-click approval. A weekly management report writes itself from your live data on Friday afternoon.
None of this is futuristic. All of it is deployable in weeks.
The honest caveat: automation done badly creates more work, not less. Automating a broken process just makes the mess faster. That’s why we always start with a workflow audit — mapping what your team actually does, finding the highest-ROI candidates, and automating those first. Quick wins build trust; trust builds appetite for the bigger transformation.
Here’s an exercise we run with almost every client in our first workshop. We ask each team member to write down everything they did yesterday. Not the job description version — the honest version.
The lists are always the same. Copying data from emails into the CRM. Chasing someone for a document. Formatting a report that looks exactly like last month’s report. Answering the same customer question for the eleventh time. Writing a follow-up email that’s 90% identical to every other follow-up email.
Then we ask one question: which of these tasks actually needed you?
The silence is usually the answer.
Industry research has put the number on repetitive, automatable work at anywhere between 20% and 40% of the average knowledge worker’s week. In our own client engagements, we’d say that’s conservative for admin-heavy businesses. A five-person operations team doing 30% repetitive work is effectively paying 1.5 salaries for tasks that software handles better, faster, and without sick days.
The shift that changed everything is that automation no longer requires rigid rules. Old-school automation broke the moment an invoice arrived in a slightly different format. Modern AI — models like Claude and GPT — can read. They handle messy emails, scanned documents, half-complete forms, and ambiguous requests the way a competent junior employee would, except instantly and around the clock.
What does this look like in practice? An incoming enquiry gets read, classified, logged in the CRM, and answered with a personalised draft before anyone opens their inbox. A supplier invoice gets extracted, matched against the purchase order, and queued for one-click approval. A weekly management report writes itself from your live data on Friday afternoon.
None of this is futuristic. All of it is deployable in weeks.
The honest caveat: automation done badly creates more work, not less. Automating a broken process just makes the mess faster. That’s why we always start with a workflow audit — mapping what your team actually does, finding the highest-ROI candidates, and automating those first. Quick wins build trust; trust builds appetite for the bigger transformation.
