Stop Replacing. Start Multiplying.

IKEA built an AI chatbot called Billie. It handles 47% of customer inquiries. Most companies would have cut the call centre headcount and banked the savings.

IKEA retrained 8,500 call centre workers as interior design advisors instead. That new service channel now generates $1.4 billion in revenue, 3.3% of total company revenue, with a target of 10% by 2028.

The chatbot is not the story. The $1.4 billion is the story. And it only happened because IKEA asked a different question. Not "how do we replace these people?" but "what could these people do if they were not answering the same ten questions all day?"

That question is the difference between a cost saving and a multiplier.

The like-for-like trap

Most AI conversations start with a cost comparison. An admin in Perth costs around A$80,000 per year (base salary plus superannuation, leave, insurance, and equipment, per Glassdoor's 2026 data and Scale Suite's on-cost analysis). An AI subscription costs A$100 per month. The maths looks dramatic: a 98% reduction.

It is also the wrong frame.

Replacing an admin with a chatbot saves you A$80,000. That is a one-time gain. It does not grow. It does not compound. And it assumes the only value your people create is the admin work they currently do.

The Fyxer Admin Burden Index surveyed 5,000 workers in late 2025 and found that high earners spend 76 minutes per day on administrative tasks AI could handle. That is not junior staff on data entry. That is your operations manager, your project lead, your business development person, spending senior-salary hours on work worth A$25 an hour.

The question is not whether to automate the admin. The question is what those people do next.

Eliminate, Restructure, Automate

Before you automate anything, ask three questions in order.

Most businesses skip straight to automate. That is how you end up with an AI that does the wrong work faster.

What happens month by month

This is not a one-shot implementation. It is a progressive build.

Month 1: the assistant. AI handles the daily admin layer. Morning briefings, inbox sorting, meeting prep, draft documents. Your team gets 10 to 15 hours per week back. The setup is free. The subscription is A$100 per month. This is the trust-builder: your team sees AI work on real tasks before anyone commits to more.

Month 2 to 3: the diagnostic. Now that the admin layer is handled, the real question surfaces: where is the rest of the time going? An operations diagnostic (A$5,000 to A$7,000 over two to three weeks) maps how work actually flows through your business. Not what the org chart says. What actually happens. AI interviews your team, cross-references their answers, and surfaces what no single person can see: the bottlenecks, the duplicate work, the processes that exist because "we have always done it that way."

This is where Eliminate and Restructure happen. In my experience, the diagnostic typically finds that 20 to 30% of what a team does can be eliminated outright, and another 20 to 30% can be restructured before anyone writes a line of automation.

Month 3 onwards: the multiplier. With the admin automated and the processes cleaned up, an ongoing retainer (A$1,000 to A$3,000 per month) builds compounding automation. Customer service response times drop. Invoice chasing runs automatically. Quoting and reporting happen in minutes instead of hours. Each month adds a new capability, and each capability frees more capacity for revenue work.

This is the IKEA pattern at small business scale. The AI handles routine. Your people move to the work that grows the business.

The evidence

Deloitte Australia surveyed over 1,000 SMBs in 2025. Businesses that moved from basic to intermediate AI use saw a 45% increase in profitability. Those that went further saw 111%. Not cost savings. Profitability. Because freed capacity went to higher-value work.

The Future Business Academy tracked the same pattern across UK small businesses in 2025. A six-person accountancy practice freed 31 hours per week from client emails, meeting notes, and report drafts. They used the time to launch advisory services, generating £6,000 in new revenue within four months. A four-person marketing firm freed 26 hours per week and took on two additional clients worth £12,000.

In customer service specifically, Tediber cut response times from 72 hours to under one hour after automating 64% of support inquiries. The staff who had been answering those inquiries moved to higher-value customer work.

Nobody planned to "increase revenue with AI." They stopped wasting time. Good people, given time, find ways to grow a business. That is what I help businesses do.

IKEA turned 8,500 admin workers into a $1.4 billion revenue channel. You do not need 8,500 people. You need your best five, doing the right work.

The setup is free. The subscription is A$100 per month. The risk is two weeks of testing on one task. The upside is finding out what your team is capable of when they stop doing the wrong work.

• IKEA/Ingka Group: AI chatbot Billie, 8,500 workers redeployed, $1.4B new revenue (3.3% of total, 10% target by 2028)
• Fyxer Admin Burden Index (January 2026): 5,000 workers surveyed, high earners 76 min/day on avoidable admin
• Glassdoor + Scale Suite: Perth admin salary data + Australian employment on-costs
• Deloitte Australia, The AI Edge for Small Business (November 2025): 45%/111% profitability uplift, 1,000+ SMBs
• Future Business Academy, How Small Businesses Achieve 200% ROI with AI (November 2025)
• Tediber/Yuma AI: 64% automation, customer service response time 72hrs to under 1hr