AI User Adoption: Winning Over Sceptical Teams

You Bought the AI. Nobody Uses It.

Here is a number that should alarm every Australian operations manager who has just signed off on an AI tool: companies spend 93% of their AI budget on technology and infrastructure, and only 7% on people-related initiatives like training and change management (HBR, November 2025).

The result is predictable. According to Deloitte Australia's November 2025 report, 66% of Australian SMBs now use AI in some form, but only 5% are fully enabled to realise its potential benefits. That is a wide gap between buying AI and getting value from it.

The people side of the rollout is the part that goes unfunded.

Having worked on large-scale technology rollouts at enterprise operations including BHP and Rio Tinto, where data platforms had to be adopted by hundreds of operators and analysts across multiple sites, I have seen this pattern play out repeatedly. The tools that stick are the ones with a serious adoption strategy behind them. Raw capability decides far less than people expect.

This guide gives you the practical, step-by-step playbook for getting resistant teams to actually use AI -- based on research from McKinsey, HBR, Deloitte, and Australia's Department of Industry.

Part 3 of 4 in our AI Launch Series. This post covers user adoption after you have verified quality and understood what makes AI launches different.

Why Your Team Resists AI (It Is Not What You Think)

The instinctive assumption is that employees fear losing their jobs. But Australian data tells a different story. The BizCover Australian Small Business AI Report (2025) found that 72% of employees view AI as an opportunity to enhance their roles, not replace them. Cost and job loss concerns are notably absent from the top barriers.

So what is actually going on?

Understanding the specific flavour of resistance in your team determines your entire adoption strategy. A skills gap requires training. A trust gap requires transparency. A leadership gap requires visible executive usage.

The Trust Deficit Is Real and Growing

Research published in HBR (November 2025) found that employees who receive hands-on AI training report 144% higher trust in their employer's AI tools compared to those who receive none. Workers with high trust save approximately 2 hours per week using the same tools as low-trust peers.

The training numbers run the other way. 30% of employees have received zero AI training, and 61% have spent less than five hours learning about AI (Slack Workforce Lab, 2024). Most businesses are asking people to trust tools they have never been properly taught to use.

The Adoption Framework That Actually Works

After studying the research and drawing from enterprise rollout experience, the adoption approach that works in Australian businesses follows a specific sequence. Skip a step and adoption stalls.

Research supports this sequence. McKinsey (2025) found that 48% of employees would use AI tools more often if they received formal training, and 45% would increase usage if AI were integrated into daily workflows. You need both -- and in that order.

Step 1: Pick the Right First Use Case

The first AI use case your team encounters will shape their attitude toward every subsequent one. Choose poorly and you create sceptics. Choose well and you create advocates.

The worst first use case is anything that requires people to change how they work before they see value. The best first use case delivers visible time savings within the first week, using a tool that lives inside software they already know.

Step 2: Build Your Champion Network

Research consistently shows that 10-20% adoption of any innovation triggers rapid acceptance by the majority (Strategico Consultants, 2025). You do not need everyone on board. You need the right 2-3 people.

What Makes a Good Champion

Champions are not necessarily your most technical people. They are:

• Respected by peers -- the person teammates check with before they commit to something
• Willing to be vulnerable -- open about their own learning curve, including the awkward parts
• In the workflow -- they do the actual work the AI tool touches, day in and day out

When champions model adoption in their own work and talk openly about what they struggled with, they create psychological safety for others to try (and fail, and try again).

The Champion Playbook

1. Identify 2-3 people who are curious (not necessarily tech-savvy)
2. Give them early access -- one week before everyone else
3. Let them customise -- ask them to find their own best use cases
4. Make them teachers -- have them show teammates in small informal sessions
5. Recognise publicly -- highlight their results (time saved, errors avoided) in team meetings

From enterprise experience: On large-scale data platform rollouts across mining operations, the teams that adopted fastest had a respected operator on the floor saying, "Look, I was sceptical too, but this actually saves me an hour a day." The quality of the training materials barely moved the needle. Peer influence beats PowerPoint every single time.

Step 3: Show, Don't Tell

Forget the corporate "AI awareness session" with 40 slides about machine learning. Research from HBR (November 2025) found that workers given hands-on practice opportunities are 72% more likely to report high trust in AI tools.

The Live Demo Formula

Instead of telling people what AI can do, show them -- on their own data, with their own tasks.

1. Take a real task from the team (an email they wrote yesterday, a report they compiled last week)
2. Run it through the AI tool in front of them -- live, not pre-recorded
3. Show the imperfect output -- do not cherry-pick the best result
4. Let the team critique it -- this gives them agency and control
5. Show the edit -- demonstrate how human review improves the output

This approach works because it addresses the three main objections simultaneously: it proves the tool is useful (skills gap), shows the output transparently (trust gap), and integrates with their actual work (workflow gap).

Step 4: Handling "AI Will Take My Job" Honestly

This conversation deserves a straight answer. Here is how to handle it with integrity.

What to say:

• "AI is going to change parts of your role. The repetitive parts. That is the honest truth."
• "What it will not do is replace the judgment, relationships, and problem-solving that make you good at your job."
• "The people who learn to use AI well will be more valuable, not less. That is the pattern across every industry."

What NOT to say:

• "Don't worry, nothing will change" (they will not believe you, and it is not true)
• "AI is just a tool, like a calculator" (this minimises their valid concern)

The reframing that works: Position AI adoption as a career skill. Employees who build AI fluency are more employable everywhere they go. That shifts the motivation from "do this for the company" to "do this for yourself."

The pattern is clear: AI handles the mechanical parts. Humans handle the judgment. The role shifts from "doer" to "reviewer and decision-maker," which is typically a more senior function.

Step 5: Training That Actually Sticks

McKinsey's 2025 research is unambiguous: when employees receive adequate training, they use AI tools more frequently as skill levels rise. But "training" does not mean a two-hour workshop and a PDF.

The Training Structure That Works

One number here matters more than the rest: weekly manager check-ins raise trust scores by approximately 60% (HBR, November 2025). The check-in signals that AI use is valued, supported and expected, so keep it well clear of compliance monitoring.

Practical Example 1: Rolling Out AI Email Drafting to a Sales Team

Consider a typical Australian business with an operations manager responsible for a sales team of 8. The team sends roughly 40 outbound emails per person per day. Each email takes 10-15 minutes to draft from scratch. The ops manager wants to introduce AI-assisted email drafting to reclaim time for actual selling.

The Setup

Tool selection: A generative AI assistant integrated into the existing email client (Microsoft Copilot in Outlook at $30/user/month, or a standalone tool like Superhuman AI). What matters most is that the tool lives where the team already works, with no new logins and no new tabs.

Budget: For 8 users at approximately $30/month each, total cost is $240/month ($2,880/year).

Week-by-Week Adoption Plan

Week 1 -- Champion Selection and Early Access The ops manager identifies two champions: one top performer who is naturally curious, and one mid-performer who is vocal and respected. Both get access a week early with one instruction: "Try using it for your actual emails this week. Note what works and what doesn't."

By Friday, both champions have found their groove. The top performer uses it for prospecting emails. The mid-performer discovered it is brilliant for follow-up sequences after demos.

Week 2 -- The Live Demo (Not a Training Session) The ops manager books a 30-minute team meeting. No slides. Instead, one champion opens their laptop and shares their screen:

• Pulls up a real prospect they are emailing
• Types a brief prompt: "Follow-up email after demo of inventory management software. Prospect was interested in reporting features but concerned about cost."
• The AI generates a draft in 10 seconds
• The champion edits it live, saying "See, it got the tone slightly wrong here -- too formal for this prospect. I'd change this line."
• Sends the email

The team sees the real output, including the imperfections, and watches the edit process. This takes 2 minutes per email instead of 12. The visual impact is immediate.

Week 3 -- Paired Practice with Low Pressure Each team member gets access. The ops manager pairs each person with a champion for their first three emails. No mandate to use it for everything -- just "try it for three emails today and see what you think."

Common pushback at this stage:

• "It doesn't sound like me" -- show them how to add tone instructions ("Write in a casual, direct Australian style")
• "It got a fact wrong" -- emphasise that AI drafts always need human review. This is editing, not blind automation
• "It takes longer to fix the draft than to write it myself" -- this is normal for the first 5-10 uses. The learning curve is real but short

Week 4 -- The Daily Challenge and Measurement The ops manager sets a light challenge: "Use AI to draft at least 5 emails per day this week." No punishment for missing it. At Friday's sales meeting, they share early results.

Weeks 5-8 -- Embed and Expand By week 5, most of the team is using AI for at least half their emails. The ops manager updates the team's email SOP: "Step 1: Generate AI draft. Step 2: Personalise and review. Step 3: Send."

The two holdouts (there are always 1-2) get individual attention. The ops manager asks: "What specifically is not working for you?" Often the answer reveals a solvable problem -- the tool does not handle their specific email type well, or they need a different prompt approach.

Practical Example 2: AI Inventory Forecasting for a Procurement Team

Consider a warehouse manager at an Australian wholesale distribution business. The procurement team of 4 currently uses spreadsheets to forecast demand, placing orders based on gut feel and last year's numbers. Stockouts cost the business roughly $15,000/month in lost sales. Overstocking ties up $200,000 in working capital.

The Setup

Tool selection: An AI-powered demand forecasting module that integrates with their existing inventory system (such as Cin7 with forecasting add-ons, or a standalone tool like Inventoro that connects to Xero/MYOB). The tool analyses historical sales, seasonality, lead times, and supplier reliability.

Budget: Approximately $300-800/month depending on SKU count and features.

The Trust-Building Approach

Procurement teams are often the most sceptical because they have deep domain knowledge. They know their suppliers, their seasonal patterns, their weird one-off customers. Telling them "the AI knows better" is a guaranteed way to kill adoption.

Phase 1: Shadow Mode (Weeks 1-2)

The warehouse manager does not replace the existing process. Instead, AI runs in parallel.

• The procurement team continues placing orders exactly as they always have
• Each week, the warehouse manager shows them what the AI would have recommended
• The comparison is purely informational -- "Here is what you ordered. Here is what the AI suggested. Let's see who was closer."

This is non-threatening. The team keeps full control. But they start seeing patterns: "The AI caught that seasonal spike two weeks earlier than we did."

Phase 2: Advisory Mode (Weeks 3-4)

Now the team receives AI forecasts before they place orders. They can accept, modify, or reject each recommendation. The key rule: no one is forced to follow the AI's suggestion.

The warehouse manager tracks three metrics:

1. How often the team accepted the AI recommendation
2. How often the AI was more accurate than manual forecasting
3. How many stockouts and overstock events occurred

Phase 3: Assisted Mode (Weeks 5-8)

Based on Phase 2 data, the team and warehouse manager review results together. Typically, the pattern looks something like this:

At this point, the procurement team has seen the evidence with their own data. The conversation shifts from "Do we trust the AI?" to "Which categories should we let it handle automatically, and which do we want to keep reviewing manually?"

Phase 4: Ownership (Months 2-3)

The procurement team now owns the AI tool. They decide which product categories get automated ordering and which get human review. The warehouse manager's role shifts from "driving adoption" to "supporting the team's decisions about how to use the tool."

The critical factor that made this work: nobody told the team to trust the AI. The warehouse manager showed them the data and let them reach their own decision. Autonomy builds trust faster than mandates.

The Manager's Weekly Adoption Checklist

Use this checklist during the first 8 weeks of any AI rollout.

What to Expect: Realistic Adoption Curve

Do not expect 100% adoption in week one. Research and practical experience suggest the following timeline for most Australian businesses.

Note: 100% adoption is not the goal. 80-90% sustained usage is a realistic and excellent outcome. The remaining 10-20% often have legitimate reasons (different role requirements, accessibility needs) that should be respected.

The ROI of Getting Adoption Right

The gap between buying AI and actually using AI is where Australian businesses are leaving the most value on the table. Deloitte Australia (2025) found that moving from basic to intermediate AI use delivers a 45% increase in profitability, and moving from intermediate to fully enabled use delivers a 111% increase.

The difference between a wasted AI investment and a transformative one comes down to whether your team actually uses it, and that depends entirely on how you roll it out.

Your Action Plan This Week

1. Audit your current state -- ask your team anonymously: "What stops you from using [AI tool] more?" The answers will tell you which resistance type you are dealing with
2. Identify 2 champions -- look for curiosity and peer respect ahead of technical skill
3. Pick one high-volume, low-stakes task as your first use case. Email drafting, meeting summaries, and data extraction are proven starting points
4. Schedule a 30-minute live demo (not a training session) within the next two weeks
5. Read the next post in this series on measuring AI success at 30-60-90 days to set up your tracking framework

Need help building an adoption strategy for your specific team? Book a free 30-minute consultation and we will map out a practical plan based on your tools, team size, and industry.

AI Launch Series Navigation

This is Part 3 of 4 in our series on launching AI successfully in Australian businesses:

1. AI Quality Verification: Ensuring Accuracy Before and After Launch
2. What Makes Launching AI Different From a Traditional Feature Launch
3. AI User Adoption Strategy: Driving Adoption Among Skeptical Teams (you are here)
4. Measuring AI Success: The 30-60-90 Day Framework

Related Reading:

• Your IT Team Isn't the Problem. Your AI Strategy Is. - Explores why AI projects fail due to strategy, not technology -- directly relevant to adoption challenges
• 7 AI Quick Wins You Can Run This Quarter - Practical first AI use cases with implementation steps, ideal for the "pick your first win" strategy
• Getting Started with AI: The Complete Guide for Australian Business Owners - Foundation guide for owners beginning their AI journey
• AI ROI Calculator: How to Justify Your First AI Project in Australia - Framework for building the business case that supports adoption investment

Sources: Research synthesised from Deloitte Australia "The AI Edge for Small Business" (November 2025), Harvard Business Review "Overcoming the Organizational Barriers to AI Adoption" (November 2025) and "Workers Don't Trust AI" (November 2025), McKinsey "Reconfiguring Work: Change Management in the Age of Gen AI" (2025), Australia Department of Industry AI Adoption Tracker (Q1 2025), BizCover Australian Small Business AI Report (2025), Slack Workforce Lab (2024), and Strategico Consultants change champion research (2025).