Nearly every small business that brings in an AI tool like Claude discovers the same thing within a fortnight: the software works, and the rollout still stalls. The blocker is rarely the model. It is the people who have to change how they work. Knowing who resists, and why, is the difference between a tool that sits unused behind a login nobody remembers and one that quietly pays for itself.
The adoption curve is real, even in a ten-person business
In 1962 the sociologist Everett Rogers described how a new idea spreads through a population: a tiny group tries it first, a slightly larger group follows once the risk looks manageable, then two big middle waves adopt, and a final holdout group resists until the old way is taken off the table. The same shape appears inside a ten-person Sydney firm as in a national market. Two or three of your staff will open Claude the day you hand it over. A larger group waits to see whether those early users get burned. A final group digs in and keeps doing things the way they always have.
The mistake owners make is treating the whole team as one audience. A single all-staff email that says "we use AI now" lands completely differently on each group. The innovators are already three prompts deep and mildly annoyed you took so long. The holdouts read the same sentence as a threat. Here is roughly how the groups break down:
Innovators (about one in twenty). They found the tool before you did and are already using a personal account. Your job is to channel them, not to introduce them.
Early adopters (roughly 15 percent). Pragmatic and curious. They will commit if you give them one clear, real task where the tool obviously helps.
Early majority (about a third). They adopt once they watch a peer, not the boss, succeed with it. Proof beats persuasion.
Late majority (about a third). They move when the new way becomes the normal way and staying on the old path costs them effort.
Laggards (the remainder). They resist until the old process is genuinely retired. Some never fully convert, and that can be fine.
The four real reasons people resist
When a rollout stalls, owners tend to blame vague "change fatigue". That label hides four specific, addressable reasons. Name the right one and the resistance usually softens.
1. Fear the tool is coming for their job
This is the quiet one nobody says out loud in the meeting. If a bookkeeper hears that Claude can draft reconciliations, the first thought is not "great, less typing", it is "how many of me does this business now need". You cannot argue someone out of this with a cheerful slide. You address it by being direct about what the tool does and does not change, and by showing the same person the parts of their role the tool makes more interesting rather than redundant.
2. Scar tissue from the last software rollout
Most Australian SMBs have lived through a CRM or accounting migration that was sold as effortless and turned into six months of double entry. Staff who carry that memory are not being difficult. They are pattern-matching, and their pattern is that new systems mean unpaid overtime. The counter is a smaller promise kept: one workflow, one week, measured honestly.
3. No visible connection to their actual work
A demo about "AI for business" means nothing to the person who spends their day chasing overdue invoices. Show that same person Claude drafting the follow-up emails they hate writing, in their own tone, and the abstraction disappears. Resistance is often just a translation problem: the value was never shown in the language of that role.
4. Trust and accuracy concerns
This one is legitimate and should be respected, not dismissed. A professional who signs off on advice cannot paste in an answer they have not checked, and under the Privacy Act they are right to ask where client data goes. The healthy response is to teach people to treat Claude as a fast first drafter whose work is always reviewed, and to set a clear rule about what information is allowed into the tool. Staff who raise accuracy concerns are usually your most careful workers. Win them and the middle follows.
How to move the curve
You do not move the curve by pushing on the laggards. You move it by making the early wins visible and letting each group pull the next one along. A few tactics that work in a small Australian team:
Start with a volunteer, not a mandate. Pick one early adopter and one painful weekly task, and make that pairing succeed publicly.
Show peer proof, not owner enthusiasm. A two-minute clip of a colleague finishing a report faster does more than any policy.
Retire the old way on purpose. The late majority only moves when the previous process stops being an option, so set a date.
Write down what is allowed. A one-page AI use policy removes the ambiguity that anxious staff fill with worst-case assumptions.
Protect the sceptics' role in review. Give careful staff the job of checking outputs. It uses their instinct instead of fighting it.
It helps to put a number on the stall. A tool at $30 per seat per month across ten staff costs about $3,600 a year, which is trivial. The real cost of a rollout that never lands is the time it was meant to give back. If Claude reliably saves each person four hours a week, and your loaded staff cost sits near a $120,000 salary equivalent, an unused rollout is quietly burning far more than the licence fee every single month. The licence is the cheap part. The adoption is the asset.
Start with the willing, not the whole org
The businesses in Melbourne and Sydney that get real value from Claude did not convert everyone on day one. They found their two willing people, gave them a task worth winning, and let the result travel. The curve did the rest. Resistance is information: it tells you which of the four fears is in the room, and each one has a specific answer. Treat adoption as a people problem with a technical assist, not a technical problem with a people cost, and the tool starts earning its keep.
If you want a second pair of eyes on where your team sits on the curve and which resistance you are actually facing, we are happy to help. You can book a short AI solutions brainstorm and we will map it with you.
