AI is changing business. Relationships matter more than ever.

Customers are not waiting for businesses to catch up. They are already bringing their own tools to the interaction.

Gartner's survey of 3,566 B2B and B2C customers, conducted in February and March 2026, found customers were approximately three times more likely to use a third-party generative AI tool than a company-provided chatbot when resolving a service issue (Gartner, July 2026).

That is worth sitting with. Customers have not rejected AI. They have gone and used it somewhere else, on their own terms, before they contact you.

So by the time someone reaches a business, they often arrive better informed and with a sharper question than they would have had two years ago.

Australian Government guidance describes AI as already affecting how organisations operate across areas such as customer service, administration, software, stakeholder engagement and data handling (business.gov.au).

The practical consequence for most businesses is not dramatic. It is incremental.

Responses get faster. Quotes get turned around sooner. Information gets found more quickly. Preparation improves.

Which means being competent at the routine parts stops being a differentiator. It becomes the entry requirement.

Accounting is going to change significantly as AI improves.

A great deal of what an accounting business does involves gathering, checking, classifying and presenting information. Technology is getting better at all of it, and it will keep getting better.

We are not defensive about that. If software can prepare a reconciliation faster and more accurately than a person, it should.

The question is what the time gets spent on instead.

That is the design decision, and it is a choice rather than an inevitability.

An owner deciding whether to hire another employee can get the general information easily. Award rates, on-costs, superannuation, the rough annual cost.

That is not the hard part.

The hard part sounds like this:

• can the business actually carry the cost through a slow quarter?
• is the constraint really staff, or is it systems, pricing or collections?
• what happened the last two times the business hired ahead of demand?
• what else is the business planning in the next twelve months?
• what do the owners personally need the business to produce?

The same pattern applies to buying equipment, taking on commercial property, adding debt, changing owner distributions, opening a second location or acquiring another business.

Those questions require context about a specific business, a view on risk, and someone prepared to be accountable for the recommendation.

There is a clear line between automation that helps and automation that blocks.

In Gartner's August 2026 research, 50% of customers said their interactions were easier when companies used generative AI — and 87% said it was essential for companies to provide an option to reach a human agent when using it (Gartner, August 2026).

Both numbers matter. Customers want the speed. They also want a way through when the automated path does not resolve the issue.

Businesses get into trouble when automation is used to reduce contact rather than to reduce friction.

A useful test: does this save the customer time, or only save us time?

Good candidates for automation are usually the things nobody values:

• chasing documents and information
• scheduling and reminders
• status updates
• data entry and reconciliation
• preparation and first-pass analysis
• finding information that already exists somewhere in the business

Poor candidates are the moments where something is at stake — a decision, a risk, a mistake, a difficult conversation, a number that does not look right.

This sounds almost too simple to write down, and it remains the thing most businesses stop doing first when they get busy.

A short call before something becomes a problem is worth more than a well-written email after it.

It is also the interaction least likely to be replicated by a competitor with the same software.

Customers will usually tell a business what is not working if asked directly and given a real way to answer.

The value is not in collecting the feedback. It is in the response — changing something and telling the customer it changed.

Few things erode a relationship faster than a customer having to explain their own history back to the business.

Context is the asset. What the business does, who the owners are, what has already been tried, what was decided last year and why.

Used well, technology makes this better rather than worse. Information that used to sit in one person's memory can be available to the whole team before a meeting starts.

That is the point of it — not to remove the conversation.

When the routine parts of a service become uniform across the market, price becomes the obvious point of comparison — unless something else is clearly better.

That something is usually specific: the business is known, the advice is timely, someone picks up, problems get raised early, decisions get thought through properly.

Software can be bought by anyone. So can templates, dashboards and automated workflows.

What cannot be bought off a shelf is several years of understanding a particular business, its numbers, its seasonality, its owners and the decisions they have already made.

As the technology levels the routine, that accumulated context becomes a larger share of what a customer is actually paying for.

None of this is sentimental. Retained customers usually cost less to serve, refer more, and produce more predictable revenue than a constant cycle of replacement.

The commercial argument for relationships is not softer than the argument for efficiency. It is the same argument, measured over a longer period.

The order matters.

Decide what the relationship should feel like, then choose technology that supports it. Businesses that do it the other way around tend to end up with efficient processes and irritated customers.

This is how we think about the way we work — and why we build tools that help owners see the effect of a decision, rather than tools that replace the discussion about it.

Predictions that AI will do everything, and predictions that it will change nothing, are both likely to age badly.

The realistic version is less dramatic. More of the routine work becomes automated. Expectations rise. People concentrate on the work that requires judgement, context and accountability.

Businesses that use the technology to spend more time with customers will probably be in a stronger position than those that use it to spend less.

We are positive about technology and intend to keep using more of it — to reduce repetitive work, improve preparation, retrieve information faster, identify issues earlier and make processes easier for clients.

We intend to keep people involved where judgement is required, where an owner is working through a decision, where there is uncertainty, and where context and accountability matter.

That is a choice about how we design the service, not a claim about what the technology can or cannot eventually do.

You should know the people working on your business — that is why we publish our team — and the business advisory relationship should be close enough that context does not need rebuilding every time something important happens.

The goal is not to choose between technology and relationships. It is to use the technology to make the relationship better.