Small Melbourne-based accounting firm Tax Window improved how it follows up with clients by using AI. The result is less manual work and communication that feels more personal.
Automating personalised client follow-ups
How Tax Window helps clients feel seen and understood without adding more administrative effort
The challenge
Tax Window provides tax and advisory services to individuals and small businesses. Their accountants focus on personalised advice, but were spending a lot of time on manual administration that didn’t add real value.
After each initial client meeting, accountants sent a follow-up email with a quote. This process was manual, repetitive and prone to error, with emails copied from generic templates.
While the advice in meetings was tailored, the follow-up emails didn’t reflect it. Clients with very different needs received almost identical messages.
This created two issues:
- Accountants spent time on a task they didn’t enjoy.
- Clients couldn’t see the personalised service being delivered.
Tax Window had already experimented with AI in an ad hoc way and saw stronger engagement and better conversion rates. But scaling this across all clients wasn’t practical without technical support.
The approach
Tax Window partnered with SMEC AI through AI adopt centres. They improved outcomes by focusing on the problem first, rather than adding new technology.
After reviewing Tax Window’s needs and existing systems, SMEC AI built a custom automation using current tools. The solution creates personalised emails based on each client’s situation, without disrupting day-to-day work.
The team tested a proof of concept before rollout and got hands-on support during implementation.
The benefits
The automation delivered clear benefits for both the team and the clients:
- Follow-up emails reflect each client’s needs.
- Conversion rates have improved as clients feel understood and supported.
- Each accountant saves around 4 hours a week on manual admin.
The accountants can now spend that time on advisory work and client support.
Importantly, automation didn’t remove human touch – it strengthened it. The advisory value was always there, but AI made it visible in every interaction.
Video transcript
Automating personalised client follow-ups
Automating personalised client follow-ups
Hi, I'm Josh from Tax Window.
We’re a small Australian tax and accounting firm.
We look after individuals and small businesses, and we really focus on personalised advice.
Like a lot of small practices, we were spending way too much time on admin that wasn’t helping our clients.
For example—after the first meeting with a new client, we’d always send a follow-up email with a quote. But it was manual, pretty template-heavy, and easy to mess up.
So we’d have a really specific conversation with the new client in the room… but then the follow-up could look almost the same for two completely different clients.
This didn’t reflect the level of personalised service we pride ourselves on, and because the process was manual, chewed up staff time.
We’d played around with AI a bit and saw better engagement and conversion—but we couldn’t roll it out properly across the whole client base without some help.
That’s why we partnered with SMEC AI through the AI Adopt Centre Program—just focusing on the business outcome we were after, rather than thinking about getting any shiny new tech.
Working with our existing setup, we built an automation that drafts a personalised follow-up email based on the client’s situation— with no new tools, and no disruption to how we work day-to-day.
And the impact’s been immediate.
Each accountant gets back about four hours a week that used to go into admin.
We’ve put that time back into advisory work and being available for our clients. The follow-up emails now reflect what we talked about—and we’ve seen better conversion because clients feel understood.
For us, automation didn’t remove the human touch—it actually helped us show more of it.
If you’re looking at AI and wondering where to start, my advice is: pick one annoying admin job that happens all the time, start small, and test it properly.