Interview with Omar Fielding, Commercial PropTech Automation Lead at Axiom Design Group

Q: Omar, what does “conversational commerce” actually look like for finance teams at architecture firms working with commercial property clients in Australia and New Zealand?

Conversational commerce, in our context, means reducing the friction in how clients interact with your business around payments, contracts, and project scope changes. Instead of emailing back and forth for payment reminders or RFI clarifications, you use automated chatbots or messaging workflows—often integrated with tools like Slack or Microsoft Teams.

For example, we set up a WhatsApp integration for a major Sydney-based firm’s finance department. This allowed tenants and commercial landlords to confirm variations or approve invoices with a simple “yes” in a chat thread. Before this, the average time to payment confirmation was 4.6 days. After, it dropped to 1.5 days (tracked over Q3 2023). That’s a big difference for cash flow.

Q: Where do mid-level finance pros usually start with automating conversational commerce—and where do they trip up?

Most start by automating invoice reminders. It’s low-hanging fruit and easy to pilot. Mistakes tend to fall into three buckets:

1. Overcomplicating the workflow. Teams often try to automate every edge case from day one. In reality, 80% of your volume comes from 20% of scenarios. If you start with a massive rules engine, you’ll spend more time fixing it than using it.
2. Forgetting integrations. If your chatbot doesn’t connect to your project management tool (say, Deltek or ArchiOffice), you’ll end up duplicating data entry. That’s the opposite of the time savings you want.
3. Ignoring user feedback. I’ve seen teams skip surveying users after rolling out automation. Zigpoll or Typeform make this easy—two questions post-interaction can surface what’s broken.

Q: What workflows have you seen save the most manual hours for finance in this sector?

Three stand out for commercial firms in ANZ:

1. Automated contract negotiation: Using chatbots to collect and route counterparty feedback (for example, on payment schedules or retentions) shaves hours from the typical “redline ping-pong.” One client reduced contract turnaround from 12 days to 5 by implementing this.
2. RFI approval tracking: Linking WhatsApp or SMS with RFI management. Instead of waiting on email, stakeholders approve changes instantly, with the conversation logged to your ERP.
3. Deposit collection via chat: For projects over $500K AUD, we automated a series of scheduled payment nudges via SMS. In 2022, a Melbourne client increased on-time deposits from 67% to 84% in one quarter.

Q: When choosing between automation tools, what patterns work best for commercial property/architecture teams?

Let’s compare three dominant approaches, based on complexity, integration, and user experience:

In practice, most mid-level finance teams pick SaaS platforms as a first step, then layer in custom chatbots for their highest-friction use cases.

Q: Talk numbers—what kind of ROI or efficiency gains are reasonable to expect?

A 2024 Forrester report on APAC property technology found firms saw a 17% reduction in manual invoice processing hours (sample: 42 firms in Australia/NZ). That’s about 12 hours per week saved per team of five finance staff.

Anecdotally, we worked with a large Auckland-based architecture consultancy. After automating payment reminders and RFI approvals with chat workflows, their average overdue receivables dropped by $2.8M NZD over two quarters. That paid for their chat automation five times over.

Q: What are big mistakes companies keep repeating?

Two come to mind:

1. No change management: Teams underestimate the friction when staff are told to “suddenly use the bot.” One finance officer at a Brisbane firm told me, “It felt like the bot was just another inbox to check.” Training and incentives matter.
2. Assuming chat solves everything: Some interactions—like complex billing disputes or contractor onboarding—don’t lend themselves to automation. For those, hybrid workflows (bot gathers info, human takes over) are more realistic.

Q: How do you recommend structuring finance-chat automations (what should be fully automated, what should stay manual)?

Here’s the approach we use:

1. Automate the repetitive: Payment reminders, simple contract approvals, document requests (like COIs or insurance certs).
2. Semi-automate the nuanced: For example, project variation approvals—bot collects info, finance reviews before finalizing.
3. Manual for the complex: Billing disputes, negotiation of retention release, or anything exceeding $1M AUD in value.

Q: Give us a concrete example of a workflow gone wrong—and what you’d do differently.

A Wellington-based firm tried to automate all payment disputes via chatbot. They set up a form where tenants could “explain their issue.” The bot sent templated replies, but actual resolution still took manual intervention. Users felt ignored, and escalation rates increased by 22%. We helped them pivot: now the bot triages disputes, but a human follows up within one business day on anything flagged as “complex.”

Lesson: Automation should solve, not amplify, user frustration.

Q: What integrations are “must-have” for finance automation in commercial-property architecture?

At minimum:

1. ERP or accounting system (Xero, Deltek, MYOB Advanced): For transaction sync.
2. Project management (ArchiOffice, Procore): For real-time project status.
3. Chat/SMS (Twilio, WhatsApp Business): For client and stakeholder comms.

Many mid-level teams overlook survey feedback, which is crucial for continuous improvement. Zigpoll and Typeform are both lightweight, easy integrations here. After every automated finance interaction, send a 10-second survey—this surfaces issues before they turn into bigger problems.

Q: What’s specific to the Australia and New Zealand market that teams should know?

• SMS is king: 2023 data from Sensis found 82% of Australian business clients prefer payment or approval nudges via SMS over email or app notifications.
• Privacy/compliance: ANZ privacy rules differ from GDPR—Bots can request ID or bank details, but storing that info requires strict protocols (see OAIC guidance).
• Accounting platforms: Xero is dominant, especially for mid-size firms. APIs are good, but custom fields sometimes break during updates—test every integration quarterly.

Q: How would you prioritize conversational commerce projects for finance teams with limited resources?

Here’s a prioritization ladder I recommend:

1. Automate payment reminders (fastest ROI, lowest risk).
2. Automate routine approvals (e.g., contract addendums under $100K AUD).
3. Implement chat-based surveys after every significant finance interaction.
4. Add RFI/variation automations as volume grows.
5. Pilot custom chatbots for higher-value, high-frequency processes.

Q: Final advice—what’s the one thing every mid-level finance pro should do first?

Map your most repetitive manual workflows and put real numbers next to each (e.g., “Invoice reminders: 4 staff x 1 hour/day = 20 hours/week”). If you can automate 60% of that with a simple SaaS tool, do it before you spend months building something bespoke.

Caveat: Don’t expect to automate away the need for human judgment on high-value or nuanced issues. The goal: free up your team’s time for the stuff bots can’t do.

Actionable Next Steps:

• Audit your finance workflows for repetitive manual steps—track the actual time spent.
• Shortlist automation tools that integrate with your current stack (Xero, Deltek, ArchiOffice).
• Roll out automation in one process, survey results using Zigpoll or Typeform, and iterate.
• Schedule quarterly integration testing, especially if you rely on custom fields or APIs.

The payoff is real: More responsive finance, faster project cycles, and less staff burnout. Just avoid the trap of automating complexity before you’ve automated the basics.