AI Tips for Strata Professionals: Testing AI Against Professional Strata Budget Expertise

Each week we bring you one low cost, easy to use tip on how we are helping our clients use AI in their strata management businesses.

Tip #3 - Testing AI Against Professional Strata Budget Expertise

The experiment:

We took a budget and three years of actuals prepared by a very experienced and competent strata manager, and tested Claude AI to see the accuracy of the AI against the strata manager's work. We gave the AI the actual amounts for three previous years but withheld the strata manager’s budget and contextual information about building-specific requirements.

What we found:

The AI-generated budget estimated expenditure at $559,000, while the strata manager's budget projected $1,002,000—a significant 44% difference. Key discrepancies appeared in major areas that were building specific:

• BMC Levy Contribution: 55% lower in AI budget ($156,434 vs $351,513).
• Car Parking Fund: 66% lower in AI budget ($87,013 vs $252,502).
• Concierge Services: 39% lower in AI budget ($146,574 vs $240,000).

What worked well:

• AI applied reasonable inflation factors (3.5-5%) to historical data.
• It maintained appropriate expense categories and accounting structures.
• Generated professional visualisations and clear documentation.
• Accurately projected operational expenses.
• Produced a fiscally conservative budget with a large surplus.
• Completed the analysis in minutes rather than hours.

What didn't work:

• AI couldn't anticipate planned upgrades or special projects without being told.
• Lacked awareness of building-specific issues requiring attention.
• Had no knowledge of contract renewals or service changes.
• Missed context about long-term maintenance planning.
• Could not factor in committee-approved initiatives.

Effective prompting techniques:

Through our experiment, we identified these essential prompting strategies:

1. Provide specific context about planned projects and known building issues.
2. Use detailed prompts that include specific adjustments and their justifications.
3. Include forward-looking information about scheduled maintenance.
4. Implement a two-stage process: first historical analysis, then contextual adjustments.

Example of an effective prompt:

"Prepare a budget based on the actuals, but make the following adjustments:

1. Increase the BMC contribution to $351,513 based on the draft BMC budget.
2. Allocate $252,502 to the Car Parking Fund for scheduled maintenance.
3. Increase concierge services to $240,000 to accommodate extended hours.
4. Include $50,000 for foyer refurbishment from the capital works plan".

The conclusion:

AI excels at analysing historical data and creating professional documentation but requires human expertise to provide critical context. The optimal approach combines AI efficiency with strata managers’ irreplaceable knowledge:

1. Use AI to handle initial analysis and standardised documentation.
2. Have strata managers provide building-specific context and future requirements.
3. Let AI generate the formatted budget and visualisations.
4. Have managers review and finalise before presentation.

This hybrid approach reduces mechanical work while leveraging the strata manager's unique property knowledge—saving approximately 60-70% of the time typically spent on budget preparation while maintaining or improving quality.

How we can help:

Contact Aleks on LinkedIn or via aleks@michaelteys.com to learn how to implement effective AI-assisted budget preparation for your strata properties, including our framework for providing appropriate context to ensure accuracy while maximising time savings.

Suggested reading: AI Tips for Strata Professionals - Streamlined Strata Budget Preparation

‍