A Practical Framework for ML in SMBs

Most advice on AI adoption is written for enterprises with dedicated data teams and six-figure tool budgets. That doesn't translate well to small and medium businesses. Machine learning can deliver measurable results at SMB scale, but the approach needs to be different.

Here's what we've found works in practice, across engagements in New Zealand and Australia.

Start with the business problem, not the technology

The useful question is "what's costing us time, money, or accuracy?" — not "how do we use AI?" The best ML projects start with a specific operational problem and work backwards to the right technique.

You don't need big data

Production-grade models don't require millions of rows. We've shipped models trained on modest datasets. What matters more is data quality and thoughtful feature engineering.

Buy vs. build depends on the problem

Off-the-shelf tools handle generic tasks well. For anything specific to your business logic and data, a purpose-built model will outperform. Most SMBs benefit from a custom model deployed into their existing systems and maintained over time.

Governance scales down too

Bias, explainability, and data privacy apply at every scale. Addressing these early is cheaper than retrofitting, and it builds trust with both your team and your customers.

Summary

Start small. Pick a real problem. Work with people who have shipped models into production — the gap between a proof-of-concept and a reliable production system is where most projects stall.