How AI and Digital Twins are Transforming Market Research: A Conversation with Xero

The Bold Transformation of Customer Insights

The adoption of AI is fundamentally transforming how organizations harness data to supercharge their customer insights and marketing functions. This transition, however, comes with very real organizational and cultural challenges.

Xero, for example, has undergone a significant evolution in its global insights function. After building a strong foundational stage and embedding insights throughout the business, growing from just a few global research pieces to managing over 300 studies, Xero is now tackling its third chapter: rebuilding and “supercharging the insights engine” with digital research tools and AI. This included making the bold decision to actually pause regular ongoing research programs to properly implement new AI tools.

The Rise of Digital Twins

A key area of experimentation involves leveraging partners like Emporia, an award-winning data collection platform. Emporia is currently utilizing methodology open-sourced by Stanford University research—which studied simulating 1,000 humans—to create synthetic personas or digital twins.

Based on two-hour-long behavioral and psychographic interviews, these AI replicas have shown reliable matches (around 85%) to how the real participants answered subsequent surveys. Xero partnered with Emporia as a design partner to test this technology on hard-to-reach audiences like accountants and bookkeepers across multiple geographies.

These digital twins can be utilized for several valuable applications today, including:

• Qualitative simulations and conversations.

• Lightning-fast message and concept testing.

• Pre-survey testing to identify design flaws before fieldwork begins with real people.

However, experts caution that high-stakes research, such as pricing and packaging studies, should still rely on real people rather than solely on synthetic personas. Even with iterative tools like Yabel (which uses virtual personas for creative feedback), the human element remains necessary for validation and achieving deep human empathy and nuance.

Navigating the Data Quality Crisis

The move toward AI has highlighted a severe challenge in maintaining data quality. AI is a "double-edged sword", as the rate of fraudulent responses and "junk data" (including AI and bots taking surveys) is spiking dramatically. Xero observed its typical scrub rate for removing junk data jump from 30% to 70% in late July following major AI releases.

To combat this, organizations must prioritize high-quality data. This means leveraging partners like Emporia, who can provide LinkedIn-verified high-quality sample for hard-to-reach professional audiences.

Strategic Advice for Marketers

For marketers and insights managers navigating this rapidly changing environment, a few key learnings stand out:

1. Slow Down to Speed Up: Organizations must dedicate adequate time and space to properly test and embed new AI tools into workflows. Trying to fit this experimentation into 10% of existing work won't provide the necessary acceleration.

2. Leverage Existing Knowledge: Businesses with large, established knowledge banks have absolute "gold" for fueling these new tools. Xero, for example, uses tools like Gemini enterprise-wide to synthesize and summarize existing data, improving speed to insight.

3. Experiment and Stay Informed: Marketers should actively experiment with new tools and keep up with the latest industry developments. Excellent resources for staying current include specific YouTube channels like those run by Matt Wolfe and Wes Roth.

4. Lean on Partnerships: If corporate procurement processes prevent internal testing, companies should leverage trusted external partners (like Insights Exchange) to test new technologies and provide honest feedback. These trusted partnerships are more critical than ever in the current AI landscape.

‍