A new study has found that artificial intelligence can forecast business-to-business (B2B) software adoption with a level of accuracy approaching that of real buyers, offering companies a faster and more cost-effective way to evaluate product pricing, packaging and market demand before launch.
The research compared responses from 160 real B2B software buyers with five different AI-generated buyer panels using conjoint analysis, a widely used market research method that measures how customers make purchasing decisions. One of the AI models predicted adoption rates within just 1.2 percentage points of actual buyer behaviour, representing a significant improvement over previous synthetic research.
According to Brandspur Brand News, the study highlights major progress in AI-powered market research while also identifying clear limitations that businesses must consider before relying entirely on synthetic consumer data. Researchers found that conventional large language models often overestimate customer willingness to purchase, with basic AI models predicting adoption rates as high as 98 per cent compared with the real-world figure of 53 per cent.
To improve performance, researchers developed AI personas using insights from more than 100 recorded customer interviews and sales conversations. They further enhanced the models by incorporating information about buyer loyalty to existing vendors, switching thresholds and realistic conditions under which customers would reject a product.
The findings suggest that AI buyer panels are particularly effective for testing pricing strategies, forecasting product adoption and screening product concepts before companies invest in expensive human market research. However, the study also found that AI remains less reliable when assessing product attributes that customers can only evaluate after using a product, such as onboarding quality, customer support and user experience.
Researchers concluded that AI should complement rather than replace traditional customer research. Instead of eliminating human studies, businesses can use AI to refine concepts, identify promising product features and narrow research priorities before conducting targeted surveys with real customers.
The study indicates that organisations adopting this hybrid approach could significantly reduce research costs and shorten product development timelines, enabling faster and more informed strategic decisions in increasingly competitive B2B technology markets.
