Lendi Measures Invisible Work Of AI

Lendi Group, an Australian home loan and property sales company, is working to improve its metrics for AI, according to AdExchanger. The company's General Manager of growth, Michael McCash, notes that measuring chatbots' impact on brand discovery is still a mystery. Lendi has developed an in-house model for multi-touch attribution and partnered with Mutinex as its marketing mix modeling vendor.

Lendi's use of marketing mix modeling has helped the company identify efficient channels, such as old-school, free-to-air television, which was previously underweighted. The company has since diverted money from less efficient channels to TV, resulting in increased ROI.

Media buyers often favor channels they use themselves, which can lead to inefficient spending. Marketers also tend to trust readily available metrics, such as clicks and impressions. However, these metrics may not provide a complete picture of AI's impact on customer behavior.

The rise of AI platforms like ChatGPT has made it difficult for marketers to track how their brands are discovered. Lendi is using Mutinex's AI Visibility tool to generate insights by mass prompting language models with pertinent searches. The company is also working with other vendors, such as Tracksuit and Askable, to measure brand metrics and track customer surveys.

Lendi is testing the effectiveness of advertising in AI chatbots, but so far, the results have been middling. The company's marketing mix modeling shows that the CPMs and customer acquisition rates are too high to justify significant investment. Instead, Lendi is focusing on measurement and understanding how chatbots impact discovery.

The company expects that AI platforms will provide more tools and data as they grow their ad businesses. This will help marketers understand how research through language models changes customer behavior. Lendi believes that accessing this knowledge is the next frontier for marketers.