Go deeper: 5 min. read
Chris Hudson, general manager of Mark Miller Subaru, incentivized staff with a $100 gift card for complaints to identify areas where AI tools could streamline processes. This 'Frustration Harvest' method helped pinpoint both digital and non-digital inefficiencies, such as unsafe parts-moving and slow Wi-Fi. Kevin Pitts of Tom Masano Auto Group similarly suggests addressing operational 'pain points' with AI. He developed tools like a recon photo tool (automating inventory photo posting), a marketing-accountability tool (identifying SEO/SEM issues and generating content drafts), and a financial-reporting tool (analyzing general-ledger data for trends).
Mike Yates of BMW of Bridgewater encourages a 'just do it' approach to AI, starting with simple tasks like using AI to review vendor contracts, which saved him from unfavorable terms. He recommends using paid AI accounts for security and data privacy, training AI to match the dealership's specific voice, and directly connecting AI to data sources like Google Analytics 4 for accurate reporting without intermediaries. He also organized an AI video contest for staff to encourage experimentation. Kevin Pitts advises using AI to proactively identify and clean up processes by directly asking it for solutions.
AI tools are not 'set and forget' solutions; continuous evolution is key. Experts emphasize validating AI outputs, such as Pitts reviewing all AI-generated reports before sharing. Critical customer-facing applications, like social media posts, require human approval, while internal tools, such as the recon-photo tool, can operate autonomously. Hudson advocates for fostering a culture where staff can independently develop and 'show and tell' new AI ideas, with viable concepts integrated into broader platforms. Legally, customer data must be excluded from AI training, and AI tools interacting with customers should disclose their artificial nature, potentially with legal review for complex implementations.