Is AI mainly a job-cutting tool, or a way to grow revenue? On the Eye on AI podcast (episode 390), host Craig Smith interviewed Jeff Lunsford, CEO of data-infrastructure company Tealium, and the answer from the vendor’s side was clear: grow.

A caution first. Lunsford sells this technology, so every figure below is his claim. Customers are unnamed and nothing is independently verified. Treat the numbers as the best case, not as settled fact.

Agents are a new sales channel

• About 20% of customers’ inbound web traffic is now AI agents, and he expects 50% by mid-next-year.
• He says agent traffic converts at roughly 38% more revenue per session than human traffic.
• Human traffic is not halved. Agents are “just another delivery channel”, like phones, mobile and the web before them, and total sessions rise because it is easier to send an agent off to search and buy.
• Sites that agents cannot transact on lose the customer to a competitor. He says young “AI natives” will not use a business their agent cannot use.

Fresh data lifts conversion

• Real-time experiences convert 30 to 200% better than delayed ones. Travel sees 30 to 40% gains moving from day-old to real-time data.
• Feeding conversion data to ad platforms lifts return on ad spend by about 25% on average. Ad budgets typically dwarf the technology budget, so that is real money.
• The standout example is a car dealership. It cleaned up its data, and lead-to-sale conversion went from the US average of about 10% to 30%. The market is zero-sum, so it took share from rival dealers and was confident enough to buy others. That is growth, not job cuts.

Customer experience

A shopper browses on a phone, then walks into a store, and the assistant’s tablet already shows what they were looking at. Call centres and abandoned-cart emails get the same live view. Websites also need a machine-readable data layer, because agents now use the site as their front door.

Costs and productivity

• One real-time data layer replaces the tangle of point-to-point integrations that consultants build and then take their knowledge with them. Lunsford says this lowers support, processing and storage costs.
• Staff can now run the platform in plain English through an AI agent, so the long training ramp-up largely disappears.
• Savings can be reinvested in growth or taken as profit.

What about jobs?

Almost nothing was said. Layoffs never come up. The work being replaced is integration and pipeline maintenance, and “humans are not going away” refers to customers, not employees. The framing is more activity and more revenue per company, not fewer staff.

The caveats

• Much of the gain comes from real-time data and better integration, with AI as the new reason to care about it.
• The 38% agent-conversion figure and the dealership result are the claims most worth asking independent evidence for.
• Smith himself noted that enterprise adoption of agents is still at an early stage.

Source: Eye on AI podcast, episode 390: Craig Smith interviewing Jeff Lunsford of Tealium.