There’s a version of AI in paid search that everyone’s been sold: plug it in, walk away, watch the results flourish. Anyone who’s managed a paid search program knows that’s not how it works. The real job is messier. It’s digging into impression share data, redoing keyword lists, writing the same QA checklist for the third time this quarter, and somehow still finding time to think strategically about what’s driving performance. AI hasn’t replaced that work. But the right AI tool changes how much time you spend on it. That’s what I found when I started working with DOJO AI on Morningstar’s Paid Search program.

DOJO AI sits in my workflow as a hands-on collaborator: I use it to research and vet keyword expansions, pull and interpret performance data, QA campaigns, and draft business cases. What it does is compress the hours between “I have a hypothesis” and “here’s the data to support it.” For a channel manager running multiple accounts simultaneously in a complex business, that compression is the difference between being reactive and having the headspace to be strategic.

The “Where Data Speaks” Case Study

Where Data Speaks is Morningstar’s B2B brand campaign targeting financial professionals and institutional investors in New York, Chicago and London. I used DOJO AI to conduct research and add new keyword sets to two geographic campaigns, Chicago and New York, which improved campaign efficiency through a few different optimizations.

Chicago: Efficiency AND Volume

Before the keyword expansion (January 19 through February 9), the Chicago campaign was generating 44 conversions at a $16.91 CPA. The keyword set was limited, and I suspected we were missing search intent. After working with Dojo to add the right keywords (February 10 through March 4), conversions jumped to 118 over the same 23-day window and CPA dropped 79% to $3.61. More than double the conversions at a fraction of the acquisition cost, because the new keywords matched intent more precisely and drove more efficient auction wins.

Anualized, the Chicago group is now delivering nearly 3x the conversion volume at a CPA that makes the program significantly more scalable.

New York: An Efficiency Win with Volume Still on the Table

The New York campaign tells an equally compelling story. CPC dropped 25%, CTR climbed 47%, and the new keywords captured 47% of ad group spend within just 18 days of launch.

DOJO also surfaced that 37% of eligible impressions are being lost to Ad Rank, pointing to a clear, actionable path to scale. With the right bid adjustments and tightened keyword-to-ad alignment, the foundation is there to grow both efficiency and reach. That full diagnosis, impression share breakdown, bid cap analysis, and keyword-level detail, would have taken hours to pull manually. Dojo and I had it in one conversation.

The Part People Talk About but Don’t Quantify: Time

The performance numbers are the headline. But the part that’s changed my day-to-day most is time.

DOJO and I have run keyword research across campaigns, QA’d campaigns built from scratch, pulled sitelink and ad performance analysis, reviewed performance for stakeholder meetings, and drafted a full business case for a site initiative. I estimate that at roughly 20 hours a month not spent deep in spreadsheets, two and a half working days returned to actual strategy.

For paid search, the value isn’t just in the hours. It’s in what you do with them. Every hour I’m not manually cross-referencing keyword lists is an hour spent on strategy, testing hypotheses, or building the next business case that drives better performance for Morningstar.

AI in paid search isn’t about automation for its own sake. It’s about whether the tool makes you a sharper practitioner. For me, DOJO AI has done that, not by replacing the work, but by getting me to the right answers faster and with better evidence behind them.

Michelle Mendoza is a Paid Search Channel Manager at Morningstar, Inc., where she manages B2B paid search programs across multiple accounts and countries.