Heidi Sturrock, a paid search consultant with 24 years of industry experience, joined me on a recent episode of PPC Live The Podcast. The episode covers a broad match mistake with an unexpected silver lining, and Heidi’s experience testing AI Max across 50+ accounts.
The broad match mistake — and the unexpected silver lining
Early in her career, Heidi ran a competitor conquest campaign for a high-spending B2B SaaS client using broad match — without adding negative keywords — and launched it on a Friday with a large daily budget. Over the weekend, the client’s call centre was flooded with angry calls from the competitor’s customers looking for refunds and tech support.
When Heidi called the client to own up, he surprised her by seeing it as an opportunity — training his sales team to handle the calls as soft pitches, offering switchers a 50% discount on their first month. The campaign was then split into two — one targeting disgruntled competitor customers, one for general competitor prospecting — giving better control over spend and intent.
The lessons: Don’t launch on a Friday, and know your stakeholders
Two clear lessons emerged from the story. First, never launch significant campaigns or budget changes on a Friday — the algorithm needs monitoring during its learning period and mistakes can compound unnoticed over a weekend. Second, always include all key stakeholders in client meetings.
Having both the entrepreneur and head of sales in the room meant everyone knew who to contact when things went wrong — and the entrepreneur’s visionary thinking turned a crisis into an opportunity.
Advice for when you’ve made a mistake
When something goes wrong, the first step is to stop the bleeding immediately — pause whatever is causing the problem rather than waiting for the algorithm to self-correct. Then call the client directly, own the mistake fully without deflecting blame, explain clearly why it happened, and come prepared with a solution and next steps.
Handling a mistake with honesty and accountability can actually build client trust rather than destroy it.
Common account mistakes that drive Heidi mad
Two mistakes come up repeatedly in audits. The first is attribution windows that don’t reflect the actual sales cycle — particularly for high-ticket or long-consideration products, where a short window starves the algorithm of conversion data and creates a cycle of frustration between client and agency. The second is fixating on secondary KPIs like CPC or CTR at the expense of the agreed primary goal.
If a campaign is hitting its ROAS target, a rising CPC is not necessarily a problem — the algorithm may simply be entering higher-intent auctions, and ten converting high-CPC clicks are often worth more than hundreds of cheap ones that don’t.
AI Max for search — what’s actually working
Heidi has tested AI Max across more than 50 accounts, with around two thirds seeing strong results and one third underperforming — typically due to insufficient historical data, conversion volume, or poorly defined targets. Her advice is to run it as an experiment first rather than switching everything over at once, and to treat the setup carefully — giving the algorithm the right first-party data, sensible targets, and constraints like landing page exclusions where needed. A step-by-step guide is coming to her blog soon.
One big takeaway
Don’t fight the changes coming to the industry — embrace them. The AI-powered features in Google Ads are genuinely powerful when set up correctly, and the marketers who take the time to master the new rules will be the ones who come out ahead.
Where to find Heidi
Heidi is active on LinkedIn and offers free guides at HeidiSturrock.com, including a free prompt for writing high-performing ad copy with LLMs. She’ll also be speaking on a panel at SMX Advanced in Boston in June. She will be part of the fully audience-driven Ask the Experts session. No scripts. No preset talking points. Just the conversations that matter most — driven by you.
Login
