Two days ago, Search Engine Land reported that Google Ads is testing a new campaign creation interface where advertisers describe their goals in plain English instead of clicking through the familiar objective tree. Google Ads Liaison Ginny Marvin confirmed the experiment:
We welcome advertisers' feedback around how and where this type of workflow would be helpful.
Ginny Marvin
The screenshot shows a single text box asking "What's your goal for this campaign?" with Google's AI promising to suggest goals and assets based on your input. A "Create campaign manually" option sits below for anyone who wants the traditional path. That's the entire interface.
For operators who live inside pipeline forecasts, this experiment raises a question that matters more than the UX convenience: what happens to your cost model when the setup process itself becomes a black box?
The Convenience Trade-Off Has a Price Tag
Google's pitch is straightforward. Typing "I want to generate qualified leads for our enterprise security product in the Northeast" is faster than navigating campaign type, sub-type, bidding strategy, and asset requirements. For a first-time advertiser or a founder spinning up their first paid campaign, that friction reduction is real.
The problem is that objective selection is one of the largest single levers on your cost per result. Choose the wrong campaign type, and you're optimizing for clicks when you needed form fills. Choose the wrong bidding strategy, and your CAC payback stretches from two quarters to four. The traditional setup flow, for all its friction, forces you to make those decisions explicitly. You see the trade-offs. You own the choices.
When the AI suggests the setup, you're trusting a model trained on aggregate advertiser behavior to understand your specific unit economics, sales cycle, and margin structure. It doesn't. It can't. The model optimizes for what Google can measure: clicks, conversions, and conversion value. It has no visibility into your pipeline velocity, your close rates by segment, or the fact that your CFO just cut your payback threshold from 18 months to 12.
This Isn't New, It's Accelerating
Google has been moving toward AI-driven campaign management for years. At Google Marketing Live 2023, the company introduced conversational campaign creation for Search ads, where you provide a landing page and the AI generates keywords, headlines, and descriptions. Google's documentation claims that small business advertisers using this conversational experience are 63% more likely to publish Search campaigns with "Good" or "Excellent" Ad Strength.
Ad Strength is Google's metric, not yours. It measures asset completeness and predicted relevance, not CAC payback or pipeline contribution.
The current experiment extends this pattern from Search campaigns to the objective-selection layer itself. Meanwhile, Google auto-upgraded campaigns using automatically created assets or campaign-level broad match to AI Max for Search in September 2026. Dynamic Search Ads migrate in February 2027. The direction is clear: Google wants to own more of the decision surface, and the natural language interface is the friendliest possible wrapper for that shift.
The Performance Data Is Mixed
Google's headline claim for AI Max is 14% more conversions or conversion value at similar CPA or ROAS. That's the vendor number. Independent analysis tells a different story.
Smarter Ecommerce's review of more than 250 AI Max campaigns found a median 13% rise in conversion value alongside a 16% rise in cost per acquisition. That's not "similar CPA." That's paying more for the incremental volume. Whether that trade-off works depends entirely on your margin structure and payback requirements, variables Google's AI doesn't see.
Fluency's 2026 advertiser survey shows Performance Max adoption jumped from 60% to 71% of advertisers in a single year. Adoption is accelerating, but adoption isn't the same as profitable adoption. The operators I talk to report a consistent pattern: AI-driven campaigns find more volume, but the marginal CPA on that volume is higher than the core. If your CFO is modeling blended CAC, the numbers look fine. If they're modeling marginal CAC on incremental spend, the picture changes.
What This Means for Your Forecast
The natural language experiment isn't a threat to sophisticated advertisers who will continue using manual setup. It's a signal about where the platform is heading and what that means for your planning assumptions.
First, expect more variance in campaign performance as Google's AI makes more decisions. The AI optimizes for conversions, not for your specific definition of a qualified opportunity. If your sales team is already complaining about lead quality from paid channels, this trend won't help.
Second, your attribution model needs to account for the fact that Google's AI is increasingly choosing which queries, placements, and creative combinations to serve. Exact and phrase match keywords can now serve ads inside AI Mode, something that was structurally impossible before September 2026. The control you thought you had over search intent matching is eroding.
Third, the reporting gap is real. AI Max traffic shows up in your search terms report with a match type labeled "AI Max," but you don't get the same granularity on which queries triggered which ads. Your incrementality testing needs to account for this opacity.
The Pilot Plan
If you're running significant Google Ads spend, here's the two-week test I'd run before the natural language interface rolls out more broadly:
Audit your current campaigns for the settings that trigger auto-upgrades: campaign-level broad match and automatically created assets. Decide whether you want those features on your terms or Google's.
Set up a 50/50 experiment on one campaign using Google's AI Max experiment framework. Test one feature at a time, not the full suite. Measure marginal CPA on the incremental conversions, not blended performance.
Document your brand exclusions, URL exclusions, and negative keyword lists before any AI features go live. These are the guardrails that survive the automation layer.
Build a sensitivity table showing what happens to your CAC payback if marginal CPA on AI-driven volume is 10%, 20%, or 30% higher than your core. Share it with Finance before the numbers show up in actuals.
The natural language interface is a convenience layer. The underlying shift, Google making more decisions about your campaign structure, is the change that hits your forecast. Model it now, while you still have a baseline to compare against.
