The newest Google Ads features aren’t always the right ones for your business. Here’s how we think about which tools to use – and when.
Google Ads is not the same platform it was five years ago. Or even two years ago. The pace of change has accelerated dramatically as Google pushes more of its machine learning capabilities into the advertiser’s workflow – automating bidding, expanding audiences, generating creative assets, broadening keyword matching, and now restructuring entire campaign types around AI-driven optimization.
For large e-commerce brands or enterprise advertisers with substantial budgets and equally substantial historical data, many of these features are genuinely powerful and worthwhile. But for local businesses – a regional law firm, a home services company, a dental practice – the equation looks very different. These advertisers typically have tighter budgets, smaller data pools, and a real need for results in weeks, not next quarter.
At our agency, we work with clients across that full spectrum. One of the most important things we do is help each client determine the right balance between embracing cutting-edge Google Ads features and trusting the proven, structured approaches that have delivered results to our clients for years. That balance looks different for every advertiser – but the framework for finding it is consistent.
Here’s how we think about it.
AI Features Need Data, Local Budgets Are Finite
To understand why this balance matters so much for local advertisers, you have to understand what Google’s newer AI-driven tools are actually doing under the hood.
The latest AI-powered features like AI Max, Performance Max, audience expansion, and URL expansion all share a common dependency: they need data to learn. They work by testing combinations – of audiences, of keywords, of creatives, of landing pages – and gradually optimizing toward the combinations that convert. The more conversion data they accumulate, the better they get.
The problem for local advertisers is that this learning phase costs money and time. Google’s own recommendation is often 30–90 days before drawing conclusions on automated campaigns. For a local business spending $3,000/month or less on Google Ads, burning several weeks of budget while an algorithm “finds its footing” is a real and significant risk. And unlike a national brand with 500 conversions per month feeding the machine, a local service business might see 15 or 20 – not enough signal to power sophisticated automation effectively.
This doesn’t mean local businesses should never use these features. It means they need a thoughtful entry point, realistic expectations, and – critically – a strong structural foundation to build from.
To read more about the latest AI tools for digital marketing, click here to visit our AI blog posts.
What "Tried-and-True" Actually Means
Before evaluating any new Google feature, we believe every account needs a solid structural foundation. Nothing flashy, just tried-and-true best practices. They’re the unglamorous, fundamental practices that build healthy, scalable campaigns and determine whether your ad spend produces reliable, readable, optimizable results.
Thematic Campaign and Ad Group Structure
One of the most impactful things we do for search campaigns is organize them thematically. Rather than scattering keywords loosely across ad groups, we build each ad group around a specific service theme – aligned across the keywords in that group, the ad copy those keywords trigger, and the landing page that ad sends traffic to.
For a smaller local advertiser, we may start with 2-3 campaigns. For example,let’s consider a local plumbing company looking for more leads through paid ads. We might start with a brand campaign for their brand name keywords and a “plumbing services” campaign where each ad group represents a distinct service theme: emergency plumbing, drain cleaning, water heater installation, leak detection. The campaign structure stays consolidated to protect a limited budget from being spread too thin, but the thematic discipline happens at the ad group level, where it matters most.
This matters for a few reasons. First, it keeps messaging tightly relevant at every stage of the funnel – the right keyword triggers the right ad, which leads to the right page. That relevance improves Quality Score and lowers cost per click. Second, as the account accumulates data – typically over the first four to six weeks – we can read performance by theme: cost per conversion, cost per click, and conversion rate broken out by ad group give us a clear picture of which service themes are working and which need attention.
From there, we work with the client to act on those insights. If a particular service is highly profitable for their business but faces stronger competition (resulting in higher CPCs and a higher cost per acquisition) we’ll often move that ad group into its own dedicated campaign with its own budget. That isolation gives us the control to scale aggressively on high-value services without those dollars competing against lower-priority themes for the same budget pool. The ad group data earns that decision. We don’t make it upfront.
Controlled Keyword Matching
Broad Match keywords have their place, but for local advertisers with limited budgets, they carry real risk. A broad match keyword like “plumbing company” can trigger ads for searches like “plumbing career opportunities,” “DIY plumbing tips,” or searches for plumbing services that the advertiser doesn’t perform. Every irrelevant click is money that doesn’t come back.
Our default for local advertisers is phrase match and exact match, supplemented by an actively maintained negative keyword list. When we do use broad match – typically in accounts with a solid conversion history and enough data to support it – we pair it with comprehensive negative keyword lists and first-party audience signals to help Google’s algorithm understand who our ideal customers are. Broad Match with no guardrails is an expensive way to generate data. Broad Match with structure and signals can be a powerful expansion tool.
First-Party Data as a Competitive Advantage
For local businesses that have been operating for years, there’s often an underutilized asset sitting in their CRM: customer lists. Uploading customer match lists, lookalike audiences built from existing customers, and website visitor audiences to Google Ads gives the algorithm something meaningful to work with – especially when you’re using automated features that depend on audience signals.
A local HVAC company that uploads its last three years of customer records and builds an audience from that data is feeding the algorithm far more signal than a competitor running the same campaign cold. First-party data is increasingly the difference-maker in automated campaign performance, and it’s an advantage local businesses often overlook.
Google's AI-Driven Features
With that foundation in place, let’s look honestly at where the newer AI features fit – and where they require caution.
Performance Max
Performance Max (PMax) is Google’s all-in-one campaign type that serves ads across Search, Display, YouTube, Gmail, Maps, and Discover from a single campaign. Google’s algorithm decides where to show your ads, to whom, and with what creative combinations based on your asset groups and conversion goals.
For local businesses, PMax is genuinely useful in specific contexts – particularly for advertisers who want to run a connected campaign across multiple Google surfaces without managing each separately, or for businesses with strong visual assets and a large enough service area to benefit from Display and YouTube reach. When built correctly with strong asset groups, clear audience signals, and properly defined conversion goals, PMax can perform well.
The risks to understand: PMax offers limited visibility into where your budget is actually going. Google doesn’t readily show you a breakdown of spend across channels or a full search terms report. For a local business that needs to understand exactly what’s working, that opacity is a real limitation. We recommend running PMax alongside – not instead of – structured search campaigns for clients who want to test it, and setting it up with as many audience signals and creative assets as possible to give the algorithm a head start.
AI Max for Search Campaigns
AI Max is Google’s newest overlay for search campaigns, adding AI-driven features including text customization (rewriting your headlines and descriptions dynamically), URL expansion (sending traffic to different landing pages than the ones you specified), broad match-style keyword expansion, and enhanced audience targeting.
These features sound compelling in Google’s framing. In practice, for local advertisers, several warrant careful evaluation:
Text customization
lets Google rewrite your ad copy based on what it predicts will perform. For local advertisers who have invested in precise, differentiated messaging, handing that control to an algorithm introduces risk. We recommend either keeping this feature off while a campaign establishes its baseline, or reviewing AI-generated copy regularly to ensure it reflects your brand accurately.
URL expansion
is the feature that raises the most concern for structured campaigns. When enabled, Google can direct traffic to any page on your site it deems relevant – not just the landing page you’ve designated. For a local business with a carefully built funnel where specific ad groups lead to specific landing pages, this breaks the thematic relevance chain. We typically keep URL expansion off or set it to “final URL only” for clients whose landing page strategy is deliberately structured.
That said, AI Max does introduce some features worth testing – particularly its enhanced audience targeting signals and keyword expansion in accounts that already have a conversion history to learn from. The key is selective adoption: turning on the components that complement your structure rather than the ones that undermine it.
Broad Match
Broad Match isn’t new, but Google has been aggressively pushing it in recent years – often recommending it alongside Smart Bidding as the “modern” approach to search. The argument is that broad match, when paired with conversion-based bidding strategies, gives the algorithm maximum flexibility to find converting traffic.
For local advertisers with limited budgets and limited conversion volume, we advise against starting here. The irrelevant query traffic is real, the cost is real, and the data required to make it work reliably is more than most local accounts generate quickly. Once an account has 60 to 90 days of structured campaign performance and a meaningful conversion history, selective broad match expansion – with negative keywords and audience signals – becomes a reasonable test.
Demand Gen
Demand Gen campaigns run across YouTube, Gmail, and Google Discover, and they’re designed for upper-funnel awareness rather than direct response. For local businesses whose customers are primarily searching with high purchase intent, Demand Gen is generally a lower priority. If budget is limited, it typically makes more sense to saturate the high-intent search placements first before investing in awareness channels.
Where Demand Gen can make sense for local advertisers: highly visual service businesses (home remodeling, landscaping, interior design) where seeing the work sparks interest, or in markets where search competition is so high that upper-funnel reach is more cost-effective.
Dynamic Search Ads and Dynamic Display Ads
Dynamic Search Ads (DSAs) use your website content to generate ads automatically, without manually specified keywords. Dynamic Display ads similarly auto-generate visual creatives from your assets.
DSAs can be a useful expansion tool for accounts with a large, well-structured website and enough conversion history to guide the algorithm. For local businesses with smaller sites and limited conversion data, they often generate irrelevant ad-to-landing-page pairings and can result in poor Quality Scores. We use DSAs selectively, primarily to catch keyword gaps in well-established accounts rather than as a primary campaign type.
Audience Expansion
Audience expansion tells Google to reach users similar to your defined audiences, even if they don’t match your targeting criteria exactly. For local advertisers with geographically constrained service areas, this can mean paying for impressions and clicks from users outside your market. We recommend turning off audience expansion or setting it to “bid only” rather than “targeting” until you have a clear picture of where your conversions are actually coming from.
Generative Creative Tools (Asset Studio)
Google’s Asset Studio and generative AI creative tools can generate images, headlines, and descriptions using AI. For local businesses without dedicated creative resources, these tools can genuinely help fill gaps – quickly producing image assets for display or PMax campaigns without a design budget.
The caution here is quality control. AI-generated creative can be generic, occasionally inaccurate, and sometimes visually off-brand. We use generative tools as a starting point, not a finish line – reviewing every generated asset before it goes live and supplementing with real photography and brand-specific creative wherever possible. Authentic local imagery (your team, your trucks, your actual work) will almost always outperform generic AI-generated visuals for building trust with a local audience.
Automatically Created Assets
Automatically created assets (ACAs) are a broad category of ad enhancements that Google generates and appends to your ads without explicit input from you. These include dynamic sitelinks, dynamic callout extensions, dynamic structured snippets, seller ratings, longer ad headlines, automated app promotions, automated location assets, dynamic images pulled from your website or landing pages, and dynamic business name and logo assets.
Google enables many of these by default, which means if you’re not actively auditing your account, they may already be running.
For local advertisers, ACAs present a familiar tradeoff: they can add useful ad real estate with minimal effort, but they come with limited control over what actually shows. A dynamically generated sitelink might point to a page that’s irrelevant to the search query. A dynamic image pulled from your website might not reflect your brand well at the size and crop Google chooses. A longer AI-generated headline might dilute or contradict the precise messaging in your manually written copy.
In cases where we enable ACAs, our approach is to audit them at account setup and again periodically as the account matures. For assets like seller ratings – which display your aggregate review score directly in the ad – we generally leave these on, as social proof is almost always beneficial. For assets that touch ad copy or direct traffic to specific pages, like longer headlines and dynamic sitelinks, we review them carefully and turn off the ones that conflict with the account’s messaging strategy. The goal isn’t to disable everything Google offers automatically – it’s to make a conscious decision about each asset type rather than letting defaults run unchecked.
How We Determine the Right Balance for Each Client
Every client that comes to us gets an honest assessment of where they are on what we think of as the “automation readiness spectrum”. It’s not a formal audit – it’s a conversation that covers a few key dimensions.
Budget and time horizon.
A client with $1,500/month and a need for leads within 60 days is likely not a candidate for a Performance Max experiment. A client with $5,000/month and a six-month runway has the data volume and patience to test more aggressively.
Existing conversion history.
If an account has 12 months of structured data with clear conversion tracking, AI features have something to learn from. If it’s a new account or one with limited history, the foundation has to come first.
Appetite for opacity vs. control.
Some clients are comfortable handing more control to Google’s algorithm. Others want full visibility into every dollar. We match the campaign architecture to that preference.
Competitive environment.
In highly competitive local markets where every click is expensive, the cost of a poorly targeted broad match keyword or an irrelevant URL expansion is magnified. In less competitive markets, there’s more room to experiment.
The result is a strategy that usually isn’t entirely traditional or entirely automated – it’s a deliberate blend, with clear guardrails on the AI features we do deploy and a strong structural foundation underneath all of it.
The Bottom Line for Local Advertisers
Google’s AI-driven tools represent a genuine evolution in what’s possible with paid search. Some of them will absolutely be worth adopting – the landscape two years from now will look different than it does today, and the agencies and advertisers who understand these tools will have an advantage.
But for local businesses operating with real budget constraints and a real need for near-term results, enthusiasm for the newest features has to be tempered by a clear-eyed assessment of what those features actually require to deliver. They require data, time, and a foundation that most new or small accounts haven’t yet built.
Our approach is simple: build the foundation first, let it prove itself, and introduce automation selectively where it has the best chance of succeeding and accelerating success. We stay current on every new Google Ads feature – not to adopt them all immediately, but to know exactly when the right conditions exist to put them to work.
That’s the balance. It’s knowing your client, knowing the tools, and making the call with confidence.
Interested in an honest assessment of where your Google Ads account stands? Get in touch – we’d love to take a look.
