A study of 15,666 Google Ads accounts found the average account wastes over $1,100 a month — and 29% recorded zero conversions over 90 days. The difference between accounts that convert and those that don’t often comes down to one thing: knowing who you’re targeting before you spend. Here’s how the targeting stack has evolved, and what actually moves the needle in 2026.

I wrote the original version of this post years ago. The core idea hasn’t changed — every click costs money, so you should be deliberate about who you’re showing ads to. But the tools available to do that have changed dramatically.

Back then, the advice was fairly simple: use demographic targeting (age, income, gender) and add interest-based layers to narrow your audience. That still works. But it’s now just the foundation — layers one and two of a four-layer targeting stack that barely existed when I first wrote this.

Layer 1: Demographics — the basics that still matter

Location. Age. Household income. These are the blunt instruments of targeting, and they’re still useful. If you’re a Wellington-based window shutter company, you don’t want to pay for clicks from Christchurch teenagers.

Google collects this data through account sign-ins, browsing patterns, and inferred signals from the Display Network. It’s not perfect, but it eliminates the most obvious waste.

Think of this as your first filter. It won’t win you campaigns on its own, but skipping it is like leaving your front door open and wondering why the wrong people keep walking in.

Layer 2: Interests and behaviour — what people are actually doing online

Beyond who someone is, there’s what they do. Google can tell the difference between someone casually browsing and someone actively shopping — and it lets you target accordingly.

Someone searching “window shutters” who’s also been visiting home renovation websites, reading interior design content, and comparing quotes is a very different prospect from someone who clicked once out of idle curiosity. Google calls these “in-market audiences” — people who are actively in the market for what you sell.

This layer helps you separate the browsers from the buyers. Combined with demographics, you’re already ahead of most advertisers — a WordStream study of 15,666 accounts found that only 3% achieve elite performance.

Layer 3: Your own customer data — the most valuable layer most businesses underuse

This is where the biggest shift has happened, and it’s worth explaining clearly because “first-party data” is one of those marketing phrases that sounds more complicated than it is.

First-party data is simply the information your business already collects about your customers. Your email list. Your customer database. The phone numbers and addresses people gave you when they signed up or made a purchase. The record of who visited your website and what they looked at.

It’s your data — collected directly from your customers through your business. That’s what makes it “first-party.” It’s different from third-party data, which is information collected by someone else (like a data broker) about people you’ve never interacted with.

Why does this matter for Google Ads? Because you can now upload that customer information — email addresses, phone numbers, even mailing addresses — and Google will match it to real people across Search, YouTube, Shopping, and Gmail. This is called Customer Match. It means you can show ads specifically to your existing customers (great for upselling), or create “lookalike” audiences of people who resemble your best customers, or exclude existing customers so you’re only paying to reach new ones.

You can also set up remarketing — showing ads to people who visited your website but didn’t get in touch. They’ve already shown interest. They’re warmer leads than a cold search.

The numbers make the case. BCG and Google found that companies connecting all their customer data sources generate 2x the revenue from each ad placement, with 1.5x better cost efficiency. McKinsey estimates that businesses without a first-party data strategy may need to spend 10–20% more on marketing to generate the same returns.

Yet most businesses barely scratch the surface. Google’s own data shows that customer lists that are regularly updated see a 17% increase in traffic and conversions. Uploading two types of customer information (say, emails and phone numbers) increases the number of people Google can match by 28%. Three types gets you 35%.

The irony is that this data is already sitting in your business. Your invoicing system, your email platform, your customer database. It just needs connecting to your advertising.

Layer 4: AI-powered targeting — letting the machine learn from your signals

This is the newest layer, and it’s changed the game. Google’s AI can now find your ideal customers in ways that manual targeting simply can’t match.

In practice, this means Google’s system can now look at thousands of signals in real time — what device someone’s using, where they are, what time of day it is, what they’ve searched for recently — and decide not just who sees your ad, but how much to bid for that specific person in that specific moment. It does this for every single search, thousands of times a day.

The results are significant. Google’s AI-powered campaign type (Performance Max) delivers on average 18% more conversions at a similar cost. L’Oréal tested another AI feature and saw 2x higher conversion rates at 31% lower cost per conversion.

But here’s the thing most people miss: AI targeting doesn’t replace the other three layers. It amplifies them. The AI needs good information to learn from — your customer data, your conversion history, who you don’t want to reach. Give it good inputs and it finds more people like your best customers. Give it nothing to work with and it’s just guessing expensively.

The wasted spend problem

All of this matters because the default state of Google Ads is waste.

WordStream’s study of 15,666 accounts found the average account wastes $1,127 per month. A quarter of all businesses haven’t added a single negative keyword. And accounts that do use negative keywords see 3x higher conversion rates than those that don’t.

29% of accounts recorded zero conversions over a 90-day period. Not low conversions — zero.

The gap between well-targeted accounts and poorly-targeted ones isn’t incremental. It’s the difference between a marketing channel that drives your business forward and one that quietly drains your budget every month.

What this means for your business

The shift from 2016 to 2026 isn’t just about having more targeting options. It’s a fundamental change in approach:

Then: You manually ticked boxes for age ranges and locations and hoped for the best.
Now: You build a targeting stack — demographics as a foundation, interests to refine, your own customer data to power it, and AI to scale it — and the system continuously improves based on what you feed it.

The businesses getting the best results aren’t the ones spending the most. They’re the ones giving Google the right information about who their customers actually are.

If you’re running Google Ads and haven’t connected your customer data, haven’t set up remarketing to re-engage website visitors, or haven’t explored what AI-powered targeting can do when it has good inputs to work with — that’s where I’d start.

If you’d like to talk through how these layers apply to your business, get in touch. We manage Google Ads for businesses across New Zealand and we’re happy to walk you through what’s possible with your setup.

Sources:
WordStream, “Google Ads Account Study” (15,666 accounts), 2025
BCG x Google, “Responsible Marketing with First-Party Data” / “The Fast Track to Digital Marketing Maturity”, 2021 (updated 2024)
McKinsey, “The demise of third-party cookies and identifiers”, 2022
Google Ads Blog, Performance Max and AI Max updates, 2023–2025