Google Ads Data Strength Update

Most advertisers now hand a large share of their decisions to Google’s AI. Smart Bidding sets the bids, Performance Max decides where ads appear, and AI Max for Search expands the queries a campaign can show for. That automation can work well, but it depends on one thing advertisers control: the quality of the data it learns from. Google has made that point more explicitly than ever this month, rewriting its Data Strength best practices from a brief overview into a detailed implementation guide. The update, first reported by Search Engine Roundtable, puts much greater emphasis on sending first-party data and identifiers back to Google.

Google describes data strength as building “a trusted decision engine with first-party data.” In practice, it means giving the platform a clearer, more complete picture of what happens after someone clicks an ad. Did they buy? What was the order worth? Were they a new customer or a returning one? Did a sale that started online finish over the phone or in a store? The more accurately those questions are answered, the better Google’s bidding and targeting can work towards outcomes that actually matter to the business.

The guide is organised around four steps: connecting data sources, maximising signals, activating Google’s AI, and proving return on investment. At its centre is Data Manager, Google’s tool for bringing data from websites, CRMs, cloud platforms and offline sources into one place for use across its advertising products.

The first step is making sure the basics are captured reliably. Google recommends tagging the whole site and upgrading to Google tag gateway, which runs Google’s tags through a business’s own domain and infrastructure so they’re less exposed to browser restrictions and ad blockers. It also recommends linking Google Analytics with Google Ads, and connecting offline sources such as CRM records and in-store sales.

The second step is where the guide gets most specific. Google wants advertisers to pass more identifiers with each conversion: consented, hashed customer details such as email addresses and phone numbers through enhanced conversions, along with IP addresses, order or transaction IDs, the Google Click ID and session information. Its recommendation is at least three identifier signals per conversion event, and to send every conversion rather than a selection. Each additional signal improves the chance of matching a sale back to the ad that influenced it, particularly as browser privacy features make traditional tracking less complete.

Google backs this with its own figures. Advertisers who upgrade to Google tag gateway see, on average, a 14% uplift in observed conversions. Advertisers bidding to conversion value who implement enhanced conversions see an average 8% incremental return on ad spend on Search, based on 99 conversion lift studies. Early adopters connecting conversions from multiple sources through Data Manager saw a 13% median increase in attributed conversions. These are Google’s own global averages, drawn from its internal data, so individual results will vary, but they show where Google believes the gains are.

The detail worth noticing is how much of the guide depends on information Google Ads can’t collect by itself. A website tag can record that a checkout was completed. It can’t easily tell Google that the order was later cancelled, that a B2B quote converted three weeks later through an account manager, that the customer has spent thousands over five years, or that a product carried a much higher margin than the one beside it. That information sits in the eCommerce platform, the ERP, the CRM and sometimes the warehouse system.

For many Australian businesses, that’s where the real opportunity lies. A retailer whose Shopify store, ERP and CRM all hold slightly different versions of the same customer will struggle to send Google clean, consistent data. A B2B business that takes most of its revenue through account terms, phone orders or sales reps may find its campaigns optimising towards the small share of activity a website tag can see, rather than the sales that actually drive the business. In both cases, the bidding is working with an incomplete picture, and doing exactly what it’s been told.

This is why Google’s emphasis on transaction IDs and offline data is significant. A consistent order identifier that follows a sale from the storefront through to the ERP lets a business connect an ad click to what happened next, whether that’s a return, an upsell or a repeat purchase. Feeding that back regularly, rather than through occasional manual uploads, gives Smart Bidding a far more accurate signal to learn from.

Sending more customer data to an advertising platform should be approached carefully. Google’s guidance refers specifically to consented data, hashed before it leaves the business, and says Data Manager uses privacy-enhancing technologies such as confidential matching. In Australia, proposed reforms to the Privacy Act are also raising expectations around how customer information is collected and used for marketing. The safest approach is to treat consent capture as part of the same project, making sure customers understand how their data will be used at the point it’s collected rather than retrofitting it later.

Google’s guide is detailed, but the starting point is a practical audit. Check that every meaningful conversion is being tracked, not just online checkouts. Confirm whether enhanced conversions are switched on and working, whether Google Analytics is linked correctly without duplicate connections, and whether Google tag gateway is worth adopting for your setup. Then look at what’s missing: offline sales, cancellations, customer lifetime value and new versus returning customers. Google’s built-in diagnostics can help identify setup errors along the way.

Some elements are still rolling out. Google notes that using IP addresses through Customer Match and a budgeting tool in Google Analytics are both “coming soon”, with no specific timing for Australia, so they’re worth watching rather than planning around for now.

The broader lesson is that Google Ads performance is increasingly decided before a campaign is built. Automated bidding has become very good at chasing the outcomes it’s shown. The businesses that benefit most will be the ones that show it the right outcomes, which usually means getting the data flowing between their platforms first.