At the end of every year I find myself thinking about the same question: what actually moved the needle this year, and what was just noise?

There's been no shortage of noise in Google Ads during 2025. AI announcements, campaign type changes, new bidding options, updated attribution models. Most of it doesn't change what you should be doing tomorrow morning.

What follows are five genuine observations from managing e-commerce accounts across the year — things I've seen make a real difference, and things I've watched go quietly wrong. None of this is prediction. It's pattern recognition.

Product Feed Quality Is the Lever Most Businesses Haven't Pulled

I've said this before, but it bears repeating because I keep seeing it: for most e-commerce businesses running Shopping campaigns, fixing the product feed has more impact than any bid change.

A 10% bid increase gives you roughly 10% more visibility. Fixing vague product titles, adding missing attributes, improving descriptions, and correcting category assignments can double impression share. The maths aren't close.

The reason it doesn't happen is that feed work is unglamorous. It's spreadsheets and product data, not campaign dashboards. But Google's algorithm is matching your product titles against search queries — if your titles are generic, you're invisible for the searches that matter.

In 2026, feed quality matters more, not less. Performance Max is increasingly the campaign type Google pushes, and Performance Max is almost entirely dependent on what you feed it. Garbage in, garbage out — except now the garbage gets distributed across six channels simultaneously.

The practical action: audit your top 20 products by spend. Are the titles specific and accurate? Do they include key attributes a buyer would search for? If not, that's where to start.

First-Party Data Is Becoming a Real Competitive Advantage

There was a period when building customer lists felt like a nice-to-have for large retailers with marketing teams. That period is ending.

With third-party cookies increasingly unreliable and Google leaning harder into its own signal ecosystem, businesses that have first-party data — actual customer lists, matched email addresses, remarketing audiences with meaningful volume — are getting better results from Smart Bidding than those who don't.

This isn't complicated to build. It requires uploading your customer email list to Google Ads, connecting your CRM or Shopify customer export, and refreshing it regularly. It requires building proper remarketing audiences: cart abandoners, past purchasers, product viewers. It takes a few hours to set up correctly.

Once it's in place, Google's bidding algorithms have more signal to work with. They can identify people similar to your existing customers. They can bid more aggressively for users showing patterns that match your past converters. The improvement compounds over time as the audience data grows.

The businesses that build this infrastructure in 2026 will have a structural advantage over those running campaigns with only cookie-based audiences or no audience data at all.

Performance Max Works When You Give It Good Inputs — and Fails When You Don't

I've now seen Performance Max work brilliantly and fail comprehensively across different accounts, and the dividing line is almost always the same: the quality of what you put into it.

When businesses give Performance Max strong creative assets (clear product images, varied headline options, specific description copy), an accurate and detailed product feed, clean conversion tracking, and a sensible audience signal — it performs. When they create a campaign with generic assets, a poorly maintained feed, and no audience signals, it underperforms and burns through budget finding nothing useful.

The insight here is that Performance Max doesn't compensate for weak inputs — it amplifies them. A good campaign structure with excellent creative will be dialled up by the machine learning. A weak structure with lazy creative will have its weaknesses distributed across YouTube, Gmail, Display, and Search simultaneously.

If Performance Max isn't working for you, the first question isn't "should I switch campaign types?" — it's "what am I feeding it, and is that actually good?"

Search Term Transparency Has Declined — Weekly Negative Keyword Reviews Matter More

Google has progressively reduced visibility into the search terms triggering your ads. This isn't a conspiracy; it's partly privacy policy and partly a consequence of how broad match and Performance Max work. But the practical effect is that wasted spend from irrelevant queries is harder to spot and easier to ignore.

The response isn't to give up on search term analysis — it's to do it more consistently, not less.

Weekly negative keyword reviews catch the queries that have accumulated over the past seven days. Monthly reviews miss things. Without any review cadence, you accumulate waste that compounds — each irrelevant term continuing to trigger ads, each pound spent there a pound not spent on something relevant.

The other side of this is that search terms also tell you about demand. New queries appearing in your account — terms you hadn't thought to target — often signal product interest you can act on. Regular review turns a defensive activity into an offensive one.

Conversion Tracking Accuracy Is Now a Competitive Edge

Google's bidding algorithms are only as good as the data they're optimising towards. If your conversion tracking is counting the same sale twice, firing on the order confirmation page and again on a thank-you page, or missing mobile transactions entirely — the algorithm is steering by a broken compass.

This has always been important in theory. In 2025 I've seen it become important in practice: accounts with clean, accurate conversion tracking consistently outperform accounts with messy data, even when bid strategies and budgets are comparable.

The common problems I find in audits are: duplicate conversion actions being counted, micro-conversions (page views, button clicks) being included in the primary optimisation goal alongside sales, missing consent mode implementation reducing match rates, and GA4 import events conflicting with native Google Ads tracking.

None of these are difficult to fix once identified. The challenge is that they're not visible in day-to-day account management — the campaign metrics look normal, because the inaccurate data is what normal looks like. An audit specifically focused on conversion setup is often where the most valuable improvements come from.

The Through-Line

None of these five things are new. Product feed quality, first-party data, input quality, search term hygiene, conversion tracking accuracy — these have mattered for years.

What's changed is the stakes. As Google automates more of the tactical execution, the inputs become more important. The algorithm acts on what you give it. The margin between a well-configured account and a poorly configured one has grown — not shrunk — as Smart Bidding and Performance Max have taken on more of the work.

The businesses winning in 2026 will mostly be the ones that got these fundamentals right, not the ones that found a clever automation shortcut.

If you'd like an objective view of where your account stands across these areas, a Google Ads audit is the place to start.