Google Ads Updates, October 2026: What Ecommerce Challenger Brands Need to Know

Google shipped or tested seven Google Ads and Merchant Center changes in the first week of October 2026: a Target ROAS calculator built on profit margin, self-serve Conversion Lift for Search and Performance Max, new AI Max reporting columns, a looser “view” definition for Demand Gen Display ads, Video Campaign Groups with shared reach and frequency, a Direct Offer promo asset for AI Mode, and a Universal Commerce Protocol (UCP) integration hub for checkout. For ecommerce challenger brands, the rule of thumb is simple: adopt the updates that let you measure Google, watch the ones that change how Google keeps score, and put hard limits on the ones that hand Google more of your budget, your discounts, or your checkout.

Credit where it’s due: most of these changes were first spotted inside live accounts by PPC practitioners and reported by Anu Adegbola at Search Engine Land, which is where we caught them. Several are betas Google hasn’t formally documented yet, so we’ve noted which is which.

Here’s our position up front, because it shapes everything below: measure first, then trust. Google’s automation is useful. It is also built by a company that gets paid when you spend, not when you profit. Those two goals overlap a lot of the time, but not all of the time, and a challenger brand with thin margins and no billion-dollar cushion lives in the gap between them. So we stay moderately skeptical of Google’s claims, especially attribution that keeps finding new ways to take credit. We test each new automated feature against our own numbers, look at total performance and not just the campaign report, and never hand over control completely.

We’ve sorted the week’s news into three groups:

1. Updates that help you hold Google accountable (use these first)
2. Updates that change how Google keeps score (adjust your reporting)
3. Updates that hand Google more control (set limits before you opt in)
4. A guardrail checklist for all seven

Updates that help you hold Google accountable

Can challenger brands now test whether Performance Max is actually incremental?

Yes, if you clear the bar: Google has opened user-based Conversion Lift to Search and Performance Max campaigns, and eligible accounts need at least 1,000 observed conversions and a $5,000 minimum campaign budget, according to Google’s Conversion Lift setup guide. Previously, Search and PMax advertisers had to go through a Google rep to run one, as Search Engine Land reported on October 7.

This is the most useful item on the list, and we’ll give Google credit for it. A lift study withholds your ads from a random slice of users (Google lets you set that holdback anywhere from 1% to 50%) and compares what they buy with what the exposed group buys. The difference is your incremental lift. That’s a very different number from the conversions Google’s attribution model hands itself credit for.

Why does that matter so much for ecommerce? Because Performance Max loves people who were already going to buy. Brand searchers, cart abandoners, and loyal repeat customers convert at high rates, which makes any campaign that reaches them look brilliant. The classic warning shot is eBay’s own experiment: economists at eBay and Berkeley found that brand-keyword ads had no measurable short-term benefit, because those shoppers simply clicked the organic link instead (NBER working paper, 2014). Google even acknowledges the overlap by offering brand exclusions for PMax.

• If you qualify: run a lift study on PMax before you raise its budget again. Pick a frequent conversion action (purchases, not newsletter signups) so the result is statistically useful.
• If you don’t: 1,000 conversions is a real hurdle for a young brand. A simple geo holdout (pausing ads in a handful of comparable regions for a few weeks) gets you a rougher version of the same answer.
• Watch the cost: a bigger holdback gives cleaner data but withholds ads from more shoppers. Ten to twenty percent is a sensible starting range for most challengers.

What do the new AI Max columns show?

Google Ads is rolling out three new columns for Search campaigns: Locations of interest, Optimized targeting, and Brand inclusions, which Vivek Gupta spotted and Search Engine Land reported. They don’t change how AI Max for Search behaves. They show you, in one table, which campaigns have which automated settings switched on.

That sounds minor. It isn’t, for one reason: automation creeps. AI Max turns on search term matching, rewrites ad copy, and can send traffic to different landing pages than the ones you picked. When a campaign’s cost per acquisition drifts, the first question is always “what changed?”, and until now the answer was buried in individual campaign settings. Add the columns, save the view, and audit it monthly.

• Optimized targeting: flag any campaign where audience expansion is on and compare its new-customer rate with campaigns where it’s off.
• Brand inclusions: confirm these sit only on campaigns meant to capture your own brand, so brand traffic doesn’t quietly inflate non-brand results.
• Locations of interest: if you ship only to certain regions, make sure “interest in” a location isn’t spending money on shoppers you can’t deliver to.

For context, Google says advertisers who switch on AI Max typically see 14% more conversions or conversion value at a similar CPA or ROAS (May 2025). That’s Google’s own number, measured with Google’s own attribution. It may well hold in your account. The columns, plus a lift test, are how you find out.

Should you trust Google’s new profit-margin tROAS calculator?

Use it as a sanity check, not a strategy: the beta calculator turns your average profit margin into a breakeven Target ROAS (a 15% margin produces a 667% target), and then shows projected clicks, revenue, spend, and profit, as Search Engine Land reported on October 6. Paid search specialist Arpan Banerjee spotted the beta in an account; we couldn’t find Google documentation for it yet.

The math is sound. Breakeven ROAS is one divided by your margin, and Google defines Target ROAS as conversion value divided by cost. If too many brands still set “400%” because it felt right, this is a real improvement.

But notice what Google is and isn’t doing. It still bids on the conversion value you report, which for most stores is revenue. The calculator doesn’t make bidding profit-aware; it only helps you pick a better number to feed it. And the number is only as good as the margin you type in. A 40% product margin can shrink to 25% once you subtract shipping, returns, payment fees, and discounts, and that one change moves your breakeven from 250% to 400%.

• Use fully loaded margin: product cost, shipping, average return rate, payment fees, and any standing discounts. Not gross margin from your Shopify dashboard.
• Segment by margin: one blended margin hides the low-margin hero SKU that eats your budget. Split campaigns or product groups by margin band and give each its own target.
• Feed it profit directly: if your catalog supports it, add cost_of_goods_sold to Merchant Center so Google Ads can report gross profit alongside revenue.
• Treat the projections as a pitch: the weekly “estimated profit” is Google’s forecast of what happens if you spend more with Google. Compare it with your actuals after 30 days.

Updates that change how Google keeps score

What changed in Demand Gen view-through conversions?

For Display ads inside Demand Gen campaigns, Google is switching the definition of a “view” from Active View viewability to a rendered impression, meaning at least one pixel on screen for any length of time, according to Search Engine Land. The change rolls out automatically over the coming weeks, and there’s nothing to opt out of.

Compare that with Google’s own Active View standard for display: at least 50% of the ad visible for at least one second. The new bar is far lower. Google’s stated reason is consistency across Demand Gen inventory, and the same one-pixel rule already appears in its help page for view-through bidding in Demand Gen.

Here’s the practical problem. A view-through conversion (VTC) gives an ad credit for a sale even though nobody clicked it. Lower the bar for what counts as a view, and more sales get credited to ads that a shopper may have scrolled past in a blink. Your actual revenue doesn’t change. Demand Gen’s reported results probably will. Google didn’t change what counts as a conversion; it changed what counts as the ad that gets credit. Phrasing matters, and so does fine print.

• Annotate the date: mark the switchover in your reporting so a VTC jump doesn’t get read as a performance win.
• Report VTCs separately: keep click-through conversions as the headline number for budget decisions, and treat VTCs as a directional signal.
• Check bid strategies: if a Demand Gen campaign bids toward view-through conversions, the algorithm now sees more “success.” Watch spend and blended store revenue for the next 30 days.

Are Video Campaign Groups worth testing?

Yes, if you run more than one upper-funnel video campaign: Video Campaign Groups let you set reach and frequency goals across several campaigns at once instead of one campaign at a time. Google made reach and frequency optimization for video campaign groups available globally in July 2026, and a beta spotted this month extends grouping to eligible Demand Gen campaigns, with options to deduplicate audiences or cap frequency across the group. Google Ads specialist Bia Camargo spotted that beta, and Search Engine Land reported it.

Frequency is where challenger brands quietly burn money. When a prospecting video, a product demo, and a seasonal promo each cap at three impressions a week, the same shopper can still see your brand nine times. Grouping the campaigns should cut that waste and push budget toward people you haven’t reached yet. “Should” is doing some work there. Whether Google actually coordinates delivery, or just reports the overlap more neatly, is the open question.

• Measure unique reach before and after: if group-level reach doesn’t climb at the same budget, the feature is cosmetic for your account.
• Set the frequency cap yourself: don’t leave it on Google’s suggestion. Two to four exposures a week is plenty for most product categories.

If video is a growing share of your plan, our video-first content strategy post covers how to build the creative so each dollar of reach works harder.

Updates that hand Google more control

How do Direct Offers in AI Mode work, and what’s the risk?

Direct Offers let Google’s AI show a promo code under your ad in AI Mode, choosing which incentive to show each shopper based on their context, and the beta asset only serves on campaigns using AI Max or text customization, per Search Engine Land. Google has run a Direct Offers pilot since January 2026 and expanded it at Google Marketing Live in May. The new asset adds single-use codes uploaded by CSV, start and end dates, and daily caps on claims or total discount value.

On paper, this is smarter discounting: offer 15% to the hesitant shopper and nothing to the one who was buying anyway. In practice, you’re asking an ad platform to decide who gets your margin. And the shoppers most likely to “claim” an offer are high-intent shoppers already near the checkout. That’s the group most likely to buy at full price. Right? Right.

Note the requirement, too. To use the shiny discount tool, you have to switch on AI Max or text customization. That’s a bundle, and bundles are rarely designed for the buyer’s benefit.

• Cap everything: set daily limits on both claims and total discount value from day one. Start low.
• Use single-use codes: unique codes stop your promo from leaking onto coupon sites and let you tie every redemption back to the offer.
• Exclude your best sellers: don’t discount products that already sell at full price.
• Test incrementality: compare conversion rate and margin per order between campaigns with and without the offer. If the discount isn’t creating sales, it’s just a rebate on sales you’d have made.

What is Google’s UCP integration hub, and should you connect?

The UCP integration hub is a Merchant Center setting that lets U.S. merchants choose how shoppers move between Google and their store: cart transfer to your own site, native checkout on Google, or identity linking. Google’s Merchant Center help page says it’s rolling out gradually in the U.S., with Australia and Canada next year. Universal Commerce Protocol is the open standard Google announced in January 2026 to let AI agents in AI Mode and Gemini browse, build carts, and buy on a shopper’s behalf.

This one isn’t an ad feature yet, but it’s the most strategic decision on the list. AI shopping is coming whether you like it or not, and being invisible to it is a real risk. The question is how much of the customer relationship you hand over in exchange for that visibility.

• Cart transfer keeps you in control: the shopper finishes on your site, so you keep the email, the upsell, the post-purchase flow, and the first-party data.
• Native checkout trades data for convenience: lower friction, but Google sits between you and your customer. Read the terms on data sharing, returns, and customer contact before enabling it.
• Identity linking is worth a hard look: it can carry loyalty status into Google’s surfaces, which helps conversion, but understand exactly what customer data flows each way.

Our recommendation for most challengers: start with cart transfer, measure it, and treat native checkout as a test with a defined end date. Your product data matters more than ever here, so make sure your feed and site are in shape (our Shopify SEO guide covers how products already show up in AI Mode and ChatGPT).

Your guardrail checklist for all seven

Why do challenger brands need guardrails Google won’t set for them?

Because Google’s incentives and yours don’t fully line up, and Google has real pricing power. In the 2024 U.S. v. Google ruling, Judge Amit Mehta found that Google had been “charging supracompetitive prices for general search text ads”. A platform with that kind of pricing power has little reason to design defaults that spend less.

You can see it in the small print, too. When Google offers Smart Bidding Exploration, which lets Target ROAS campaigns dip up to 30% below target to find new queries, its help page advises against raising your target to offset that. Translation: accept a lower return on the spend, please. That’s a reasonable trade for some brands. It’s a terrible default for a brand running at a 20% margin.

None of this means “don’t automate.” It means automate inside a fence you built.

How do you know if automation is stealing credit from your manual campaigns?

Look at total performance, not the campaign report: if an automated campaign’s results jump while your manual campaigns, branded search, or organic sales slip by a similar amount, the automation likely moved existing sales into its own column rather than creating new ones. Attribution creep (each tool counting a bit more as its “own”) makes this easy to miss, because every report Google shows you can look better at the same time your bank balance doesn’t.

• Track one blended number: total store revenue divided by total ad spend (often called MER, or marketing efficiency ratio), pulled from your store, not from Google Ads. If campaign ROAS climbs and MER doesn’t, something is double-counting.
• Watch the neighbors: when you launch or scale PMax or AI Max, chart conversions for your manual Search and Shopping campaigns over the same weeks. A matching drop is a red flag, not a coincidence.
• Check brand and organic: branded search clicks and organic revenue that fall as automated spend rises usually mean the automation is buying traffic you already owned.
• Count new customers: automation that “wins” mostly on returning customers is collecting from people you’ve already paid to acquire.
• Change one thing at a time: roll out one automated feature, let it run for a few weeks, and read the full picture before turning on the next.

If you want the search side of this to pull its weight, pair the guardrails with smarter keyword targeting. Our guide to buyer intent keywords shows how to find the terms that convert, which is a much better starting point for AI Max than “let Google figure it out.” And if you’re weighing which AI tools deserve a place in your workflow at all, we’ve written about putting generative AI to work without losing control.

Final thoughts

This week’s updates point in two directions at once. Google is giving advertisers better ways to check its work: lift tests, settings visibility, margin-aware targets. It’s also moving more decisions into its own hands: what counts as a view, which shopper gets your discount, and where the checkout happens. Both trends will continue.

Challenger brands win this by being the adult in the room. Measure first, then trust, and only as far as your own numbers allow. Use the measurement tools before anything else, because they tell you which automation is earning its keep. Adopt the features that shift control slowly, with caps and end dates. Judge every change by total store performance, not by how good one campaign looks. And never let a platform that’s paid on spend be the only voice deciding how much you spend.

Want a second set of eyes on your Google Ads guardrails? Contact Phrasing to audit your Performance Max, AI Max, and Merchant Center setup, and to build a testing plan that measures what Google’s automation is actually worth to your margins.

Key Takeaways

What were the biggest Google Ads updates in October 2026?

Google opened Conversion Lift to Search and Performance Max, added AI Max reporting columns, tested a profit-margin Target ROAS calculator, loosened the Demand Gen Display view definition, expanded Video Campaign Groups, tested Direct Offers in AI Mode, and rolled out a Merchant Center UCP checkout hub in the U.S.

Which update should ecommerce brands use first?

Conversion Lift. It’s the only one of the seven that measures how many sales your Search and PMax spend actually created, rather than how many Google credits itself with. Eligible accounts need 1,000 observed conversions and a $5,000 campaign budget.

Does the new tROAS calculator make Google bid for profit?

No. It converts your profit margin into a breakeven ROAS target, but bidding still optimizes the conversion value you report, usually revenue. Enter a fully loaded margin that includes shipping, returns, and fees, or the suggested target will be too low.

Why will Demand Gen view-through conversions go up?

Google is changing what counts as a view on Demand Gen Display ads from Active View (50% of the ad for one second) to a rendered impression (one pixel, any duration). More ads qualify as viewed, so more sales get credited, even if real sales don’t change.

How can you tell if an automated campaign is cannibalizing manual ones?

Compare total store revenue against total ad spend, not campaign-level ROAS. If an automated campaign’s reported results rise while manual campaigns, branded search, or organic sales fall by a similar amount, it’s likely taking credit for existing sales rather than creating new ones.

Are Direct Offers in AI Mode worth testing?

Only with strict limits. They require AI Max or text customization, and Google’s AI chooses which discount each shopper sees. Set daily caps on claims and value, use single-use codes, exclude full-price best sellers, and test whether the offers create new sales or just discount existing ones.