E-commerce
October 28, 2026
Wondering which AdWords updates can truly move your online sales? Four recent announcements deserve your attention: Google Optimize to test your pages, life events targeting, In-Market audiences for search, and Google Attribution to move beyond last-click. Together, they touch the heart of your ad profitability.
The trap would be to activate everything at once, without method or reflection. Each update requires a minimum of testing before consuming budget, and some remain reserved for the largest accounts. So, what AdWords updates truly help e-commerce merchants? On the agenda:
Why does testing your landing pages change your ad results?
How does life events targeting speak to your customers?
What are In-Market audiences used for in search campaigns?
Why abandon last-click in your attribution?
How does Qstomy extend these ad gains after the click?
Let's go.
Summary
What should merchants take away from Marketing Next?
Google used its Marketing Next event to announce several changes to its advertising product suite. Four of them directly concern online merchants: a page testing tool, life events targeting, in-market audiences, and a new attribution model.
These announcements are not trivial. They address three well-known pain points in e-commerce: pages that convert poorly, overly broad targeting, and performance measurement that credits the wrong touchpoint. In other words, they touch both acquisition and understanding of your results.
Good news: part of these features is offered for free to AdWords advertisers, whereas competing tools charged for this type of service so far. The question remains in what order to activate them. The following sections detail each one, including their limitations and precautions before spending.

Convert over 2,000 customers on average per month with Qstomy.
The world’s 1st Shopify AI dedicated to customer conversion



Empowering 200+ e-commerce merchants
Why does testing your landing pages change your results?
Google Optimize is Google's new testing tool, integrated into AdWords. Its goal: create and test variants of your landing pages without writing a single line of code and without relying on a developer. You duplicate a page, modify a title, an image, or a button, then apply the variant to any combination of campaigns, ad groups, and keywords.
The benefit for a merchant is twofold. First, you test on the real page receiving paid traffic, not on a mockup. Second, you isolate what works: an argument, a delivery promise, a product photo. Since ad traffic is paid per click, every conversion point gained translates directly into a lower acquisition cost.
Ce que propose l’outil
A/B tests on your pages.
Native integration with Google Analytics.
Advanced targeting by geographic area or user behavior.
Support for web apps on a single page.
The visual editor displays a preview of changes and maintains a complete activity log. If you wanted to test more but lacked resources, that excuse is now gone.
How does life events targeting speak to your customers?
Life events targeting enriches YouTube and Gmail ads. It targets specific moments: a graduation, a wedding, a birth. For some sectors, like wedding service providers, the use is obvious. But many of these events involve a gift, which opens the function to the majority of e-commerce merchants.
The right reflex is to align three levels of your setup. First, write an ad that names the targeted event, such as 'the ideal gift for a graduation.' Next, repeat this message on the landing page to avoid a disconnect between the promise and the page. Finally, test several variants of this page rather than freezing just one.
This is where new features combine: targeting identifies the right moment, and page testing verifies the message resonates. Without this verification, you pay for intent without knowing if it converts. To go further on how your customers phrase their questions after a campaign, see our analysis of e-commerce conversations: understanding real customer questions.
What are In-Market audiences for search?
In-Market audiences target people actively searching for and comparing products or services similar to yours. Google uses machine learning and artificial intelligence to sort through available data and estimate when someone is in the buying phase. The AdWords interface had nearly 500 at the time of the announcement, each with a viewable overview.
These audiences also provide basic demographic insights. However, they need to be tested before allocating a significant budget. Their value is real: they have greatly aided display campaigns, and advertisers can now layer them onto search campaigns.
The cautious approach
Add the audience in 'bid only' mode with a 0% modifier.
Compare its performance against general traffic.
Increase bids on audiences that outperform.
Consider dedicated campaigns to further refine the message.
To cover searches related to your offer, complement with solid SEO work: optimizing your e-commerce site for Google.
Why abandon last-click in your attribution?
This is the most significant change. Google Attribution is a major effort to move beyond last-click attribution, the current default setting in AdWords, Google Analytics, and most other channels. The product is available to all AdWords advertisers.
It allows you to choose from several attribution models: data-driven, last click, first click, linear, time decay, and position-based. You then link this model to a conversion. If you use automated bidding in AdWords, this model helps the algorithm make better decisions.
The data-driven model was previously reserved for Google's paid offerings, such as Analytics 360 and the premium version of DoubleClick. It is now free, but with conditions: at least 20,000 clicks on the account, and a total of at least 800 conversions in the last 30 days. These thresholds likely exclude many small and medium advertisers.
It is easy to see why Google requires this data volume: without it, there is no reliable attribution, and a high risk of distortion. The key takeaway remains the same: last click has many weaknesses, and getting used to fairer models is becoming a necessity.
How to read an attribution report without making mistakes?
Changing attribution models is not a snap decision. Before switching, take time to compare what each model tells you about your customer journey. A search campaign may start the conversation, a display campaign may re-engage it, and an email may close it: last click will only credit the final link.
Specifically, note the differences between the last-click model and a linear or position-based model. If a channel suddenly gains weight, it may deserve more budget than today. Conversely, a channel overvalued by last click might be consuming resources without delivering its share of results.
Keep in mind that attribution feeds your automated bidding. A poorly chosen model sends bad signals to the algorithm, which then optimizes toward the wrong objective. To properly link your customer data to this analysis work, rely on a clean foundation: using the right customer support data.
Should everything be activated at once?
No. The temptation is strong to launch all four new features on the same day, but that would be the best way to understand nothing about the results. Testing one variable at a time remains the golden rule for interpreting what actually produced the observed effect.
A reasonable order exists. Start with page tests, as they directly impact conversion rates and do not rely on high data volumes. Proceed with In-Market audiences on search, initially in observation mode, to identify responsive segments. Add life event targeting when a specific occasion matches your catalog.
Leave attribution for last, once your campaigns are stable and conversion volumes are sufficient. Changing models during the learning phase clouds the reading. Also note that the data-driven model imposes high thresholds: if your account does not meet them, it is not a failure, just a matter of timing.
What limits to monitor before spending?
Every new feature has its shadow, and ignoring it is costly. In-Market audiences require proof before absorbing a significant budget: without a testing phase, you are betting on a hypothesis. Life event targeting, for its part, speaks only to a portion of your audience, depending on the nature of your products.
Data-driven attribution remains conditioned on a high volume of clicks and conversions. Many merchants will therefore have to settle for other models, less precise but already far fairer than last-click. And Google Optimize, despite its simplicity, requires time: creating variants and analyzing results is not done in an afternoon.
Habits to keep
One test at a time, with a quantified objective.
Sufficient duration before concluding.
A verification of the reliability of collected data.
Good data is better than a sophisticated model poorly fed.
How to link these new features to your customer questions?
A successful campaign does not stop at the click. Once the visitor arrives, they have questions: delivery times, returns, availability, compatibility. If you have targeted a life event or an in-market audience, these questions will be even more precise, as the purchase context is stronger. Your pages must answer them directly.
This is where advertising work meets customer service. The same formulations that trigger a click must appear in your responses and documentation. A customer arriving via an ad for a baby gift should not encounter generic information.
Two levers help: reduce ticket volume upstream, and transform responses into useful content. On the first, see reducing e-commerce tickets with AI. On the second, integrating your customer service responses into an e-commerce SEO strategy. And if an offer was pushed offline, keep the same thread: managing customer questions for an offer seen in an offline ad.
How to measure the real gain on your sales?
The real gain is not read in the advertising interface alone. It is measured across the full chain: acquisition cost, conversion rate, average order value, and revenue per visit. A page test that gains two conversion points is immediately visible on acquisition cost, without changing bids.
For targeting, compare the behavior of targeted segments to that of general traffic: higher basket, lower return rate, better recurrence. For attribution, the effect is more subtle: it does not change the sale, it changes the budget decision. A fairer model reallocates your spending toward what truly drives purchases.
Follow these indicators over time, not just for a week. Automated bids need time to learn. And to link the quality of your product recommendations to the average order value, keep an eye on smart product recommendations.
How does Qstomy extend these advertising gains?
Qstomy is an AI agent for Shopify stores. It guides the customer toward purchase, manages recommendations, cross-selling and up-selling, the cart, package tracking, and after-sales service. Over 100 merchants use it. Where AdWords new features act before the click, Qstomy takes over once the visitor arrives.
The link is direct. You invest to attract a precise audience: Qstomy helps convert them by answering questions at the right moment, and increases the basket with relevant suggestions. Package tracking and customer service then limit bad surprises, which cost negative reviews and returns.
The stakes are clear: do not pay for a click your store cannot handle. An agent who responds quickly, gives accurate recommendations, and reassures customers about delivery transforms costly traffic into satisfied customers. To structure these interactions, rely on explicit rules: write a clear e-commerce support policy.
What checklist before activating these new features?
The checklist
Ensure conversion tracking is reliable before any test.
Prepare page variants with only one change at a time.
Add In-Market audiences in observation mode without modifying bids.
Check your click and conversion volumes before data-driven attribution.
Align ads, pages, and customer responses with the same message.
In short
Four new features, one thread: better targeting, better conversion, better measurement. Test in order, without activating everything at once.
Frequently Asked Questions
Is Google Optimize really free? Yes, it is offered free to AdWords advertisers, where competing tools charge for this service.
Is data-driven attribution accessible to everyone? No: it requires at least 20,000 clicks on the account and 800 conversions in the last 30 days.
Does life event targeting apply to all merchants? No, but many of these events involve a gift, which expands usage.

Enzo
October 28, 2026


