Martech trends for November 2026: what AI rules out, what Google keeps and what shoppers remember

Here are five changes from the past few weeks worth taking into December planning.

1. The shortlist now gets written before the visit

Most of the conversation about AI search has focused on whether a brand appears in the answer. New data suggests the more expensive moment comes earlier, when the assistant helps a shopper decide what to leave out. In research from Attest, 62% of US AI-assisted shoppers ruled out at least one product based on AI’s advice without checking elsewhere. The same research found that AI introduced 81% of US AI-assisted shoppers to products and services they were unfamiliar with, while seven in ten US AI-assisted purchases (73%) went to a brand the buyer already recognised.

AI widens the field and narrows it within the same conversation. A product can be discovered and dismissed before a single page view lands in analytics. The reason may be an outdated specification, a misread review or a description that never answered the question the shopper actually asked.

Marketing leaders are starting to treat this as brand management. According to Gartner, 42% of CMOs have taken steps to monitor and correct misinformation about their brands in AI-generated output, and 61% have adapted content strategies to improve machine readability and visibility in AI-driven experiences. Consumer research from Accenture earlier in the year points the same way: 37% of behaviourally loyal consumers would let an agent switch brands for a better fit.

For an online store, most of the work sits in product data and on-site content. That means specifications that are complete and current, comparison information written around the questions buyers really ask, and merchandising decisions that someone has actually written down. The Attest figures are US-only and self-reported. Even so, they describe a stage of the journey that most dashboards cannot see.

2. France gets AI Overviews, and Europe gets a clean before-and-after

In July, Google switched on AI Overviews and AI Mode in France, closing the last major gap in its European rollout of AI-generated search results. The launch had a single date and the reporting was already in place: Google added generative AI reports to Search Console on 3 June 2026, first for a subset of websites, and then extended them to every site as of 31 August 2026. As a result, France has given Europe its clearest before-and-after picture so far. In August we looked at the first official numbers on AI visibility. France shows what that visibility costs in clicks.

Ahrefs tracked 963 domains through the launch. The domains most exposed to AI Overviews lost more than a fifth of their click-through rate in nine days, and 82.2% of highly exposed domains lost CTR. Across the whole panel the losses were smaller: the median was −5.7% CTR and −8.8% clicks, and the pattern was close to linear, with roughly one point of CTR lost per point of exposure.

The averages hide a very uneven picture. Data published by search consultant Glenn Gabe in early November covered five French sites. Three of the sites saw clicks or click-through rate fall by roughly 45% to 70% almost overnight, by PPC Land’s calculation, while two saw little change. Retail looks less exposed than publishing so far: an Ahrefs analysis for mind Retail put the e-commerce click-through rate down by 5.4% since Google AIO launch in France.

For e-commerce teams elsewhere in Europe, a useful exercise is to compare the share of their own queries that trigger an AI Overview with the click trend on the same pages. The French window is short, so the long-term effect could turn out larger or smaller. Either way, organic traffic forecasts built on last year’s click-through rates have become harder to defend.

3. Consent moves from the cookie banner into the inbox

Two dates in October changed the consent picture for European marketers. On 21 October, EU member states agreed the Council’s position on the Digital Omnibus, the package that rewrites parts of the GDPR and the ePrivacy rules. The mandate keeps cookie rules in a rewritten Article 5(3) of the ePrivacy Directive, rather than moving them into the GDPR as the Commission proposed. Consent exemptions include first-party audience measurement, media audience measurement, technical security, and measuring contextual advertising, and after a refusal, a provider must wait at least four months before asking again for the same purpose. This is the Council’s negotiating position, and Parliament has not yet adopted its own, so the final text may still change.

Eight days later, Italy’s data protection authority began applying its rules on email tracking pixels. By 29 October 2026, anyone who uses email tracking pixels to reach people in Italy, the tiny images that tell a sender when a message has been opened, must comply with new guidelines from Italy’s data protection authority. In practice, senders now need notice, consent and an easy way to withdraw it before tracking opens. France’s CNIL reached the same position earlier in the year, and the Garante has said that checking how the guidelines are applied is, in fact, part of the inspection plan the authority has set for the second half of 2026. Across all three changes, measurement that touches the recipient’s device is treated as a permission of its own, separate from permission to receive the message.

In Manago AI, email tracking consent is handled as its own setting. There is an account-level switch, a tracking status on each contact and a tracking strategy on each sender account, and the status sits next to the existing marketing consents. Contacts who have never given tracking consent start as opt-out. For those recipients, a sender can remove tracking from the email altogether or choose a mode that keeps campaign-level counters and leaves the activity off the contact’s profile. The full setup is explained in our article on the new email tracking rules in France and Italy.

4. Black Friday meets a shopper with a price history

Shoppers are still buying this season, but more of them arrive knowing what the product cost last month. In Feedvisor’s survey of 1,006 US consumers, 45% are comparing prices more than they did a year ago, 43% think deal quality has declined or feels artificial, or report sale fatigue, and nearly 60% would use AI to track historical prices and request retroactive price matches.

The tools for that already exist. In the US, Amazon’s shopping assistant can track prices, create deal alerts, and auto-buy items when they hit your target. A discount built on a price that was quietly raised a few weeks earlier is now easy to spot for anyone whose assistant keeps price history.

Most shoppers still want to complete the purchase themselves. The same survey found that 62% would never let an autonomous agent order and pay without manual checkout. For now, the part being handed over is the checking: watching prices, comparing offers and flagging a promotion that is not what it claims.

That changes what a promotional calendar needs to protect. Adobe expects US discounts to build through the season, with 14% off in early November, 21% off just before Thanksgiving, and 30% off once Cyber Week hits. Shoppers who follow that rhythm will notice a brand whose prices move the wrong way first. In the EU, the rule on the lowest price in the previous 30 days already requires reductions to be calculated honestly. Price-tracking assistants give every shopper a way to run the same test, at any time of year. Our Black Friday preparation guide covers how to check discounts before campaigns go live.

5. Every agent action now arrives with an invoice

Forrester’s predictions for 2027 put a number on something many marketing teams already feel. The firm expects that B2C marketing leaders will cut about one-third of their AI initiatives because costs are rising, partly because usage-based pricing from martech vendors makes each agent interaction and automated workflow a billable event. In its latest survey, 82% expect to provide stronger justification for future AI investments, and 78% say executive leadership demands stronger evidence of AI value than it did 12 months ago.

That pressure lands on a measurement gap. MarTech reported at the end of October that most teams know what they want AI to improve, but few define the metric, record the starting point, or set a clear threshold for success. Under flat pricing, a vague result is an annoyance. Once every action is metered, it becomes a line in the budget review that nobody can defend.

The initiatives most likely to survive are the ones that began with a baseline: a conversion rate, a response time, the hours a report used to take. Before adding the next capability, it is worth asking how a platform’s AI will actually be evaluated and what it will cost once usage doubles.

Before the December planning meeting

December planning usually starts with last year’s results and next year’s targets. This year, three questions deserve a place on the agenda before either.

What does an AI assistant currently say about our best-selling products, and would we agree with it? How much of our email and website measurement still works once consent rules apply in full, in Italy, in France and wherever the Digital Omnibus finally lands? And which of our AI projects started with a number we can compare against today?

Answering them takes no new tool. It takes someone willing to look at data the business already holds, on the assumption that a shopper, a search engine, a regulator or a finance director may be looking at it too.