In the last week of October, Meta changed three things in your ad account, and you didn’t change any of them.

Threads became a default placement in every active campaign on October 26. Product browsing switched itself on for ads that were already running on October 28. And Meta’s own Q3 numbers, published the week before, showed the average price per ad up 14% year on year. All of this landed with Diwali and Black Friday a few weeks out, on creative you had already tested.

That’s the lead story this month, because it changes what your dashboard means before it changes what your customers do. Below it: Google and OpenAI both switched on ads inside AI answers in India in the same week, the festive numbers that show where Indian demand actually moved, and a playbook for the one thing I’d do before scaling anything in November.

Operator’s Notes comes out once a month. This issue covers October.

The big move: your October numbers aren’t comparable to your September numbers

What happened

Three defaults flipped in the last ten days of October, none of them visible in a standard weekly report.

  • Threads became a default placement on October 26 (Confirmed). Every active campaign picked it up. Threads inventory runs at roughly 20 to 40% cheaper CPMs than Facebook and Instagram, and with placement exclusions being phased out of ad sets, you can no longer fully opt out at the ad set level. (GoodMorning)
  • Product browsing switched on for existing ads on October 28 (Confirmed). When someone taps your ad, Meta may now show an in-app product experience with personalised suggestions before, or instead of, sending them to your site. It was opt-out via a button in a notice emailed around October 14. After the 28th, it’s a per-ad toggle under Advantage+ Creative and Destination settings. There is no reporting column that separates in-app conversions from landing-page conversions. (GoodMorning)
  • Price per ad rose 14% year on year in Q3, with impressions up 10% (Confirmed). That’s from Meta’s own earnings release on October 21. Expect the festive auction to be the most expensive one you’ve run.

Meta justified product browsing with one number: businesses saw 3% more conversions on average, from an experiment across about 117,700 advertisers in August. No median, no distribution, no breakdown by vertical.

Why it matters

Each change is defensible on its own. Together, they break the one thing a performance team relies on in November: the ability to compare this week’s creative report with last week’s.

Think about what a creative leaderboard is actually measuring now. Threads pulls blended CPM down, so a campaign can look more efficient while its mix quietly moved to an inventory you never evaluated. Product browsing changes what happens after the tap, so a static that “doubled” its ROAS this week may be winning because Meta’s merchandising took over after the click, not because of the image. And the higher auction price means the same budget buys fewer impressions than it did in September.

The damage isn’t that any of these is bad. It’s that your automated rules, your scaling decisions and your creative briefs are all reading an input that changed, and nothing in Ads Manager tells you so.

There’s a second cost that doesn’t show in the account. Product browsing puts a Meta-built product page between your ad and your site. For brands with a thin or stale catalogue, that page is now a creative-performance input you don’t control. For brands with licensed or co-branded products, Meta choosing which products to show is a legal question, not just a performance one.

The India angle

This matters more for Indian D2C than it might appear. Diwali is on November 8 this year, three weeks later than 2025, so the festive peak sits exactly on top of these changes. Most Indian brands run Meta at far higher budget intensity in the Dhanteras-to-Bhai-Dooj window than at any other time, with automated rules doing the scaling. Those rules are reading changed inputs.

And there’s one Indian-specific wrinkle. Product browsing shows a Meta product page with personalised suggestions. If your catalogue on Meta doesn’t carry accurate COD eligibility, delivery estimates and pin code availability, you may be paying to show shoppers a product experience that can’t actually complete the order the way your site can.

What to do

  • Do this month: Open three active ads and check whether product browsing is on. Decide per account, not globally, whether to leave it on. If your feed is complete and your product pages convert well, keep it. If taps need to land on a specific landing page, switch it off.
  • Do this month: Annotate October 26 and 28 in every report. Any week-over-week creative comparison that straddles those dates is comparing two different post-click journeys.
  • Do this month: Pause automated rules that scale or kill creatives on ROAS or CPA thresholds for the next two weeks. Then look for creatives whose ROAS jumped without a matching move in CTR or CPM. A post-click change should move conversion rate, not click-through. That gap is the only tell you have.
  • Do this month: Pull a placement breakdown and see how much spend has drifted to Threads since October 26. Judge it on your real business metric, not on CPM.
  • Plan this quarter: Rebuild placement exclusions as account-level controls or value rules, so the logic survives the ad-set rollout.
  • Watch: Whether Meta ships a reporting column for product browsing. Until it does, treat creative-level ROAS as a blended number.

Platform desk

1. Ads inside AI answers arrived in India, from both Google and OpenAI, in the same week

Confirmed. Do this month.

What changed: On October 28, Google expanded ads in AI Mode to ten more countries including India, Australia, Canada and Singapore, and launched multi-turn ads in the US, which can appear during follow-up questions in a conversation. Any Search, Shopping or Performance Max campaign is now eligible, and that spend lands in your regular Search reports with no separate line. A week earlier, on October 21, OpenAI took ChatGPT ads to 89 countries, said more than 100,000 advertisers had used it, and made its Shopify integration available internationally. On October 30 it switched on AI creative enhancements by default for all campaigns that hadn’t opted out.

Why it matters: On the earnings call, Google’s chief business officer said sponsored product listings in AI Mode already get click-through rates comparable to organic ones. That’s a strong signal for a brand-new placement. But you can’t opt out of it and you can’t see it separately. Your Search ROAS now includes a placement you’ve never evaluated.

Do this: You can’t isolate AI Mode spend in reports, so watch the shape instead. If Search impressions rise while CTR falls and conversion rate holds, AI Mode inventory is probably entering the mix. For ChatGPT, check whether creative enhancements are on, and decide if you want OpenAI rewriting your headlines without review.

2. Google AI Overviews appear on almost every brand search, and the click data is finally in

Early signal. Do this month.

What changed: Last issue, AI Overviews had jumped to 80% of branded queries. Slate Research then tested 100 brands and found 93 showed an AI Overview for their own name. More interesting: 71% of the 652 sources Google cited were pages the brand does not own. Click data from an analysis of 21.7 million Search Console impressions puts the click loss for navigational queries at 8 to 13% when an Overview appears, against 55 to 61% for informational queries.

Why it matters: Brand traffic is eroding more slowly than other traffic, which is good news. The bad news is that the paragraph a customer reads about your brand is now mostly written from other people’s pages: review sites, Reddit, marketplaces, news. You’ve lost editorial control of your own name.

Do this: Search your brand name and your top three products. Write down every source the AI Overview cites. Those pages are now your brand page. Fix what’s wrong on them before you spend another rupee on brand search ads.

3. Meta’s new Commercial Terms slipped to November 20

Confirmed. Do this month.

What changed: Meta moved the effective date of its updated Commercial Terms and Self-Serve Ad Terms from October 30 to November 20, Reuters reported, after publishers and advertisers complained to EU and UK competition regulators that the terms give Meta a broad right to use their content for its AI services. Meta disputes that reading.

Why it matters: Every advertiser accepts these terms by continuing to run ads. If you use Meta’s AI creative tools or care how content from your website and ads trains Meta’s models, this is the one legal document worth reading this month.

Do this: Have someone read the updated Commercial Terms before November 20. For agencies, this is a client conversation, not an internal note.

4. Customer Lifecycle Strategy opened to every Meta advertiser

Confirmed. Do this month.

What changed: As of October 6, any ad set can be declared as acquisition, retention or both, and Meta will exclude existing customers from acquisition using pixel, CAPI and uploaded list data. (GoodMorning)

Why it matters: This is the most useful October release and it got the least attention. Meta’s delivery buys the cheapest conversions available, and your existing customers are almost always the cheapest. An “acquisition” campaign on default settings can post a healthy ROAS while mostly reselling to people who were going to buy anyway.

Do this: Write down your new-customer CAC first. Then switch one acquisition ad set to “acquire new customers”, not all of them. Blended ROAS will fall, because the cheap repeat conversions leave the campaign. New-customer CAC becomes real for the first time. If nobody writes down why the ROAS dropped, it will look like a performance problem on Monday.

5. Festive e-commerce: the basket has changed shape

Confirmed. Plan this quarter.

What changed: Redseer’s read of the first two days of the festive sales on October 11: roughly ₹24,000 crore in GMV, 8 to 9 crore orders, 60 to 65 million shoppers, two-thirds of them from Tier 2 and rural areas. First-time shoppers doubled on last year. Gen Z was about a third of opening-day buyers. Amazon and Flipkart’s quick commerce arms ran at 4 to 5 times their normal daily GMV. For the full season, Redseer projects $15 to 16 billion, with grocery growing around 50%, beauty 35 to 40% and quick commerce more than doubling, while mobiles’ share falls from 33% to 29%.

Why it matters: The festive buyer is younger, further from the metros, and more likely to be new to online shopping than ever. That’s a different person from the one most D2C creative is written for.

Do this: Pull your festive new-customer cohort by city tier and age. If two-thirds of the market’s new shoppers are outside the top eight cities and yours aren’t, your targeting is narrower than the opportunity. Then plan the January retention flow for this cohort now. First-time online shoppers churn fastest, and the brands that win them twice are the ones that planned the second purchase before the first one shipped.

6. Blinkit posted ₹20,940 crore in quarterly order value, and raised its store target again

Confirmed. Plan this quarter.

What changed: Eternal’s Q2 FY27 results (July to September 2026) on October 21 showed Blinkit’s net order value up 79% year on year, 398.9 million orders in the quarter, and 308 net new dark stores taking the network to 2,751. Adjusted EBITDA was positive for a third straight quarter. The company now targets 3,500 stores by March 2027 and about 5,000 by March 2028. Meanwhile, Datum Intelligence estimated brands would spend ₹2,200 crore on quick commerce ads this festive season, up 50% year on year.

Why it matters: Blinkit is now profitable at the EBITDA line while adding more than a hundred stores a month. That changes the negotiation. A platform that needed brands to fund its growth is becoming one that brands need for distribution, and the ad rates will reflect that.

Do this: If your q-comm ad spend is up sharply this festive season, run the same incrementality discipline you’d run on Meta: hold back a few cities, compare sell-through, and work out what you’re really paying for a rupee of incremental sales. And watch regulation: FSSAI is moving to hold platforms liable for misleading food claims, which means platforms will start policing your listings harder than they have.

The number: 71%

That’s the share of sources Google’s AI Overview cited about a brand that the brand doesn’t own, in Slate Research’s October test of 100 brand-name searches. The brand’s own site usually got the first citation and little after it. Review sites, Reddit threads, marketplaces, news coverage and comparison pages wrote the rest.

Here’s what I think the number means. For twenty years, searching a brand’s name took you to the brand. Now it takes you to a summary of what everyone else says about the brand, with the brand’s own site as one voice among seven.

Most marketing teams still treat their website as the brand page. The AI treats the internet as the brand page. Your reviews on Amazon and Nykaa, the Reddit thread from 2024, the comparison article a competitor paid for: that’s the paragraph a new customer reads before deciding whether to click.

The practical implication is that reputation work and SEO have merged. The page that most needs fixing is probably one you don’t own. Pick the three sources Google cites most for your name and treat each like a landing page: accurate, current, and telling the story you’d tell.

Overhyped / underrated

Overhyped: Meta’s AI assistant that “runs your ads.”

It was the headline of Advertising Week on October 6: an assistant that creates campaigns, changes targeting and moves budget from a chat prompt. It’s also still in test, it stages every change for approval, and Meta’s supporting claim (5% more conversions for advertisers who follow its recommendations) was published without a method, a comparison group or a time period.

My objection isn’t that it won’t work. It’s structural. The platform that spends your money would also be the one judging the result and adjusting the spend. Every step is reasonable on its own. The combined position is that nobody outside Meta ever looks at whether Meta’s delivery is good. Test it by all means. Don’t let it become the only opinion on your own account.

Underrated: the new Indian festive shopper

Two-thirds of the 60-odd million people who shopped on the opening two days of the festive sales came from Tier 2 cities and rural India, and first-time shoppers doubled on last year. That line got one sentence in most coverage. It should be the start of a planning document.

These shoppers found online buying through a phone discount or a value-fashion deal on Meesho or Flipkart. They’ll arrive at your brand next year through a quick commerce app or an AI answer. The brands that win in 2027 are the ones that stop writing for the South Delhi and Bandra customer and start designing for the Lucknow one: price points, pack sizes, language, COD trust, and a delivery promise that holds outside the metros.

The playbook: the pre-scale check for a changed account

The single most expensive mistake available this month is scaling a creative because its reported ROAS jumped after October 26. Here’s the 30-minute check I’d run before moving budget to anything on Meta between now and the end of the year. Save it, and run it every time a platform changes a default under you.

Step 1: Confirm what changed in the account

  • Open three active ads and check Advantage+ Creative and Destination settings. Is product browsing on?
  • Pull a placement breakdown for October 20 to today. What share of spend is now on Threads?
  • Note the date of any automated rule that fired since October 26.

Step 2: Separate the real winners from the artefacts

For every creative whose ROAS rose more than 30% week over week, fill in this line:

Creative · CTR change · CPM change · Conversion rate change · Placement mix change

Then apply two rules:

  • A real creative winner moves CTR first. People click more because the ad is better.
  • A post-click change moves conversion rate while CTR stays flat. That’s product browsing or a destination change, not your image.
  • A CPM drop with no CTR change usually means the placement mix moved toward Threads.

If a creative’s ROAS jumped without a CTR move, it’s not a winner yet. It’s a question.

Step 3: Rebuild the comparison

  • Annotate October 26 and 28 on every dashboard, including the ones your agency or client sees.
  • Compare post-28 performance only to post-28 performance. Build a fresh two-week baseline before judging anything.
  • For one acquisition ad set, switch Customer Lifecycle Strategy to “acquire new customers” and record new-customer CAC before and after. Expect blended ROAS to drop. Write down why before anyone asks.

Step 4: Check the destination you don’t control

If product browsing stays on, Meta’s product page is now your landing page for some share of taps.

  • Every hero SKU in the catalogue has a current price, stock status and image.
  • For India: COD eligibility and delivery estimates are correct in the feed.
  • Compare site sessions against reported link clicks for the week. A widening gap tells you how many taps are resolving in-app.

Step 5: Decide, and write it down

One line per account: Product browsing: on/off. Reason. Date. Owner. The brands that get hurt in November won’t be the ones that chose wrong. They’ll be the ones that never noticed there was a choice.

One question

In the last issue I asked whether you’d let an AI agent buy your product without ever visiting your site. The most common answer was a split: yes for repeat purchases, no for the first one. A few of you argued the opposite, that the first purchase is exactly where you need the agent, because that’s where friction kills the sale. I’m still thinking about that.

This month’s question is simpler. When a platform changes a default in your account without asking, who on your team is responsible for noticing?

If the honest answer is “nobody, specifically”, that’s the gap to close before December. Tell me how you’ve solved it, and I’ll share the best systems next month.

If this was useful, send it to whoever reads your creative report on Monday mornings.

Until next month,

Shiva.

Sources

Shiva Sharma is a global growth marketing leader and the author of The Movement and Nobody Is Thinking About You. More about Shiva

Read next

How to set break-even ROAS for every campaign A step-by-step method for working out the ROAS each campaign needs to make money, from your own margins, and using it to decide what to fix, cut and scale. How to compare couriers on your own data Rate cards tell you what a courier costs. Your orders tell you what it really costs. A method for ranking couriers on delivered rate, COD RTO, speed and cost per delivered order.