Days 31 to 90: content, feeds and AI

Day 31. Every product has one URL, the best sellers have real titles and valid schema, and the keyword map says which collection each guide should feed. What the store still lacks is content that catches buyers before they know which product they want, a feed that puts the products in front of Shopping, and the reviews and mentions that make an assistant name you when someone asks what to buy. The next sixty days add those three things in that order, because each one depends on the last.

Why guides first, then the feed, then the signals

Which page catches which search gives the rule: category searches go to collections, product searches to products, questions to guides. Month one fixed the first two. Guides come next because a buyer who types "trail shoes for wide feet" isn't ready for a product page yet, and a guide that answers the question and then hands them the collection is the page that catches them. It also passes the links and trust it earns down to the collection, which is the point of Internal links to money pages.

The feed comes second because it depends on month one's product work. Google's Merchant Center help describes free listings: products submitted in a feed can appear in Shopping results without an ad, provided the feed meets its product data requirements, including prices and availability that match the landing page. Broken product pages produce a broken feed.

Signals come last because reviews accumulate over weeks and a third party needs something worth mentioning. Being the product AI recommends puts clear product pages and consistent specs ahead of mentions for that reason.

Weeks 5 to 8: publish the guides that feed collections

One cluster per top collection: a buying guide, a comparison, and the questions buyers ask, all linking to the collection and to the products they discuss. Take collections in the order of revenue the baseline showed.

1. Week 5: brief and publish the first buying guide using Buying guides that sell. It answers the question a buyer asks before choosing, then hands them the product.
2. Week 6: publish the comparison, following Comparison and FAQ content. "X vs Y" and "is X worth it" searches sit late in the buying process. Write the FAQ for the reader and for AI retrieval, not for a results-page feature, because Google's Search Central updates record that FAQ rich results ended in May 2026.
3. Week 7: the second collection's guide, plus a short questions section on the best-selling product pages.
4. Week 8: the link pass. Content-to-cart internal links puts a link to the collection or product at the point where a reader is ready, not only at the end. Shape each cluster as Topic clusters describes, and structure the pages as Writing for AI citations recommends: a direct answer in the first screen, headings that read as questions, sections that stand alone.

In place by day 58: two clusters live, each guide linking to its collection and products, and old posts linked to the collections they mention.

Weeks 9 to 10: clean the product feed

Product feeds and free listings covers what the feed must contain and the errors that block it. Most platforms supply the feed through their own Google channel app or plugin, so the work is checking rather than building.

1. Week 9: set up Merchant Center, connect the feed, and work through every disapproved and warned item. Missing identifiers, images that break Google's requirements and prices that don't match the page are the usual causes. Variants and duplicate content explains why each variant is its own feed item even when the page is shared.
2. Week 10: align feed and page. Price and availability in the feed must match what the page shows, and Merchant Center's automatic item updates, documented in Google's Merchant Center help, read the Product structured data you validated in month one to correct mismatches. Products that sold out or were discontinued follow Out-of-stock and discontinued products in both places.

In place by day 72: a feed with its items approved, prices and availability matching the pages, and a weekly check on new disapprovals.

Weeks 11 to 12: earn the signals Google and AI use

Reviews first, because they take longest to accumulate. Reviews, UGC and brand signals sets up a guideline-safe request routine after delivery. Mark up genuine customer reviews of a specific product, and nothing else: Google's review snippet guidelines make a business rating itself ineligible, so the markup belongs on product pages, not the homepage.

Then consistency. Being the product AI recommends and AI shopping answers explain that assistants name products described the same way everywhere. Check that names, specs and prices match across your site, your feed, marketplaces and any retailer that stocks you, and that your brand is described consistently, as Brand mentions and entity SEO recommends.

Finally, mentions. Pick one asset worth citing, a buying guide with a point of view or data from your own orders, and send it to the publications, communities and partners that cover your product category, following Digital PR for small teams and the test in Backlinks that matter. Google's spam policies treat bought links as link spam, so this is earned or not at all.

In place by day 86: a review routine running, specs consistent everywhere you're listed, and outreach sent and logged.

Days 87 to 90: measure against day one and set the next quarter

Put the baseline sheet beside today's numbers, using Tracking revenue from search and AI: organic sessions and revenue per collection, the clicks Merchant Center reports for free listings, and the tracked keywords read as a trend. Repeat the AI snapshot with the same assistants and the same questions in the same words, following Measuring your AI visibility.

Claim only what the data supports. The newest guides have been indexed for a few weeks, the feed has only just settled, and Reading rank movements honestly guards against crediting yourself for noise. Write one page: what's live, what moved, what didn't, and what you don't know yet.

Then rerun The 30-minute store quick-wins audit, take the next collections from the keyword map, and keep the feed check, the review routine and the weekly review from A weekly SEO review with Claude running. The plan is a cycle that has completed its first turn.

In place by day 90: a report against the baseline, the next collections chosen, and the routines still running.

What to take away

• Guides catch buyers before they know the product, and their job is to hand the reader and the links to a collection.
• The feed only works when price, availability and images match the product pages, which is why the feed work follows the product fixes.
• Reviews, consistent specs and earned mentions are the signals assistants use, and they build up over weeks rather than being bought.
• Measure days 31 to 90 against the day-one baseline, claim only what the data supports, and run the loop again.

Next

The plan repeats from here, and the mechanics under every step are in the SEO + AI Search Academy, starting with How Google ranks a page.