The first Black Friday I ran a real store, I spent the whole month obsessing over the deal. The discount percentage, the banner, the email sequence, the ad budget. I did almost nothing to the catalog itself, because it worked fine in October and I assumed it would hold up. Then the traffic hit, and I watched a lot of expensively acquired visitors land on a collection page, fail to narrow it down to what they wanted, and leave. The offer was good. The finding was broken.
That is the lesson I keep relearning: on BFCM, discovery is the bottleneck, not desire. People arrive ready to buy, often on mobile, often from an ad or a gift-guide roundup or an AI assistant that pointed them at you. They have minutes, sometimes seconds, of patience. If your catalog cannot get them from landing to the right product fast, the discount does not matter. And the thing that makes a catalog fast to search is not design. It is data.
With a few weeks left, here is how I would prep, working backward from the sale.
Three to four weeks out: audit and fix the attribute gaps
Everything downstream depends on your products having clean, complete attributes, so start here while there is still time to fix things.
- Export your catalog and look at the structured fields, not the pretty page. For each product, do you have a specific product type, color, size, material, and the handful of attributes a shopper would actually filter on? Count how many products are basically blank underneath. It is usually more than you think.
- Find the fragments. The same concept stored three ways (Blue, blue, navy) quietly breaks filters and search. Pick one value per concept and standardize.
- Fill the gaps that matter for your category. A clothing store lives and dies by size, fit, and material. An electronics store by compatibility and specs. Prioritize the attributes your buyers qualify on, not every field that exists.
This is the slow, unglamorous part, and on a large catalog it is genuinely a week of work by hand. I stopped doing it manually and now let our Scurix Product Attributes AI app read each product and generate the missing attributes and metafields in bulk, which is the only reason my catalog is actually ready before the rush instead of half-done when it starts. However you do it, do not skip it, because the next two steps are only as good as the data they sit on.
Two to three weeks out: make the storefront searchable
Once the attributes are solid, turn them into something shoppers can use.
- Set up real filters on your collection pages. Filters are powered by product attributes and metafields. If a shopper can narrow by size, color, price, and one or two category-specific attributes, your conversion on a crowded collection page jumps. If they cannot, they scroll and bounce.
- Check your on-site search. Type the queries a real buyer would use, including qualifiers (red running shoes mens 10). Does it return the right products? Thin attributes usually mean bad search results, which is another reason the audit comes first.
- Build your BFCM collections now. Create the sale collections, gift guides, and under-$50 style pages ahead of time. Use attribute-based automated collection rules where you can, so they populate themselves and stay accurate as inventory moves.
- Tidy the mobile experience. Most of your BFCM traffic is on a phone. Make sure filters are reachable, images are compressed, and the path from collection to product to cart is short.
The goal for this stage is simple: a stranger who has never seen your store should be able to find a specific product in under a minute.
One to two weeks out: metadata and the search surfaces
Now make sure the machines that send you traffic can read your catalog too, because a meaningful share of your shoppers will arrive through Google Shopping, rich results, and AI assistants.
- Confirm your product schema is valid. Run key products through Google's Rich Results Test. Price, availability, brand, identifiers, ratings. This is what earns the rich result that out-clicks a plain link.
- Check your product feed. If you sell through Google or Meta Shopping, your feed quality depends directly on your attributes and identifiers. The audit you did in week one pays off here automatically.
- Make your products legible to AI shopping. Assistants answer buying questions by reading structured attributes, not marketing copy. The stores with complete, specific data get recommended; thin ones get skipped. The work is the same attribute hygiene, which is why it compounds.
- Leave time to recrawl. Search surfaces do not update instantly. Changes you make now need days to be recrawled and trusted. This is the real reason not to leave it to the final weekend.
The final week: freeze, test, and leave margin
In the last stretch, stop making structural changes and start pressure-testing.
- Freeze the catalog structure. No renaming collections or reworking attributes the day before. Let search engines settle on what you have.
- Do a cold run-through. On your phone, from an ad or an email link, try to buy three different products as if you were a customer in a hurry. Note every place you hesitate. Fix the worst ones.
- Double-check inventory and availability data. Nothing wastes BFCM ad spend like driving traffic to a product whose availability is wrong, or that is quietly out of stock.
- Keep an alert on the essentials. Low stock, high-value orders, payment issues. During the rush you want to know about problems in minutes, not when you check the admin the next morning.
The quiet advantage
None of this is as exciting as the deal itself, which is exactly why most stores underinvest in it, and exactly why it is an advantage if you do not. A good discount with a findable catalog converts. The same discount with a catalog shoppers cannot search just burns your traffic and your ad budget.
The best part is that this work does not expire on December 1. Clean attributes, working filters, and valid structured data keep paying off all year, through every search shift and every new AI surface that starts reading product data. Black Friday is just the deadline that finally forces the catalog prep you should have done anyway.
