CarParts.com is opening its marketplace and ad network together in December. A countdown for retailers planning the same

What CarParts.com announced

On October 22, 2026, Mirakl published the news that CarParts.com, the online auto parts retailer, plans to launch CarParts Connect, a marketplace and retail media network running on Mirakl's platform, expected in December 2026. It brings four pieces together at once: marketplace operations for recruiting and managing sellers, catalog management with AI-driven matching of seller product data, retail media such as sponsored listings and banner ads, and seller payouts.

The new categories are telling: vehicle accessories, tools, travel and road-trip essentials, pet and passenger products, and outdoor gear. The marketplace isn't there to sell more brake pads. It's there to sell the things someone buying brake pads also needs for the trip.

Mirakl's post argues that launching the pieces separately costs more than it looks: a catalog built before the marketplace may not fit what sellers send, adding retail media later means onboarding your best sellers twice, and payouts handled outside the platform leave finance reconciling by hand.

We agree with the argument. Launching everything together only saves work if the decisions are made in the right order.

If you're a retailer thinking about a marketplace for next year, here's that order, counted back from a launch date. The weeks below are our illustrative plan, not CarParts.com's.

Sixteen weeks out: who gets in

The first decision is about sellers, and it comes first because everything else depends on it. Mirakl's post puts it as a question: who will you let in, and how will you check them?

Write down:

• Who qualifies. Business verification, insurance, return rates, shipping speed, the categories they're allowed to sell in.
• How you check. Documents, a test order, references, a probation period with lower limits.
• What happens when they slip. Late shipments, wrong items, policy breaches. Decide the steps before you need them.

This is also the week to decide whether sellers are invited to advertise during onboarding. Mirakl recommends bringing sellers into retail media in the same onboarding, and it's much easier to design one process than two.

We've written about what the first suppliers need from onboarding. The short version is: be specific, be quick, and tell them what good looks like.

Twelve weeks out: the catalog

The catalog takes the longest, and it's where most marketplace launches lose time. Sellers send data in their own formats, with their own names for things. It has to be mapped to your categories and attributes, matched to products you already list, and checked before it goes live.

Mirakl's post makes a point specific to car parts that applies more widely: universal parts fit every vehicle, but everything else has to be matched to the vehicles it fits. Any category with compatibility has the same problem: printer ink, phone cases, replacement filters, bike components. If your customers ask "will it fit?", this is the decision that needs the most time and the most care.

Decide now:

• Which attributes are required for each category, and which are nice to have.
• How matching works, automated, manual or both, and who reviews the matches the system isn't sure about.
• What blocks a listing from going live, especially missing compatibility data.

A practical test: take fifty real products from three sellers you'd like to recruit, in whatever format they'd send them, and run them through your mapping. Count how many go live cleanly, how many need a person, and how long that person takes. That one exercise tells you more about your launch date than any project plan, and it gives you a real answer when a seller asks how quickly their products will appear.

Eight weeks out: where the ads go

Retail media is the revenue line that makes many marketplace business cases work, and it's the one most likely to annoy customers if it's placed badly. Mirakl's question: decide where sellers and brands can advertise on your site and how ad placements fit with search and product pages.

Settle:

• Placements. Which search results, category pages and product pages carry sponsored listings, and how many per page.
• Labeling. How sponsored items are marked, consistently.
• Relevance rules. Whether a sponsored listing can appear for a product it doesn't fit. In a fitment category, an ad for a part that doesn't fit the customer's car is worse than no ad.
• Who can buy. Only marketplace sellers, or brands whose products you sell directly too.

Four weeks out: money

The last of the four decisions is the one that keeps sellers. Mirakl's question: what's your payout cycle, and how much cash will you hold in reserve so returns can be paid out smoothly?

• Payout cycle. Weekly, every two weeks, or monthly, and how many days after delivery.
• Reserve. How much of each seller's balance you hold back for returns and disputes, and when you release it.
• Commission and fees, stated in one place, with examples.
• What happens when a payout fails. We've written about what to do when payouts stop; the time to design that process is before the first one fails, not after.

Finance should sign off on this week's decisions before any seller sees a contract. Customer service should see them too, because they'll be the ones explaining to a shopper why a refund from a marketplace seller takes a different path than one from you.

Launch week, and the week after

By launch, the decisions above should be settled and tested with a small group of sellers. Launch week is for watching:

• Listings rejected by the catalog checks, and why.
• First orders from each seller, and whether they ship on time.
• Ad placements on the busiest pages, and whether they look like your store.
• The first payout, end to end, including a return.

The week after is for fixing the process that broke. One will. The point of the countdown is that it breaks in week one with ten sellers, not month three with two hundred.

If you're weighing a marketplace for next year, or you've launched one and the catalog or payouts are taking more hands than they should, our marketplaces team builds and fixes these operations. Get in touch and tell us which of the four decisions you're least sure about.

Related work: Spin Market.