How do I advertise on Amazon and my own store?

Marketplace advertising and advertising to your own store do different work, and the budget argument only makes sense once you have named the work.

Marketplace ads sit in front of people who have already decided to buy something in your category and are choosing between options. The intent is high, the comparison is brutal, and the win is usually decided by price, reviews, images and delivery speed. You are harvesting demand that already exists, including demand you created elsewhere.

Ads to your own store do two other things. They create demand among people who were not shopping yet, and they capture the customer relationship: the email address, the subscription, the second order at full margin. That is why a store channel with a lower reported return can still be the more valuable one.

So set the jobs explicitly. Marketplace campaigns defend your listings on your own brand terms, take category terms where your listing genuinely competes, and win the moments when someone is comparing. Store campaigns carry the brand work, the ranges the marketplace handles badly, bundles that cannot exist there, and everything aimed at people who will buy repeatedly.

Once the jobs are separate, the campaigns stop competing for the same credit. You are not asking which channel gets the order, you are asking whether each channel is doing the job you funded it to do.

A large share of marketplace sales for a known brand come from people typing the brand name, and many of those people learned the brand somewhere else. That overlap is real, it is hard to trace, and pretending it does not exist leads to the wrong budget decision.

Here is how it shows up. You run store side ads or social content, demand rises, and marketplace sales rise too, with the marketplace campaign taking the credit because it served the last ad. Turn the store side spending off and both numbers fall together, usually with a lag of a few weeks.

Two practical responses. First, keep defensive brand campaigns on the marketplace running but capped, because they are cheap insurance against another seller appearing above your own listing, not a growth channel. Judge them on share of your brand searches held, not on return.

Second, test the direction of the halo deliberately. Change store side spending in one region or one product family, hold it elsewhere, and watch what happens to marketplace sales of the same products over the following weeks. That is a rough test, but it is far more informative than any single platform's attribution report, because neither platform can see the other.

Also watch your own search terms. If marketplace ads are winning searches that used to come to your site organically, you are paying for traffic you already had. Branded searches that grow on both sides at once is the pattern you want.

The two channels have different economics, so a single target return on ad spend applied to both will overfund one and starve the other.

On the marketplace, the contribution per order is the price minus cost of goods, referral fees, fulfilment fees, storage and returns. Those fees are substantial and they come off before any advertising. Work out the break even return with the fees included, then set the target above it.

On your own store, contribution is the price minus cost of goods, payment fees, shipping and packing, and whatever you carry for returns. Shipping is usually the big variable, especially under a free shipping threshold. But the store also earns the second order, so it can carry a lower first order return if the repeat rate is real. Only claim that allowance if you can show the repeat rate in your own data.

Write both break even numbers down and put them next to the campaign targets. Two different ceilings, both honest.

Then allocate at the margin. Move the next increment of budget to whichever channel returns more contribution per extra dollar, not to whichever has the better headline number. In practice that usually means the marketplace absorbs money quickly up to the point where you have won the searches worth winning, and further growth has to come from the store side, from new products, or from a wider catalogue.

Inventory is the constraint that overrides all of this. If one channel runs the stock down, the other channel's ads keep paying for pages that cannot ship.

You cannot follow a shopper from one platform to the other, so stop trying. Use two numbers that survive the gap.

The first is total marketing cost against total revenue across both channels in the same period. Add marketplace ad spend, store ad spend, agency or tool costs, and, if you want the honest version, the marketplace referral and fulfilment fees, because those are the price of that channel's traffic. Compare that to total revenue from both. Track the ratio weekly and watch the direction rather than the absolute value.

The second is new customers. Count first time buyers on the store side, and use the closest proxy the marketplace gives you for new to brand purchases. Growth that only comes from existing customers rebuying is not the same as growth that widens the base, and blended revenue hides the difference.

Set a reporting rhythm that matches the lag. Weekly for spending decisions, monthly for the blended ratio, quarterly for the question of whether the mix is right.

Then keep the channels honest on data. On the store side you own the email list, the subscription, the browsing history and the ability to talk to buyers whenever you like. On the marketplace you own the listing and little else, so use packaging inserts, warranty registration and product documentation to invite people into your own channel where the platform's rules allow it. Every customer who moves across is one you can market to for free next time.

And keep both accounts in your own name, with your own billing, so the history stays yours whoever runs the campaigns.