Ecommerce Advertising: Buy Ecommerce Ads Traffic

The difference between e-commerce and most other verticals is that the conversion is a purchase, not a form submission. The visitor has to trust the store, find the product, accept the price and complete a checkout: four places where the funnel leaks, none of which the ad network controls.

What the traffic source does control is who arrives and how much they cost. That makes granular targeting and a per-country bid the two levers that matter most: the same product can be profitable in one market and hopeless in the next at the same bid.

How to plan the first e-commerce campaign

Start with one country, one device type and one product category. Broad campaigns across several markets hide which market is working, and in e-commerce the difference between two neighbouring countries is routinely larger than the difference between two creatives.

Choose products with a clear visual and a price that survives being shown in the ad. Native works because the creative can carry the product image and the price; if your product needs a paragraph of explanation before the price makes sense, it belongs in a longer funnel than a single click.

Set the budget you are willing to lose, not the budget you hope to earn. The purpose of the first campaign is a conversion-rate number per country, not profit.

What it costs to start

A store can measure itself faster than most advertisers, because the conversion is a purchase and the value of it is known to the cent. The minimum deposit of $50 buys enough clicks in a cheap market to see whether a product page holds a stranger.

What it will not tell you is margin. A first test usually finds the traffic price, not the profitable traffic price, and those are different numbers. Decide before you start what a purchase may cost you, and stop the test when the cost per purchase passes it rather than when the budget runs out.

Tracking that lets you read the result

Pass macros for country, device, operating system and traffic source into your analytics or tracker. Without that split the campaign reports one blended conversion rate, and a blended number in e-commerce is close to useless: it mixes markets with three-fold differences in purchase behaviour.

Measure to the purchase, not to the add-to-cart. Add-to-cart is a useful intermediate signal for creative testing, but the decision to raise or lower a bid should be based on cost per purchase and average order value in that country.

If you use a postback integration, verify that it fires on the confirmed order and not on the thank-you page load. The two diverge more often than teams expect, especially with local payment methods that redirect.

Creatives: product, price, reason to act now

Three elements do most of the work in this vertical: the product image, the price and a reason the visitor should not postpone. The reason can be a discount, a deadline or stock, but it has to be true: inflated urgency raises the click-through rate and lowers the conversion rate, which is the worst possible trade at a per-click price.

Show the product in use rather than on a white background. Catalogue photography works in a catalogue, where the visitor is already searching; in a feed the ad competes with content, and context sells better than a cut-out.

Rotate creatives on a fixed schedule for retargeting campaigns. The same audience seeing the same banner for the third week is where cost per purchase quietly rises without anything appearing to change.

Bidding by country and when to scale

Recommended bids differ between markets by an order of magnitude, and so does purchasing power. The right way to compare countries is cost per purchase divided by average order value in that country, not the bid.

Scale in steps rather than jumps. Doubling a working campaign's budget usually changes which inventory you win, so the result is not two times the same traffic. It is a different mix that has to prove itself again. Raising by 30 to 50 percent at a time keeps the change readable.

When a market stops responding to bid increases, the ceiling is the available inventory rather than the bid. The traffic table on this page shows what is available per country, so you can tell the difference between a bid problem and a volume problem before spending more.

Retargeting store visitors

Most first-time visitors from paid traffic do not buy on the first visit, and in e-commerce that is normal rather than a failure. Banner retargeting is the cheapest way to bring back the ones who reached a product page.

Separate the audiences: someone who abandoned a cart needs a different message from someone who viewed a category. A cart abandoner already accepted the price and stopped at checkout, which is a delivery, payment or trust problem, not a discount problem.

Cap frequency on retargeting harder than on acquisition. The same visitor seeing the same reminder ten times a day is an expensive way to be annoying.

Seasonality and how to prepare for it

E-commerce demand is not flat, and neither is competition for inventory. In the weeks around major sales events, recommended bids rise because more advertisers want the same impressions. A campaign that was profitable in October can lose money in late November on the same bid and the same creative.

Prepare two things before the season: the creatives, tested at a low budget while traffic is cheap, and the answer to how much you are willing to pay per purchase when bids rise. Testing creatives during the peak is the most expensive way to learn.

After the season, expect the reverse: cheaper traffic and a lower-intent audience. It is the right time to test new markets, because a mistake costs less.

E-commerce, lead generation and sweepstakes: what to expect

E-commerce asks the visitor for money, which is the hardest conversion on this list. It rewards products with a clear visual, a defensible price and working local payments.

Lead generation asks for contact details, converts several times better and moves the difficulty downstream to whoever calls the lead. Sweepstakes ask for one action and convert best of all, at the cost of audience quality.

If you are entering paid traffic for the first time with a store, it is worth running a lead-generation offer in parallel on the same traffic. It tells you whether the problem is the traffic or the store, an answer that saves a lot of budget.