Amazon PPC dayparting: what it is, when it pays, and how to start

Dayparting means changing your bids or budgets by the hour, so you spend more when shoppers are likely to buy and less when they're browsing. Amazon doesn't offer it as a simple switch for most sponsored ads, so it usually means a tool or a person making scheduled changes.

It's worth doing. But it's often sold as the fix for problems it doesn't solve, so here's an honest version.

Start with the problem dayparting actually fixes

When we pulled the change history on our own account, two things stood out:

• Changes happened once a day, in one batch. About 80% of all logged changes landed between 2pm and 4pm Pacific. Nothing changed in the morning or the evening, whatever shoppers were doing.
• Budgets ran out early. One of our audience campaigns ran out of budget before noon on 13 of the 28 days we checked. For the rest of each of those days it showed no ads at all.

The second one is the expensive one. A campaign that goes dark at 11am misses the evening, which for many consumer brands is when people buy. That's a pacing problem, and it's the first thing dayparting should fix.

The data you need first

To bid by the hour, you need conversions by the hour, per campaign. The standard search term and campaign reports in the Amazon Ads console give you days, not hours. Hourly data comes from Amazon Marketing Stream, which pushes hourly traffic and conversion data to an app you connect through the Amazon Ads API.

We're being straightforward here: as of writing, we don't have hourly conversion data for our own account yet, because it only starts flowing once you're connected to Marketing Stream. So we're not going to show you a chart of "our best hours." Anyone who shows you one without that data is guessing.

A sensible way to start

1. Fix pacing. Find every campaign that runs out of budget before evening. Either raise the budget on the ones that are profitable or lower bids so the budget lasts the day.
2. Collect two to four weeks of hourly data before changing bids by hour. One week is too noisy, especially on campaigns with a few orders a day.
3. Look at conversion rate by hour, not clicks. Clicks at 2am can be cheap and useless. What you want is the hours when a click is most likely to become an order.
4. Change in small steps. Raise bids 10% to 20% in your best hours and lower them in your worst. Big swings make it hard to tell what worked.
5. Keep the weekend separate. Saturday at 10am and Tuesday at 10am are different shoppers.

When dayparting doesn't pay

• Low-volume campaigns. If a campaign gets two orders a day, there isn't enough data to say which hour is better.
• When the real problem is targeting. If a campaign is paying for search terms that never convert, it'll waste money at every hour. Fix the negative keywords first.
• When bids are already near your break-even. Raising evening bids on a product with thin margin just moves the loss. See ACoS vs TACoS for setting targets per product.

Placement multipliers aren't dayparting

Amazon lets you raise bids for top-of-search or product-page placements. Many tools adjust these once a day and call it optimization. It's useful, but it's about where your ad shows, not when. A campaign can have perfect placement settings and still run out of money by lunch.

How Adtender handles it

Adtender connects to Marketing Stream, paces budgets so profitable campaigns run all day, and only starts shifting bids by hour once there's enough hourly data to trust. Every hourly change is logged with the numbers behind it.