Prime Day Advertising Strategy: Manage Your History, Then Buy the Right Position
Prime Day advertising strategy in two parts: protect your price history and sales rank before the event, then predict conversion and buy the right position.
Last updated 11 min read
Advertising strategy for Prime Day, Prime Big Deal Days or any big Amazon sale comes down to two jobs. Before the event, manage the history you can still change: your price history decides which deal you can run, and your recent sales feed the rank you start from. Then predict how well clicks will convert at the deal price, and buy the right position at the right time instead of raising every bid. Start early: a deal is checked against up to 60 days of your price history, so list every promotion planned before the event next to the deal price you want to run.
Most event advice comes down to raising budgets and bids. That only pays when the extra clicks convert well enough at the deal price, and your normal-day numbers can't tell you that. During a big event your price is lower, your competitors' prices are lower too, and shoppers compare deals across the whole page.
Why a short sale event changes your advertising strategy
Prime Day and Prime Big Deal Days each run for two to four days of deals for Prime members. A short event limits what you can learn during it: if you spend the first day measuring, only the remaining days are left to act on, and the last of them is also the day your increases need to come off. So the conversion rate you plan around, the deal price you submit and the cost per click you will accept all need to exist before the event starts.
Strategy 1: Manage your history before the event
Two kinds of history are still in your hands in the weeks before the event: the prices your product sold at and the sales that feed its rank. Both build up slowly, so they have to be handled before the event, not during it.
Price history decides which deal you can run
When you create a deal, Seller Central shows a Max Price, the highest deal price that still qualifies. Amazon's rules for event deals build it from several inputs, including the lowest price your product sold at in the past 30 and 60 days. Every promotion and price cut in those windows becomes part of the history your deal is measured against. A promotion priced below your planned deal price can pull that ceiling under the deal, and a deal above its Max Price gets suppressed. Amazon's own tip is to check the deals dashboard for Needs Attention alerts starting two weeks before the deal.
Managing price history means protecting the deal you plan to run. It doesn't mean raising your price beforehand so the discount looks bigger: Amazon's pricing policies prohibit misleading reference prices, and a higher price also costs you sales in the weeks that build your rank.
Sales history feeds the rank you start from
The case for a lead-in promotion is sales history. Amazon's own explanation of Best Sellers Rank says recent sales count more than older ones, and that a product selling well has a better chance of ranking higher in both organic search and Best Sellers lists. Amazon doesn't say how much a short promotion moves either one, so treat a lead-in promotion as a bet on sales history, with no guaranteed ranking gain.
Track the result instead of assuming it. In the weeks before the event, record each deal product's Best Sellers Rank once a day, along with the ranks of the two or three competing products closest to it. That gives you a starting position to compare against during the event, and it shows whether the lead-in promotion moved anything.
If you run a lead-in promotion:
- Keep its price above the deal price you plan to submit, and check the Max Price again after it starts.
- Remember that deal discounts combine with other promotions running at the same time, so a coupon left running into the event deepens the discount your margin math assumed.
- Judge lead-in ad spend against your normal break-even ROAS, since the product isn't discounted yet.
Strategy 2: Predict conversion, then buy the right position at the right time
The history you can't change is the best forecast you have. Your past discounts and your last Prime Day show how much conversion rises at a deal price, and that number decides what a click, and a better position, is worth during the event.
How much does a lower price lift your conversion rate?
Your own past discounts are the best estimate you have. Conversion rate, the share of clicks that become orders, is the number that decides whether extra ad spend during a deal can pay, and it varies a lot between products and discounts.
For each product you plan to discount, find past promotions or deals and compare them with two or three ordinary weeks before:
- Conversion rate: orders divided by clicks, from the Sponsored Products advertised product report for the promotion days and the weeks before.
- Average selling price: ordered product sales divided by units ordered, from Business Reports in Seller Central.
- The ratio between them: how much conversion rose for the price you gave up. Leave out the first day of each promotion, which often reads low while the deal ramps up.
Then compare that history with what the discount requires. A discount comes out of the margin that pays for your clicks, so the ROAS you need to break even rises. Break-even ROAS, the ratio of ad sales to ad spend at which the margin exactly pays for the ads, is 1 divided by your margin after the discount.
| Margin before discount | Discount | Margin after discount | Break-even ROAS before | Break-even ROAS during deal | Conversion lift to hold profit per click |
|---|---|---|---|---|---|
| 40% | 10% | 33.3% | 2.50 | 3.00 | +33% |
| 35% | 15% | 23.5% | 2.86 | 4.25 | +75% |
| 30% | 20% | 12.5% | 3.33 | 8.00 | +200% |
| M | D | (M minus D) ÷ (1 minus D) | 1 ÷ M | 1 ÷ margin after discount | M ÷ (M minus D), minus 1 |
The margins are illustrative. At a 35% margin, a 15% discount means conversion has to rise about 75% before a click at today's bid earns the profit it earns today. If your past discounts lifted conversion by a fraction of that, the deal may still be worth running for sales volume, but it doesn't justify paying more per click.
How did your products and campaigns respond last Prime Day?
Your last Prime Day is the closest rehearsal you have, so read it product by product and campaign by campaign against the two weeks before. During a large event competitors cut prices at the same time you do and shoppers move toward whatever is on a deal, so each of your products reacts differently, and an account total blends those reactions together.
| Compare, per product and per campaign | What it tells you | Where to find it |
|---|---|---|
| Impressions and clicks | Which campaigns scaled with event traffic and which stayed flat or ran out of budget | Campaign report, by day |
| Conversion rate and average order value | Which products converted better, and whether a lower order value ate the gain | Advertised product report and Business Reports |
| Cost per click on the same keywords | What a click actually cost you, since account averages move when clicks shift between products | Targeting report, keywords with clicks in both periods |
| ROAS, split by products with a deal and without one | Where extra spend paid and where it didn't | Mark each product by whether its selling price dropped |
In one account we manage, conversion rose about 15% during a recent Prime Day, but several products were discounted, so average order value fell about 15% and ROAS ended flat. Split by product, the picture changed. Discounted products converted about 25% better and carried the gains. Products without a deal paid about 15% more per click while converting only about 10% better, so their ROAS slipped. None of that showed in the account's average cost per click, which barely moved.
When a discounted product converted worse than its discount suggested, look at what the products next to it charged during the event. A competitor that cut deeper takes the clicks your deal was meant to win. Plan the next event for that product with the lower lift it actually showed, and put those competitors on the list you will check on event mornings.
Use your own version of that table to plan the next event. Products and campaigns that paid off last time are where extra budget has a track record. Products that won't have a deal shouldn't get event-sized bid increases by default. And the conversion lift your discounted products showed last time is the number to check against the lift the deal math above requires.
Buy the right position at the right time
Once you know roughly how well each product converts at its deal price, put the extra money where that conversion shows up instead of spreading it across the account.
- Position: pay for the top slot only where your deal wins it. A top-of-search multiplier is the extra percentage added to your bid when the ad competes for the top of the first page of search results. In one campaign, raising it doubled the top-of-search share of impressions, from about a third to about two-thirds, while total impressions stayed flat. The extra money bought a better slot for roughly the same impressions. During an event, that slot is worth paying for on searches where your deal is the better offer on the page and the conversion lift supports the cost; elsewhere it mostly moves impressions you already had.
- Time: keep increases inside the deal window. A Lightning Deal runs for 4 to 12 hours, or until its units sell out, and your price is only discounted inside it. Any bid increase for that product belongs inside the window, within the cap below, and comes off when the deal ends; outside it you pay event prices for clicks at your normal price. A schedule-based bid rule raises bids only during the hours you set, or you can change bids by hand. If your deal runs for the whole event, the window is the event itself.
- Budget: keep deal campaigns running until the deal ends. Event traffic can spend a normal daily budget early in the day, and a campaign that runs out sits out every hour after that. Look at when your deal campaigns stopped spending on your last Prime Day, and set event-day budgets, by hand or with a schedule-based budget rule, so they last until the deal ends.
If you want an outside read of the account before the event, a free read-only audit shows where spend is wasted and which changes to make first.
What to do during and after the event
Leave bids where they are on the first morning, cap what a click may cost, and take every increase back the day the deal price ends. The same rules answer whether to raise bids during Prime Day or during any promotion you run yourself.
- First day: measure. The first day tends to read low, because orders from recent clicks are still being credited. In one deal we reviewed, bids went up on the first day: the account paid about 85% more per click for a price cut of about 3% and finished the deal well below its ROAS target.
- Every morning: check competitor prices. Search your top keywords the way a shopper would, and note what the products next to yours cost and whether they carry a deal badge or coupon. Where a competitor's price undercuts your deal, clicks on that search convert worse than your deal average, so don't raise bids there; put the increase on searches where your deal is the better offer.
- From the second day: raise only up to a cap. Max cost per click = normal cost per click × (deal conversion rate ÷ normal conversion rate) × (deal price ÷ normal price). If normal ROAS already sits below the deal's break-even ROAS from the table, cap lower still.
- The day the deal ends: revert. Increases sized for the deal price keep buying clicks at the normal price. Expect ad sales to drop when the price steps back up, and check what else to rule out when ad sales drop suddenly before changing anything more.
- The week after: wait. Judge the event after about a week, and keep the event days out of every baseline. Then read what the extra spend returned with marginal ROAS: the sales the additional spend brought in, divided by that additional spend.
What to check in the weeks before the event
Manage your history
- Deal eligibility. In the Deals dashboard, note each product's Max Price and any Needs Attention alerts. Check any promotion or coupon scheduled before or during the event against that price.
- Rank baseline. Record each deal product's Best Sellers Rank once a day, with the two or three competing products closest to it.
Predict conversion and plan position
- Past discounts. For each deal product, compare conversion rate and average selling price during past promotions with the ordinary weeks before, leaving out each promotion's first day.
- Deal math. Add a row per product to the table above using your real margin and planned discount, and compare the conversion lift it needs with what your past discounts delivered.
- Last Prime Day. Pull the campaign, targeting and advertised product reports for your last Prime Day and the two weeks before. For each product and campaign, compare impressions, conversion rate, average order value, cost per click on the same keywords and ROAS, split by products with and without a deal.
- Competitor prices. For each deal product, search its top keywords and list the competing products shown next to it, with their current price and any deal or coupon. Recheck the same list each morning of the event.
- Deal windows and budgets. Note when each deal runs, set the bid changes for those hours and the changes that undo them, and set event-day budgets that last until each deal ends.
- Reversal plan. Download a bulk file from Bulk operations now, so the reversal on the last day is one upload, and name the person who will do it.
Copy the deal math table into a spreadsheet, add a row per product, and put your past conversion lift next to the lift each deal needs. If you also want the account itself looked at, start with a free Amazon PPC audit.