A new listing with a coupon slapped on day one looks like momentum. Units move, the velocity graph ticks up, and it feels like the launch is working. Then the coupon ends, the price snaps back, and sales fall off a cliff. The buyers who showed up were never buying your product. They were buying the discount, and now they are gone, along with the margin you spent to get them.
This is the trap most brand owners fall into during a launch. They treat price and promotion as one lever instead of two, and they pull it in the wrong order. Coupons and deals are genuinely useful tools for a launch. They can seed the review base, feed the velocity Amazon's algorithm rewards, and buy you visibility you have not earned organically yet. But the sequence matters more than the size of the discount. Get it backwards and you build a customer base that only responds to markdowns.
Why sequence beats size
Amazon's ranking algorithm cares about sales velocity relative to competitors in your category, not about how deep your discount is. A 15 percent coupon that converts a well-targeted click is worth more to your rank than a 40 percent coupon that converts a bargain hunter who was never going to buy again. The mistake is discounting before you have proven the listing can convert on its own merits.
If your images and A+ content are still unfinished, or if your title and backend keywords are not indexing for the terms you actually want to rank on, a coupon does not fix that. It masks it. You get a burst of orders from price-sensitive shoppers, the conversion rate looks fine for the wrong reasons, and once the discount disappears, your true conversion rate against full-price traffic is what determines whether you keep the rank you bought.
The right order is: fix the page, confirm it converts, then use price and promotion to accelerate what is already working.
Week one: prove the listing before you discount it
In the first several days, resist the urge to launch with a coupon live. Instead, run your starting PPC campaigns at a price close to where you intend to sell long term, even if that means slower initial velocity. Watch the actual conversion rate on that traffic. This is the number that tells you whether the listing itself is doing its job.
If conversion is weak, the fix is the page, not the price. Check for the common listing mistakes that quietly suppress conversion: a hero image that does not stop the scroll, bullet points that describe features instead of answering objections, or a price that looks wrong relative to the value shown. Fix those first.
Week two: introduce a limited, purposeful discount
Once you can see the listing converting on undiscounted traffic, a coupon or a Prime-exclusive discount earns its place. Use it with a specific purpose: to close the gap between "interested" and "bought" for shoppers who are already on the fence, not to manufacture demand that is not there.
Keep the discount modest and time-bound. A coupon that runs indefinitely stops being a launch tool and starts being your price. Set an end date, and treat the discount as a lever you pull to convert marginal traffic, not a permanent floor. If you find you need the coupon live constantly to hold your unit velocity, that is a signal the underlying price or the underlying listing is misaligned with what the market will pay, and the fix is elsewhere.
A coupon should close a gap the listing cannot close on its own. If the coupon is doing the listing's job, fix the listing.
Reviews change what pricing has to do
Early reviews shift the entire calculus. A listing with zero reviews needs more help from price and promotion because shoppers have nothing else to trust. A listing with a growing base of genuine reviews needs less, because the reviews themselves are doing the conversion work that the discount used to do.
This is why it pays to run your review velocity plan in parallel with your pricing sequence rather than after it. As reviews accumulate, you should be able to taper the discount without losing velocity, because trust signals have replaced price as the reason to buy. If you find that velocity only holds when the coupon is live, even after you have a healthy review count, that is the bargain hunter problem showing up. Your buyer base has learned that the product goes on sale, and they are waiting.
Deals and event pricing are a different tool
A scheduled deal, whether a Lightning Deal or a seasonal promotion, is not the same tool as an evergreen coupon, and it should not run on the same logic. A deal is a short, high-intensity push meant to move inventory and spike rank around a specific date. It works because it is scarce. Run deals too often on a new product and you train the same bargain-hunting behavior, just on a calendar instead of a coupon badge.
If your launch timeline happens to fall near Prime Day or another major event, treat that overlap deliberately rather than accidentally. The pre-event checklist for locking inventory and pricing applies just as much to a new launch riding an event as it does to an established product.
Watch the exit, not just the entry
The clearest sign that promotion sequencing went wrong is not what happens while the discount is live. It is what happens the week after it ends. If units drop back to near zero, you did not build a customer base, you rented one. If units settle at a level above where you started, meaningfully above your pre-promotion baseline, the discount did its job: it introduced shoppers to a product that could stand on its own once they tried it.
Track this exit metric explicitly for every launch. It tells you whether your next launch should lean more on discount or less, and it is a far more honest signal than the vanity metric of units sold during the promotion window.
What to do this week
If you have a launch in progress or coming up in the next month, map out the sequence before you touch price. Confirm the listing converts on close-to-full-price traffic first. Set a firm end date on any coupon before it goes live. And build in a check, seven days after the discount ends, to see whether velocity held. That single check will tell you more about whether your launch worked than any number you saw while the discount was running.