I have optimized 14,000+ hero images across 50,000+ listings, and home health is the category where I most often see good products losing to worse ones for a reason nobody diagnoses: the images were built like consumer electronics when the buyer is in a completely different mental state.
Amazon medical device product images — knee braces, blood pressure monitors, TENS units, thermometers, mobility aids — sell to someone who is hurting, scared, or shopping for a parent who is. That buyer does not browse. They evaluate. And most listings in this category answer none of the questions that evaluation is actually running.
This playbook covers the hero stack for home health and medical devices: what the buyer is silently asking, the five-layer hero framework adapted for this category, the compliance minefield that suppresses listings, the FSA/HSA badge most sellers never claim, and the sizing problem that quietly feeds the returns machine.
The Buyer Is in Pain, or Buying for Someone Who Is
Every category playbook starts with the one-second questions. In home health, you first have to understand who is asking them, because this category has three distinct buyers and most stacks are built for none of them:
The self-buyer in pain. Their knee has hurt for three weeks. They typed "knee brace for meniscus tear" at 11pm. They are not shopping for features — they're shopping for relief, and they are suspicious of anything that looks flimsy.
The caregiver. Adult child buying a blood pressure monitor or a rollator for a 74-year-old parent. Buyer is not the user — the same split I wrote about in the pet supplies playbook, but with higher stakes. They're asking: will Mom actually be able to use this? Can she read the display? Will it feel safe?
The prescribed shopper. "Doctor told me to get a home BP monitor / compression sleeve / cervical pillow." They arrive with a vague spec and no brand knowledge. They want the listing to look like something a clinician would nod at.
All three buyers share one trait: anxiety raises the trust threshold. In drinkware, a mediocre hero loses you a click. In home health, an untrustworthy hero disqualifies you. The one-second questions in this category are: Will this actually help my specific problem? Is it safe and legitimate? Will it fit me / can they use it?
Your hero stack either answers those or the shopper scrolls to a listing that does.
The Five-Layer Hero Stack for Home Health
Same five-layer discipline I use in every category playbook, tuned for the anxiety purchase.
Layer 1: Instant identification — on the body, not in the bag
The single most common hero failure in this category is the flat-lay: a knee brace photographed like a folded t-shirt, a TENS unit as a box of parts. A brace flat on white is an ambiguous fabric object. A brace on a knee is a product with a job.
For wearable supports — braces, sleeves, posture correctors — the category norm on page one is on-body or on-limb presentation, and it wins for a reason: it answers identification and placement in one glance. Check your top ten competitors before you commit either way, but if your hero is a flat-lay and the page-one incumbents are on-body, you've found your first test. Standard main-image rules still apply — pure white background, the product occupying 85% of the frame, nothing in the shot the buyer doesn't receive.
For devices — monitors, TENS units, nebulizers — the hero is the device itself, three-quarter angle, with its interface visible. Which brings us to the layer this category gets uniquely wrong.
Layer 2: The living display
If your product has a screen, the screen is your capacity number. In the small kitchen appliance playbook I called the dead display the cardinal sin — in home health it's worse, because the reading is the product. A blood pressure monitor exists to produce that number.
Show the display ON, with a realistic, healthy-adjacent reading — large, high-contrast, legible at 280 pixels. A BP monitor showing 120/80 in fat digits does three jobs at once: proves the device works, demonstrates the display is readable (the caregiver's #1 worry for an aging parent), and hands the machine-reading layer clean OCR text. A dead gray screen does none of them and quietly reads as "stock photo of a product that may not exist."
Layer 3: The fit signal
Sizing is the return engine of this category — more below — but the hero itself should telegraph fit. For a brace: which joint, which side of the body if it matters, and visible adjustability (straps, dials, open patella). For cuff-based monitors: cuff on an arm, because cuff fit is the top functional complaint in reviews. For mobility aids: a human-scale cue, since a cane or rollator with no reference reads as either toy-sized or industrial.
Layer 4: Trust cues without claims
This is the tightrope. You want legitimacy signals in the hero zone; you cannot put badges, seals, or claim text on the main image. The move: let build quality do the talking in the hero (real materials, visible stitching, metal where there's metal), then spend slot 2 or 3 on a dedicated trust panel — FDA-registered/cleared status where genuinely applicable, testing standards, warranty terms handled carefully. The compliance section below covers what will get you suppressed.
Layer 5: One differentiator
One. In this category the differentiator hierarchy is roughly: clinically-relevant design feature (hinged support, clinically validated accuracy, dual-user memory) > comfort/wearability > accessory count. The listing that leads with "comes with a free carry bag" while the competitor leads with "hinged lateral support" loses the meniscus-tear buyer every time.
The Compliance Minefield: Claims That Kill Listings
No category punishes sloppy image copy harder. The A+ rejection playbook covered banned language broadly — here's the medical-specific tier, because these mistakes don't just bounce an A+ module, they suppress listings.
"FDA approved" is the phrase that ends listings. Most home health products are FDA cleared (510(k)) or the facility is FDA registered — legally distinct from approved, and Amazon's automated sweeps know the difference even when sellers don't. Baking "FDA APPROVED" into an infographic is a suppression waiting for its enforcement pass. Use the precise term your regulatory status supports, or don't reference FDA at all. And never use the FDA logo — the agency itself prohibits it, full stop.
Disease treatment claims are the second rail. "Cures plantar fasciitis." "Treats arthritis." "Eliminates sciatica." The defensible register is support and management language — "targeted compression for knee support" — not medical outcomes. The dividing line: describe what the product physically does, not what condition it fixes.
Before/after pain framing — the pain-scale-10-to-2 graphic — is outcome-claim territory dressed as design. It also reads as infomercial, which is precisely the wrong register for the anxious buyer.
Borrowed authority — white-coat stock models, stethoscope props, "#1 Doctor Recommended" without a citable survey — is both a policy risk and a trust anti-signal. Shoppers in this category have seen a thousand fake-clinic renders. Specificity beats theater: "tested to ISO 81060-2" out-converts a stock doctor every time it's true.
The uncomfortable part: sellers respond to these constraints by stripping images down to nothing, which is its own failure. Compliance doesn't require blandness. Materials, mechanisms, measurements, fit, and honest capability statements are all legal — and they're the content this buyer actually wants.
The FSA/HSA Badge: The Free Conversion Lever Most Sellers Never Claim
Here's the part of this playbook that pays for the read.
Amazon runs an FSA/HSA program: eligible products show an "FSA or HSA eligible" badge in search results and on the detail page, and shoppers can filter search results to eligible items only. Eligibility flows from the SIG-IS list (the industry group that maintains eligible-product data under IRS Section 213(d)) — Amazon matches UPCs from that list to ASINs automatically.
Why creative people should care about a catalog mechanism: the badge changes the price math in the shopper's head. An FSA/HSA purchase is pre-tax money that expires — for that shopper, your $89 monitor competes at an effective ~30% discount against non-eligible alternatives, and the filter removes those alternatives from the results page entirely. It's the rare badge that both boosts CTR and shrinks your competitive set.
The catch is that automatic matching misses products constantly — wrong UPC history, catalog quirks. If your product type is plainly eligible (most braces, monitors, TENS units, first-aid) and you're not badged: confirm your product category on the SIG-IS list, then open a Seller Support case citing it, with your UPC. Sellers get badges added this way routinely. And once you have it, merchandise it — a "save with your FSA/HSA dollars" callout on an A+ module or later-slot image, especially in Q4 when use-it-or-lose-it deadlines drive a genuine December surge in FSA spending. The badge itself does search; your stack should close the loop on the page.
Sizing: Where This Category's Returns Are Manufactured
Fit-dependent products — braces, sleeves, compression wear, cuffs — live and die on a boring image: the sizing chart. Most listings bury it in A+ or skip it entirely. Then the returns arrive: too small, too tight, doesn't fit my arm.
The compounding problem I've covered before applies with full force here: expectation-gap returns accumulate toward the frequently-returned-item badge, and that badge takes a 20-50% CVR hit on everything the listing does afterward. In a category where fit is genuinely hard, the sizing image isn't housekeeping — it's badge insurance.
What the sizing slot (2 or 3 — early, not buried) needs:
- A measure-how graphic, not just a table. Show where on the body the tape measure goes. Half of brace returns trace to people measuring the wrong spot on the joint.
- Fat, thumbnail-legible ranges. S/M/L means nothing; "fits 13"–17" above-knee circumference" means everything.
- Honest range edges. If your XL genuinely fits a 22" thigh, say so; if it doesn't, the flattering copy costs you a return plus a badge contribution.
- Cuff ranges for monitors. "Fits 8.7"–16.5" upper arm" printed large. Large-cuff shoppers search for exactly this, and Rufus fields the question constantly.
One more honesty point: adjustability is a feature, but "one size fits most" is a claim that reviews will audit. If your review page says it doesn't fit larger athletes, your images should stop implying it does. I'd rather lose the click than eat the return and the badge math that follows.
Demographic Honesty and the Machine Layer
Two quieter levers.
Show the hands that will actually use it. The lifestyle-demographic mismatch problem is at its most extreme in this category: rollators shot with 30-year-old models, arthritis gloves on hand-model hands. The 68-year-old buyer — or their adult child — runs a half-second "is this for someone like me/Mom" check, and a listing full of young athletic bodies fails it. Match the model to the buyer's reality. For dual-market products (a knee brace selling to both runners and retirees), split the lifestyle slots and cover both instead of averaging into neither.
Feed the machine specific text. Home health queries are among the most specific on Amazon — "blood pressure monitor large cuff upper arm," "knee brace for meniscus tear left knee," "TENS unit for lower back." Rufus and the AI layer resolve those against your structured attributes and the OCR-readable text baked into your images. High-contrast callouts with the exact spec vocabulary — cuff range, joint, intensity levels, weight capacity — keep you in the consideration set for the long-tail queries where this category's margins live. Thin decorative type fails the thumbnail and the machine at the same time.
Subcategory Rules
Braces and supports. On-body hero, joint obvious, adjustability visible. Slot 2 = measure-how sizing. Differentiate on support mechanism (hinged vs. sleeve vs. strap), not accessory count.
Blood pressure monitors. Display ON with a fat, legible reading. Cuff range printed large. Accuracy/validation standard on the trust panel if you genuinely hold one. Caregiver features — memory for two users, backlit display, app sync shown with a real screen — earn slots here more than anywhere.
Thermometers. Speed and mode are the differentiators (forehead/ear/oral; 1-second read). Show the reading, show the distance for no-touch models. Parents buy these at 2am — the stack should work for a panicked phone glance.
TENS units. The intimidation-factor category. Hero: device with pads placed on a real back, screen on. Slot 2: what's in the box, because pad count and lead configuration drive the "do I need to buy more parts" hesitation. Keep intensity claims mechanical, not medical.
Mobility aids (canes, rollators, shower chairs). Weight capacity in big type — it's the #1 filter question and a safety issue. Show fold/transport state, show real human scale, and respect the dignity point: the buyer is often easing a hard family conversation, and imagery that makes the product feel stigmatizing loses to imagery that makes it feel like independence.
Pulse oximeters and heating pads. Oximeters: display on, finger in device, one honest accuracy line. Heating pads: size relative to a body ("covers the full lower back"), controller legible, auto-shutoff as the trust feature.
Nine Anti-Patterns I See Every Week
- The flat-lay brace — folded fabric on white, answering nothing.
- The dead screen — a monitor that shows no reading is a monitor that shows no proof.
- "FDA APPROVED" baked into an infographic — wrong term, suppression bait.
- The pain-scale before/after graphic — outcome claim in costume.
- The stock-photo doctor — borrowed authority that reads as fake.
- The icon wallpaper — six vague medical icons ("PREMIUM QUALITY," a heart, a shield) instead of one real spec.
- Sizing buried in A+ — the return machine, fully assembled.
- The 25-year-old rollator model — demographic mismatch at its most absurd.
- Generic medical blue — the entire category is blue-and-white; the listing that owns a disciplined non-blue system stands out on the grid without gimmicks.
The Six-Step Audit
- Pull your hero at 280px next to the top five competitors. Can a stranger tell what it is, where it goes on the body, and that it's built well — in one second?
- Screens on? If the product has a display and your images show it dead, fix that this week. Cheapest CVR test in the category.
- Run the claims sweep. Every image and A+ module: any "approved," any disease-cure language, any outcome numbers you can't substantiate. Fix before an enforcement pass finds it.
- Check your badge. Search your product logged-out with the FSA/HSA filter on. If eligible-but-absent, that's a Seller Support case this week — near-zero effort, real CTR return.
- Read your last 60 days of return reasons. Fit/size complaints mean the sizing slot moves earlier and gets a measure-how graphic. The reasons are literally telling you which image to build.
- Match the demographics. Compare the humans in your lifestyle slots against who actually buys. If they've never met, re-cast the primary lifestyle image first.
FAQ
Can my main image show the brace on a person? On-limb presentation is the working page-one norm for wearable supports, within standard main-image rules (white background, product dominant, nothing extra in frame). Check the top ten results for your exact query — if incumbents are on-body and you're flat-lay, test the switch; it's usually one of the biggest CTR moves available in this category.
Is it worth pursuing the FSA/HSA badge for a small brand? Yes, and it's cheap: verify eligibility against the SIG-IS list, open a Seller Support case with your UPC. The badge shows in search results, survives into the filtered results page, and effectively discounts your price by the buyer's tax rate. Very few levers in this category cost an hour and change the search grid.
Can I say my device is "FDA approved"? Almost certainly not — most home health devices are FDA cleared or the facility is FDA registered, and the wrong term is a suppression risk. Use the exact term your regulatory status supports, keep documentation, and never use the FDA logo.
What's the single highest-leverage fix if I only do one thing? Fit-dependent product: move a measure-how sizing graphic to slot 2. Display product: turn the screen on with a big legible reading. Both are afternoon fixes that attack this category's two biggest leaks — sizing returns and dead-screen distrust.
How do I stand out when every listing looks clinical? Not by being louder — by being specific. Real materials at macro distance, a real reading on a real screen, real sizing honesty, and one owned non-blue accent in a disciplined system. The anxious buyer isn't looking for excitement. They're looking for the one listing that looks like it was made by people who understand their problem.
Home health is the clearest case I know for merchandising over design. A beautiful listing that ignores the buyer's fear loses to a plainer one that answers it. If you're running braces, monitors, or mobility products and your CVR doesn't match your product quality, start with the audit above — and if you want the deeper pattern library, the supplements playbook, the frequently-returned badge breakdown, and the personal care playbook all border this territory.