Find the pillow your neck actually needs

Nothing. The app is free to use, there is no paid tier or subscription, and you do not need an account to get a full result. The browser version offers optional one-time support amounts of $5, $10, $20 and $50, and contributing changes nothing at all about your access, your fit score, or which products appear at the top of your list — that isolation is stated in the product's own documentation.

No. The front and side photos are analyzed on your own device, and only the resulting measurements — things like shoulder breadth and neck-to-shoulder distance — go into the recommendation. The images are not sent to Pillow Pair, not attached to your saved results, and not kept on a server. In the Android app the whole assessment and vision path makes no network requests at all, and the app holds no camera permission, so photos have no route off the device even in principle. You can also skip the photos entirely.

Because commercial status is not an input to the score. It is recorded as ordinary data on each product record, and the ranking rules never read it. The rules look only at compressed-height match, your sleep positions against the manufacturer's own guidance, firmness, adjustability, your clinician's constraints, and your material and cooling preferences. A test in the suite marks one of two otherwise identical products as an active affiliate and asserts they score the same. As it happens, no product in the catalog has an active commercial relationship today.

If you sleep well and wake up without a stiff neck, you probably should not bother. This is for the case where you have replaced pillows repeatedly and cannot work out why none of them are right, or where something changed — a new mattress, a new sleeping position, a shoulder that started hurting — and the pillow that used to work stopped working. Getting a target height range takes a few minutes and tells you whether height is the problem at all.

The web version is live and you can use it right now in a browser, or install it to your phone from the browser. The Android app is in early access with an invited group while the public store listing is finished. What is genuinely early is the science: the formulas connecting your measurements to a pillow height are transparent, research-informed estimates that have not been validated against direct measurement, and the app says so itself on its methodology page rather than pretending otherwise.

As a guest, your profile and saved results stay on your device. The browser version does send de-identified counters with no identifier attached, and a check-in you submit as a guest is sent anonymously including any note you write — no name, no account, no photos, no profile. If you create an account, Download my data gives you a full copy of your account, profile, saved results and check-ins, and Delete my account is a two-step confirmation that permanently removes them, cascading through your saved results and follow-ups. Both controls are inside the app in English and Spanish. The one thing kept is the record proving you consented to and later opted out of text messages, if you ever used that. It is detached from you but retained, because messaging rules require that proof to survive.

Your result is not locked inside the app. Download it as a plain report, share it, email it to yourself, or save it to Google Drive, and it will still name every recommended pillow by brand, model, exact variant, size and published height, with a link to the manufacturer's own page. There is nothing to cancel and no subscription to lose, and the report stays readable with or without the app.

Carefully, and the app is built to say no when it should. Six answers stop it outright and tell you to get urgent care: trouble breathing or swallowing, new bowel or bladder changes, new weakness, new trouble with balance or walking, recent significant trauma, and rapidly worsening severe unexplained pain. New numbness or tingling, fever with neck or back pain, a recent injury, post-surgical status with no clinician guidance entered, or symptoms that are getting worse hold the recommendation until you have been evaluated. In both cases you get no product list at all. If your condition is stable and already evaluated you can include it as context, and anything your clinician told you about pillow height, contour or adjustability becomes a hard rule that removes products which break it.

Yes to the first. Guest mode gives you the complete flow — intake, safety screen, photos if you want them, the support profile and the ranked matches — with nothing to register for. On Android the questionnaire and the recommendation engine run entirely on the device with the catalog built in, so they work with no connection; signing in, cloud history, export, deletion and the manufacturer links are the parts that need a network.

Up to six matches, ranked, with the top three laid out side by side for comparison. Each one shows its estimated compressed height for every position you sleep in, the reasons it placed where it did, and any cautions — including a standing caution that its dimensions are manufacturer-published and have not been compression-tested by Pillow Pair. If your constraints eliminate everything, you get a breakdown of which constraint removed how many products instead of an empty screen.

We would rather you hear this from us than discover it later. As of 2026-09-08:

• The formulas that turn your measurements into a pillow height are informed by published ergonomics research but have not been validated against tape measurements, 3D scans or sleep studies. The app says exactly this on its own methodology page rather than burying it.
• Catalog heights come from what each manufacturer publishes, not from Pillow Pair measuring the pillows itself, and all 19 records are labeled that way — every product match carries that caution on its face. A physical measurement program is written up and the logging path exists, covering an unloaded baseline, loaded readings out to eight hours and rebound readings at both the center and the edge, but no product has completed it yet.
• The per-material compression factors that turn a published height into an expected compressed height are development assumptions, not measurements. A real measurement for a specific product would replace the assumption for that product, and none exists yet.
• The head-support estimate rests on a published regression from a study of 15 cadaver heads, whose own authors noted the measured masses were probably low. That is why the app shows a range with a confidence label instead of a single number, and widens the range further when you have not entered a head measurement.
• The catalog is a starter set of 19 variants across 7 brands. It is not the whole market, and a pillow that would suit you may simply not be in it yet.
• Only 3 of the 19 variants are adjustable and all three come from a single maker, and only 5 are contoured across two makers. A clinician instruction requiring an adjustable or a contoured pillow therefore narrows the list hard. The app names the constraint that emptied your results rather than showing a blank page.
• Guest mode keeps your profile and saved results on your device, but it is not total silence: the browser version sends de-identified operational counters, and a check-in you complete as a guest is sent anonymously, including any free-text note you type. Neither carries your name, account, photos or profile, and the Android assessment sends nothing at all.
• Prices are not stored, because they change constantly. You check the current price on the manufacturer's page.
• The Android app is in early access with an invited group and is not downloadable by the general public.
• The fully bilingual assessment and US-unit defaults landed in the Android update currently with the store for review; the build the invited group is running still has an English-only assessment flow. The browser version has been bilingual throughout.