CRO for Shopify+ Brands
Data-driven, no guesses

mirrormate

MirrorMate October Concepts

MirrorMate · Convertibles · October 2026

October roadmap

Built on the final read of the frame-builder test, 624 answers to the new measuring survey, 15,000 post-purchase survey answers, the builder step tracking from July, the results of the last five tests, and 53 new customer reviews.

72%of people who typed in their measurements in the builder test did not order. 564 measured, 156 bought, on the current flow.
23%of buyers say a builder step made them consider not ordering. The steps named are the space, clips and vanity questions, not the tape measure.
53buyers asked, unprompted, for confirmation that they measured correctly before ordering. The top request in the measuring survey.
1SlotFrame builderFollow-up to the September test

Keep the People Who Already Measured

What the data says

September's builder test lost. The loss is after the tape measure, and it is about worry and price, not typing.

V1
Removing the size-and-price step cost orders. Orders among openers down 9%. "Buy now, measure later" fell from 12% to 4.5%. People stop at the price, not the screen. Hiding it only moved the price later.
Frame builder test · 27 Aug to 15 Sep · about 1,550 openers per version
V2
The fraction slider got more people through and fewer bought. 42% finished width and height, against 35%. 19% of them ordered, against 27.7%. Easier typing is not what people ask for: 7 of 451 in the survey.
Frame builder test. Measuring survey · 451 answers
V3
The exit popup was the only change that moved orders up. Orders among openers went from 8.2% to 10.8%. A quarter of openers saw it. Email use comes from Klaviyo.
Frame builder test · 404 people saw the popup
V4
The clips pictures did not add the clarity people need. Orders did not move. The pictures show what a clip is. The worry is "will this work on my mirror", and the question still asks people to read, decide and type.
Frame builder test. Measuring survey · 23 buyers name the clips step
05
Three sources, one answer. The test: 72% of measurers left (564 measured, 156 bought). July's tracking: they leave at clips (15%), space around (13%), review (8%) and cart (18%), not at width and height. The survey: buyers name the same steps, vanity 23, clips 23, space 20, measuring 11. Top ask: confirmation that they measured correctly, 53 of 451.
Frame builder test. Builder step tracking · July · 4,479 measurers. Measuring survey · 24 Aug to 16 Sep · 460 answers
The mirrors are flush to the wall on both sides. When I wrote 0 for left, right and bottom and it questioned it, I second guessed myself.Measuring survey, September
TODAY, CURRENT FLOW, 20 DAYS 1,599 opened the measuring steps 564 measured 156 ordered 408 measured, then left. 72%. THE BUILDER, STEP BY STEP WHERE EACH IDEA LANDS Size andprice Width andheight Spacearound Clips Vanity Review andadd to cart 13% leave 15% leave 8% leave price shock not the problem 5 1 2 4 4 4 3 1 The exit popup on its own, on today's inputs 2 The popup asks "what's stopping you?" and offers a call-back 3 "We check every measurement before we cut", with phone, email and chat 4 Space, clips and vanity on one drawing of the mirror, answers attached 5 A short intro above the sizes: five quick steps, every measurement checked IDEA 4: ONE DRAWING wall top 1/8 left 0 right 0 bottom 0 0 means flush. Tap where your clips are. Bulky clips? Tick yes, we send new ones free. Frame size and price update as you go IDEA 3: REVIEW STEP We check every measurement before anything is cut. Not sure about a number? Ask before you order. Call us, we pick up Chat now Email us Annandale White Slim, 30 3/4 x 40 1/4 $184

The concept

Leave the measurement inputs alone. Work on the people who have already measured: the worry that they got it wrong, the three questions after the tape measure, and the price. Five ideas to pick from on the call. Every step gets tracked.

  • Retest the exit popup on its own. The only change that moved orders up, and it ran on top of the slider that lost. Same trigger, today's inputs.
  • Ask why they are leaving, and offer to call. One question on the way out, "What's stopping you?", four answers: not sure I measured right, price, want to see it first, something else. Leave a number and someone calls. Kevin's exit survey, plus a list of people to help.
  • Say the measurements get checked, and show how to reach a person. On the review step: "We check every measurement before anything is cut." Phone, email and chat right under it. The top ask in the survey, 53 of 451, answered where people hesitate.
  • Put the space, clips and vanity questions on one drawing of the mirror. Today they are three written questions with boxes to fill, and this is where measurers drop. Instead, one picture of a mirror with the wall, the light and the vanity around it: type the gap on each side straight onto the drawing, "0 means flush", tap where the clips are. Attach the answer to the question: "Bulky clips? Tick yes and we send replacement clips free." A small summary of mirror size, frame size and price updates as they go, so they can see they got it right.
  • Make the size-and-price step an intro, not a wall. It stays. Add a short line above the sizes: five quick steps, a few minutes, every measurement checked before cutting. The price reads as the start of something easy.
2SlotProduct page, second screenFollow-up to the top-screen test

This Frame, in Real Bathrooms

What the data says

The customer proof on every product page is the same six videos, whatever frame the visitor is looking at, and it sits well below where most phone visitors stop scrolling. Buyers ask for photos of this frame, on this page.

01
Seeing it in person is the biggest thing that nearly stops an order, and customer photos are the most-asked-for fix. 3,231 of 6,619 buyers nearly stopped because they wanted to see the frame first. 649 of 4,276 said more customer reviews and photos on the product page would have made deciding easier, the top request after the virtual view.
Post-purchase survey · January to August · 6,619 and 4,276 answers
02
The proof that exists is generic and buried. "See what people think" shows the same six videos on a white frame's page and a black frame's page. On a phone the page runs to about 20,000 pixels and the average visitor stops around a third of the way down. The top-screen test launching this week measures how many people reach that block today.
MirrorMate product page · captured August 2026. Microsoft Clarity · August
03
The change is what customers talk about, and they want to see it before they buy. "Transformed" a builder-grade mirror is the most common thing reviewers say, about 40 of 92 reviews in June and again in September's batch. 13 of the 53 new reviews carry customer photos. The question the team hears most is "how will it look in my bathroom".
Customer reviews · June and September batches · 92 and 53 reviews. MirrorMate team · 27 August call
After receiving several samples, we selected the Annandale White Slim.Customer review, August
TODAY, SECOND SCREEN ADD $2 sample Ask our assistant three questions Description How to measure How to install customer videos 12,000 pixels down, the same six on every frame no proof about this frame PROPOSED, SECOND SCREEN Annandale White Slim, in real bathrooms photos and clips from customers who bought this frame Before and after, same mirror, drag to compare BEFORE AFTER Need help deciding? ADD $2 sample proof about this frame, one swipe in

The concept

Put customer photos and videos of the frame on that page on the second screen, starting with the top sellers, and add a same-mirror before-and-after to the gallery. If it moves frame orders on the top sellers, roll it out to all 50 styles.

  • Per-frame customer photos and short videos in the "See what people think" spot, moved up to the second screen. MirrorMate's team curates the top ten; the site has plenty of Instagram content to draw on.
  • A before-and-after in the gallery, slot two or three: one bathroom, one camera position, the mirror bare and then framed, shown as a wipe. Uppercase generates one per finish family.
4Collection pageFollows the descriptions test

Compare Frames Side by Side, Visually

What the data says

People already narrow the grid by colour: 42% of visitors who saw the new colour blocks clicked one. The next job is choosing between the three or four frames left, and today that means opening each page and coming back. Matching cabinets, taps and counters is what 76 of 944 written answers about nearly stopping were about.

01
The August call settled the shape. Width is the only spec worth showing; a table of style, colour family and collection is filler. Make it visual: the chosen frames on the same mirror photo, with the vanity and hardware pairings underneath.
September concepts call · 27 August
02
Nobody knows whether the existing compare table is seen. Every product page ends with a fixed "Compare Similar Styles" table. The top-screen test launching this week measures how many people reach it, which decides how much a dynamic version is worth.
Intelligems · product page test · from this week
1. TICK ON THE GRID Compare 3 frames 2. SEE THEM ON THE SAME MIRROR Chelsea BlackAnnandale WhiteHighline Nickel 2 1/4" wide1 1/2" wide1" wide with brass tapswith a white vanitywith chrome from $129from $99from $119 Order all three as $2 samples widths and pairings come from the descriptions test

The concept

Tick two to four frames on the grid, see them on the same mirror at the same size, pairings underneath, and one button to order all of them as $2 samples. The pattern is the one Crutchfield uses: tick, then compare side by side. Runs after the descriptions test, which gives each frame the words the comparison needs.

Earliest startWhen the collection descriptions test has read, and once the compare-table baseline is in. Not before November.
5Homepage, top two screensFree to run now

The Top of the Homepage Sells the Company, Not the Frames

What the data says

The first two screens hold one product photo, a row of press logos and the how-it-works steps. The first frame anyone can shop sits three screens down, and that block draws four times the clicks of the banner button. Asked what gave them the confidence to order, buyers put brand trust last.

01
Trust already ranks last with buyers. Brand trust came last of the five reasons offered, at 5%. The press logos hold a whole screen near the top of the page.
Post-purchase survey · January to August · 4,917 answers
02
The frame picker block pulls hard from three screens down. Clicks inside it, colours and styles combined, total 5,876 against 1,463 on the banner button, while 45% of visitors leave within one screen of the top.
Microsoft Clarity · desktop homepage · 30 days · 13,626 views
03
August's banner result points at the frames, not the picture. The more before-and-after a version kept, the less it lost, and the only one that did no measurable harm was the one with a colour picker.
Intelligems · homepage test · concluded 20 August · 35,866 visitors, 771 orders
TODAY, SCREENS ONE AND TWO SHOP NOW IN THE PRESS How it works not one frame to shop yet the frames start three screens down PROPOSED, SCREENS ONE AND TWO tap to change it Shop the frames $99 $99 $120 by colour · by style · by room press logos move further down frames to shop on screen two cut the way people browse the frames come to the visitor

The concept

  • Shoppable frame tiles in the top two screens. Real frames with prices, above or below the banner, or head to head against the press logos in the same spot. Cut them by colour, by style, by room.
  • Rebuild the banner comparison as one mirror, wiped, and keep the colour picker.
  • Make the colour swatches look tappable. Outline them, name the colour, mark the selected one.
Earliest startNow. Nothing is running on the homepage.
6CartFree to run now

A Real Cart Page Instead of the Side Panel

What the data says

The cart is a narrow panel that slides in from the side, and there is no cart page at all. Two jobs do not fit in that width: entering the discount code people already hold, and confirming the measurements they just spent minutes taking. July's test showed the panel is not short of confidence. It is short of room.

01
The discount code is the most written-about thing that nearly stops an order. 127 of 944 written answers, ahead of price at 115. Kevin's read on the August call: people sign up for the welcome offer and buy the same day, so a code landing in spam stalls the order.
Post-purchase survey · January to August · 944 written answers
02
One in five buyers says measuring nearly stopped them, and the cart never shows the measurements. 1,403 of 6,619. The one box the cart does show is "order special instructions", which people use to repeat clips and lights the builder already captured, and which stops the order going straight to production.
Post-purchase survey · 6,619 answers. MirrorMate cart · captured August 2026
03
Adding reassurance to this panel has already been tried. Across 67,174 visitors neither version moved sales and one lost 13% of revenue per order.
Intelligems · cart test · concluded July
TODAY Your cart (1) Satin Nickel $140 no measurements order instructions CHECKOUT no box for a code a narrow side panel, and no cart page PROPOSED, A REAL CART PAGE Your cart Highline Satin Nickel Slim Cut for 48 by 36 inches change $140 WELCOME10 applied ✓ filled in for them Subtotal $140.00 Welcome discount −$14.00 Total $126.00 Questions before you order? Call us: 9 to 5:30 Eastern · 5 year guarantee the code and the measurements are both settled before checkout

The concept

  • Send the cart link to a real cart page. The panel keeps doing the "added to cart" job.
  • Apply the code people already have, from the popup, the email or the text, so there is nothing to hunt for.
  • Restate the measurements on the line, with a link back to change them.
  • Take the instructions box out, and put the phone number in its place.
Earliest startNow, once the codes that arrive and then do not apply are fixed. A cart test on top of broken codes reads as nothing.
7Product pageFollows concept 2

Product Page, Round Three: the Words and the Second Screen

What the data says

What is left of the August product-page list once the top screen and the customer proof have been tested. Each of these is real, and each would muddy the two tests ahead of it if it ran at the same time.

01
Offering the sample while someone is still choosing costs basket size. The collection test's winning version removed the sample button from the grid and basket size went from $182 to $206. The product page gives the sample equal billing with "Start framing".
Intelligems · collection test · 25 August to 9 September
02
Installation is the second-biggest confidence driver and the description does not lead with it. 724 of 4,917 buyers credit the install instructions. The description and the accordion under it open with the same two sentences about the finish.
Post-purchase survey · 4,917 answers. MirrorMate product page · captured August 2026

The concept

  • The description written to four angles: what it looks like in a room, whether it fits, what it matches, how it goes on. Also the copy AI shopping tools pull, so it should be tested rather than assumed.
  • One description, not two. The intro and the accordion say the same thing.
  • The sample block's weight, now that the collection test says the mid-choice sample offer pulls people off the frame.
  • The assistant's real estate, once we have its usage by page.
Earliest startAfter concept 2, on the same page. Not before November.