ItemFits · Q3 2026 edition
ItemFits Q3 2026 Public Report
ItemFits answers one question with geometry instead of guesswork. Our proprietary language-to-geometry engine turns a plain-English description into a measured scene and solves it: will this couch, mattress, dresser or TV get through that door, up those stairs, into that elevator or into the back of that car? Three years in, this is what that looks like.
Published
Publisher: ItemFits LLC · Reporting period: Jul – Sep 2026 · Production data as of , before the quarter ended
Quarter over quarter
Q3 against Q2
Q2 is April to June 2026. Q3 is July to September 2026, counted through the last week of September.
| Metric | Q2 2026 | Q3 2026 | Change |
|---|---|---|---|
| Clicks from Google (definition) | 10,761 | 29,855 | +177% |
| Google search impressions (definition) | 902,411 | 2,484,413 | +175% |
| Page views (definition) | 56,769 | 210,576 | +271% |
| Conversations (definition) | 5,085 | 19,151 | +277% |
| Fit decisions (definition) | 4,282 | 15,335 | +258% |
| New accounts (definition) | 378 | 995 | +163% |
Growth
Ten times the search traffic since March
Every figure below is counted from production records, March to September 2026. Internal test accounts are excluded. September is a partial month and is shown lighter.
Clicks from Google
10,957 in August, up from 1,051 in March
Values (clicks)
| Month | Clicks |
|---|---|
| March 2026 | 1,051 |
| April 2026 | 2,308 |
| May 2026 | 3,311 |
| June 2026 | 5,142 |
| July 2026 | 10,075 |
| August 2026 | 10,957 |
| September 2026 (partial) | 8,823 |
Conversations
6,767 in August, up from 848 in March
Values (conversations)
| Month | Conversations |
|---|---|
| March 2026 | 848 |
| April 2026 | 885 |
| May 2026 | 1,259 |
| June 2026 | 2,941 |
| July 2026 | 6,209 |
| August 2026 | 6,767 |
| September 2026 (partial) | 6,175 |
Fit decisions
5,420 in August, up from 345 in March
Values (fit decisions)
| Month | Fit decisions |
|---|---|
| March 2026 (partial) | 345 |
| April 2026 | 756 |
| May 2026 | 1,300 |
| June 2026 | 2,226 |
| July 2026 | 5,139 |
| August 2026 | 5,420 |
| September 2026 (partial) | 4,770 |
New accounts
372 in July, the best month so far
Values (new accounts)
| Month | New accounts |
|---|---|
| March 2026 | 43 |
| April 2026 | 78 |
| May 2026 | 96 |
| June 2026 | 204 |
| July 2026 | 372 |
| August 2026 | 365 |
| September 2026 (partial) | 258 |
Demand
People bring real, specific problems
The answer is usually no
More than half of the definite answers we have given were won’t fit or likely won’t fit. Each of those is a return, a failed delivery or a sofa stuck in a stairwell that someone found out about on their phone instead.
What we learned this quarter
Findings from 15,335 fit decisions
Every fit check is a small measurement of the real world. Here is what Q3’s added up to.
Cars dominate the question
9,557 of Q3’s fit decisions were about getting something into a vehicle. Doorways came next at 10%.
Jul – Sep 2026 · September partialSource: ItemFits Q3 2026 dataset · How this is countedCheck an item against a vehicle →No is more common than yes
Of 13,594 definite answers in Q3, 6,926 were won’t fit or likely won’t fit.
Jul – Sep 2026 · September partialSource: ItemFits Q3 2026 dataset · How this is countedWhat stops things fitting →A fit is often a close call
Of 6,155 fits with a measured clearance, 1,208 had less than 2 inches to spare. That is too close to eyeball with a tape measure.
Jul – Sep 2026 · September partialSource: ItemFits Q3 2026 dataset · How this is countedHow to measure accurately →Asking beats guessing
Times in Q3 the engine stopped to ask for a missing measurement instead of guessing one. That was 9% of fit checks.
Jul – Sep 2026 · September partialSource: ItemFits Q3 2026 dataset · How this is countedHow the checker decides →What people tried to fit
Sofas, mattresses and TVs, into Corollas and RAV4s
Most checked items
Fit checks by item family. Aug – Sep 2026 · September partialSource: ItemFits fit-check records · How this is counted
What kind of space
Share of 15,335 fit decisions. Jul – Sep 2026 · September partialSource: ItemFits fit-check records · How this is counted
Most picked vehicles
- Toyota Corolla
- Volkswagen Golf GTI
- Toyota RAV4
- Toyota Camry
- Tesla Model 3
- Subaru Outback
- Honda Civic
- Chevrolet Equinox
By body style: SUVs 43%, sedans 30%, hatchbacks 14%, minivans 7%, wagons and pickups 3% each. Jul – Sep 2026 · September partialSource: ItemFits vehicle-grid picks · How this is counted
Where the questions come from
All 50 states · 112 countries · 12 languages
Top countries after the US
- Canada
- United Kingdom
- Australia
- Germany
- Netherlands
- France
- New Zealand
- Norway
- Sweden
Ranked by questions asked. Jul – Sep 2026 · September partialSource: ItemFits request locations · How this is counted
Where it works
One engine, wherever the question comes up
The same language-to-geometry engine answers every surface, so a shopper gets the same verdict on a retailer’s product page as on itemfits.com.
itemfits.com
Describe any item and any space in plain words. Available in 12 languages, with vehicle, door, stair, hallway and elevator guides.
Open the fit checker →ItemFits: Products & Spaces
For furniture and appliance retailers. A fit button on product pages, a whole-cart check, a route check from front door to room, an AI assistant and fit analytics for the merchant.
View the Shopify app ↗ItemFits — Will It Fit? Any Item, Any Space
Reads a product’s dimensions straight from the retailer’s page and checks them against your door, stairs, elevator or car.
12,139 answers with a permanent link
Every answer can be sent to a partner, a mover or a delivery team, with the full calculation attached.
Mar – Sep 2026 · September partialBrowse shared results →Our findings and our math are public
Anyone can read our studies of real fit checks, such as what most often stops an item from fitting and which big items people check most. Every formula we use is published too, each with a diagram and a worked example.
The formulas →What makes it different
A proprietary language-to-geometry engine
You describe the problem the way you would to a friend: “will my 84-inch sectional make it up a narrow staircase with a turn at the top?” Our engine reads that sentence, identifies each object and space, attaches real dimensions from what you said or from our catalogs, builds a geometric model of the scene, and hands it to the right solver. Cubic feet tells you almost nothing about whether a 7-foot sofa gets through a door. The engine works out how the object actually has to move.
It turns the object, the way a mover would
Every orientation is tried: on its side, on end, tilted through a doorway, pivoted around a hallway corner, stood up in an elevator cab. 89 purpose-built solvers cover doors, stairs, turns, elevators, vehicles, rooms and containers.
See the rotation and tilt formulas →The math is deterministic
Language models read your words. The verdict comes from geometry code with no randomness, so the same inputs always give the same answer. Nobody edits a verdict by hand.
The opening and the space are separate checks
A car can have room for a dresser that will not get past its liftgate. We check the way in and the space inside independently, which is exactly the point Chevy Truck Expert picked up on.
Vehicle cargo dimensions →It says when it’s guessing
If a size comes from a catalog standard or an estimate rather than your tape measure, the answer labels it “Standard size” or “Estimated” so you know which numbers to double-check.
It shows its working
Each answer comes with the full calculation, the tightest point, the best approach and a 2D or 3D visual, so you can check the reasoning before anything gets lifted.
Built on published standards
Standard doors, halls, stairs and elevators come from the building codes, not rules of thumb. A ±¼ in tape-measure tolerance is built into every check.
Our methodology →Authority
Cited by the people who know the vehicles
Collaboration finalized Sep 2026 · credit live on itemfits.com since Sep 2026
“The worksheet was developed in editorial collaboration with ItemFits, whose Suburban calculator treats the cargo opening and the interior space as separate geometric checks.”Chevy Truck Expert, Suburban Cargo-Fit Measurement Record
Chevy Truck Expert is an independent enthusiast publication for Chevrolet truck and SUV owners. We worked with them on a free, printable worksheet for 2021-and-later Suburban owners that records the liftgate opening, the seat configuration, the item and any loading-path limits in one place. Their page credits ItemFits by name, and our Suburban page links back to their worksheet. No money changed hands in either direction.
The collaboration came out of a rebuild we had just finished. We replaced hand-typed spec-sheet numbers on our vehicle pages with the same measured catalog the calculator uses, and the page now says plainly which layouts we have measured and which we have not.
Chevrolet Suburban cargo calculator → · The worksheet on Chevy Truck Expert ↗
Partners
Privacy and trust
Your data is yours to delete
ItemFits follows the GDPR right to erasure (Article 17). Deleting an account takes one click and nobody has to email anyone.
Immediate
Deletion from Settings → Account removes your profile, chat history, saved projects, room scans, API keys and email preferences on the spot, and emails you a receipt listing what was erased.
De-identified
Records we must keep for payments or abuse prevention have your account, email, IP address and device fingerprint stripped, so nothing links back to you.
Never sold
We do not sell or rent personal information or share it with third parties for their marketing.
Encrypted
Data is encrypted in transit and at rest. Card payments go through Stripe, and we never store card numbers.
30 days
The longest we take to answer a privacy request.
No fake reviews
We don’t publish invented testimonials or star ratings. When we have enough verified ratings, we will show them.
Read the full privacy policy.
How it’s built
Tested like the answer matters
ItemFits was founded in 2023 and formed as ItemFits LLC in September 2026. The current platform was rebuilt from the ground up. What we care about most is that every answer can be checked.
verdicts written by a person. Every answer is computed from the measurements
hand-verified known-answer cases, checked automatically on every change, across 11 validation programs
automated tests covering the geometry, the language parsing and the answers
re-runs anonymized real customer questions to catch any change in an answer before release
Also: 9,398 commits since Dec 2025, 602 database migrations, 354 API routes and 44 automated checks on every change. Counted Sep 2026.
These describe how the checker is tested, not how often it is right in the field. See the solver validation study and the formulas.
Coming in future editions
What we are measuring next
- Speed. Median and slowest-case time from question to answer.
- Retail outcomes. What shoppers do after a fit check on a retailer’s site: add to cart, purchase, return.
- Sales redirected. How often a won’t-fit answer leads to a different product that does fit.
- Where things get stuck. Which point on a route blocks an item most often: the door, the stair turn, the hallway corner or the vehicle opening.
- Measured vs. estimated. How many answers used only measured or manufacturer dimensions.
Methodology
How these numbers are counted
Every figure carries its own date range. Fit decisions are rows in our fit-check results record from Mar 2026, when that record began; earlier checks are not counted. Internal test accounts are excluded from fit decisions, conversations, questions and new accounts. A conversation is a chat session with at least one question in it, across the web app, extension and Shopify widget. New accounts are registered (non-anonymous) sign-ups.
Page views are our own tracked page-visit events and exclude visitors who block scripts; the monthly average includes the partial month of September. Search figures come from Google Search Console. Countries and US states are counted from requests to the fit checker, recorded since Jul 2026. Vehicle rankings come from picks on the vehicle grid. Item families group the item names people typed; generic names are left out. The item count uses a field we began recording in Aug 2026. Code figures are counted from the git repository.
This edition was drawn from production data on the day it was published, before the end of September, so every window that includes September is partial.
Every term is defined on How ItemFits data is counted. For the living, regularly updated figures, see the ItemFits Fit Check Dataset. This report is a historical snapshot and does not change when new data arrives.
Data notes
- The September fit-decision bar (4,770) came from a data pull a few hours earlier than the quarter total (15,335), so the July–September bars add up to 15,329, six fewer than the total.
- The fit-check record began on March 9, 2026, so the March fit-decision bar (345) covers only part of the month. It is drawn lighter here; the original edition did not mark it.
- The space-type shares are rounded to whole percentages and sum to 100%. Recomputing them from the underlying records puts “Other” nearer 2% than 3%. The grouping of checker routes into space types was not recorded when the edition was prepared, so the published 3% cannot be reproduced exactly.
- The last-4-weeks search figures (16,510 searches, 1,729 pages, average position 6.9) were read from Search Console at publication. Google revises recent Search Console data, and the exact end day of that window was not recorded, so they cannot be reproduced to the unit today.
Corrections
No corrections since publication.
Quoting a single figure? Give the window printed next to it, and link this page. For the current value of a figure, use the Fit Check Dataset, which is updated as new data arrives and can be downloaded as CSV or JSON.