Ask ten small business owners how they collect Google reviews and you will hear two answers: "we print a QR code" and "we text a link." Both are defensible, but they convert at completely different rates depending on where the customer is standing, what they just did, and how much friction sits between them and the review screen.
Below: conversion benchmarks by touchpoint, the four print mistakes that kill roughly 60% of printed codes, sizing and error-correction specs you can hand straight to a printer, five collateral formats with realistic performance ranges, and a complete A/B blueprint you can run in six to eight weeks.
The Head-to-Head Conversion Battle
The number that matters is reviews posted per 1,000 customer exposures, not raw scans. Scans are a leading indicator, and chasing them produces the classic trap: a countertop code that gets 400 scans and 11 reviews loses to a texted link that gets 60 opens and 7 reviews, even though the QR's engagement numbers looked far better.
Ranges below are directional aggregates of field campaigns run by print shops, agencies, and review-management vendors. Treat them as priors to test against, not guarantees for your business.
| Touchpoint | Interaction rate | Interaction → posted review | Reviews per 1,000 exposures |
|---|---|---|---|
| Countertop tent card (in-store) | 5–15% of transactions | 15–35% | 8–53 |
| Cash register receipt footer | 0.5–2% of receipts | 15–30% | 1–6 |
| Vehicle decal or service magnet | 2–8% of visits | 15–30% | 3–24 |
| Post-service leave-behind card | 10–25% of jobs | 20–40% | 20–100 |
| Packaging insert (e-commerce/retail) | 1–3% of orders | 10–25% | 1–8 |
| SMS shortlink request | 5–20% of messages clicked | 3–12% | 2–24 |
| Email shortlink request | 2–10% clicked | 1–6% | 0.2–6 |
In-person and counter touchpoints
A physical code wins here, and it is not close. The customer is standing in front of you, the transaction is done, and a scan takes three seconds with no typing. A texted shortlink in the same scenario competes with the customer's pocket and usually loses. If you can only run one channel in-store, run the print.
Field service and job-site touchpoints
Field service is the strongest case for physical, because the technician controls the moment. The leave-behind card handed over at the driveway and the van decal parked in front of the customer produce the highest rates in the table, because the request lands on visible satisfaction instead of a schedule you chose. Unprompted SMS from an unknown number is increasingly filtered.
Digital follow-up touchpoints
Shortlinks win the follow-up. An email or text arrives days later with no friction, no print cost, and instant reach across every customer on your list. This is also where dynamic routing pays: you can route to Google, Yelp, or Facebook based on the customer's prior rating without changing any artwork.
A QR code is a location-based ask. A shortlink is a relationship-based ask. They succeed for opposite reasons, which is why running both beats optimizing one.
Scan Failure Points: What Ruins 60% of Printed QR Codes
Print audits consistently land near 60% of printed codes failing on first attempt, and almost none of those failures come from the code itself. They come from four decisions made in the design tool before the file reaches the printer.
| Failure point | What goes wrong | Fix |
|---|---|---|
| Too small | Modules fall below ~0.4–0.5 mm and blur together in print; nothing decodes | 1 inch minimum for handheld, 3+ inches for signage. Scale the artwork, never the module count |
| Inverted or low-contrast colors | White modules on a dark background fail on a meaningful share of older Android camera apps; pale-on-pale never scans | Dark modules on a light background, contrast ratio of at least 3:1 and ideally 7:1 |
| Wrong error-correction level | A logo punched into the center of a Level L/M code corrupts modules and the code silently stops decoding | Level Q or H for any logo centerpiece or overlay over 10% of the code area |
| Cropped quiet zone or poor output | The required four-module white margin is trimmed, or 300 dpi heat, glare, and dot gain swallow the smallest modules | Four-module quiet zone on all sides, 300 dpi minimum, matte stock, rich black rather than 100% K on gloss paper |
Environmental damage is the second tier: laminated cards curl, window decals catch a glare band at midday, and bent paper or a bumpy vehicle panel distorts the grid. Matte vinyl and flat placement solve more of these than error correction does.
Technical Best Practices for Printable QR Codes
Minimum physical size by viewing distance
A common rule of thumb is one inch of code for every ten feet of scanning distance, but that assumes a large, flat, high-contrast code and a modern phone. Treat it as an optimistic floor, design to the far end of the range, and test with the oldest phone in your audience.
| Placement | Scan distance | Minimum code size |
|---|---|---|
| Handheld / countertop | 1–2 ft (arm's length) | 1 inch (25 mm) |
| Reception desk, checkout counter | 4–6 ft | 1.5–2 inches |
| Door, elevator, hallway | 8–12 ft | 3 inches (75 mm) |
| Vehicle window or side panel | 6–10 ft, moving | 3–4 inches, matte vinyl |
| Storefront or parking lot | 20–40 ft | 5–6 inches, or a separate eye-level sign |
Short URLs matter here: every extra character adds modules, which shrinks each module at a fixed print size. A 20-character URL scans at half the size of a 60-character one. Use a branded short domain and keep the destination under 30 characters.
Error correction levels
| Level | Recovery | Use for |
|---|---|---|
| L | 7% | Plain URL, no artwork, digital display only |
| M | 15% | Default for print; logo overlay under ~10% of the code |
| Q | 25% | Branded centerpiece with a logo covering 10–20% |
| H | 30% | Heavy branding, bottle wraps, outdoor or scuffed surfaces |
Raising error correction adds modules and forces a smaller code at the same print size, so level and physical dimensions must be decided together. Do not shrink the artwork to preserve dimensions and expect it to scan.
Dynamic versus static QR codes
A static code freezes its destination into the printed pixels. If you later change domains, abandon a shortlink provider, move platforms, or want a different landing page per location, every piece in the field becomes a dead end, and print cannot be recalled. A countertop tent card that stays up for a year is a permanent liability you cannot fix.
Dynamic codes store the destination server-side and redirect on scan. You can swap the destination without reprinting, send different traffic to different locations from one file, add scan-level analytics without UTM bloat, and retire a bad offer on the spot.
Dynamic QR done right: point your own branded domain at the redirect service (for example yourdomain.com/r/review), never at a third-party shortener. If you later migrate the service, you can repoint your own domain and every printed piece keeps working. If you printed a static code pointing at a third-party shortener, switching providers means reprinting everything.
The trade-off: a dynamic code adds a redirect hop of roughly 100–300 ms, and it only works while the service behind it is alive. Owning the domain is what makes that risk manageable.
5 High-Converting Physical Collateral Formats
1. Countertop acrylic tent cards
A 3.5 × 5 inch or A6 acrylic tent sits at eye level on the counter or beside the payment terminal, so the code is read at 1.5–2 feet. It produces the highest volume of any in-store format because it is in the customer's line of sight at the moment of payment. Pair it with a staff prompt instead of expecting the code to sell itself, rotate the design monthly so it doesn't become wallpaper, and budget $0.15–0.60 per unit in quantity.
2. Cash register receipt footers
Receipt footers scale with volume but convert poorly: the paper is small, thermal-printed, and discarded within seconds. Use a code of at least 0.6–0.75 inch, keep it 0.25 inch clear of the paper edge and away from thermal smudging, and print a short URL beneath it as a fallback. The strongest version pairs the code with a stated value ("$5 off your next visit") rather than a bare request.
3. Vehicle window decals and service magnets
These work passively, at traffic speed, and only if placed correctly: put the 3–4 inch code at eye level on the rear quarter panel or the driver's-side door magnet, never below the window line where it disappears behind the sill, and never on glass that catches a hard glare. Use matte vinyl, because gloss reflects sun into the camera. Expect to reapply every 12–18 months as UV degrades the vinyl.
4. Post-service inspection leave-behind brochures
This is the highest-converting format in the table, and the reason is simple: the customer is physically holding the card at the moment of visible satisfaction. Include the technician's name, one line describing the work performed, and a single instruction on the back. Hand it over at the door rather than adding it to a stack of paperwork, and fold it so the code lands on the outside panel.
5. Packaging insert cards for e-commerce and retail
Insert cards convert at 1–3% of orders, which looks weak until you multiply by order volume. A 2 × 3.5 inch card on 300 grams of cardstock fits a 1 inch code, but placement decides everything: a card buried under the invoice and packing slip is a wasted impression. Ship monthly or quarterly in a plain white envelope, and use a dynamic code so you can rotate the offer without reprinting.
Designing the Call-to-Action Frame
A bare QR code is a URL-shaped object most customers do not consciously recognize. Field tests put the gap at roughly 70% more scans for a framed, captioned code than for the same code dropped onto a design with no instruction. The frame does three jobs: it tells the eye what the object is, tells the reader what they get, and removes the decision of whether to bother.
- Headline: two to four words, a verb plus an object. "Scan to Review" or "Rate Us in 10 Seconds" beats "QR Code".
- Value line: what the reader gets for acting. A stated benefit, a discount, or a thank-you gesture all outperform a bare request.
- Instruction: tell them exactly what to do, including that the phone camera works. One line is enough.
- Contrast band: a solid light band behind the code, even inside a dark layout, so the quiet zone survives the design.
- One code per panel: two codes split attention and measurably reduce scans on both.
- Brand color outside the code: the code stays dark-on-light even when the rest of the collateral is saturated.
If a stranger cannot tell in three seconds what happens when they scan and what they get for it, the design is not finished.
Complete A/B Testing Blueprint
Print is expensive enough that test design matters more than artwork. This is the sequence that works.
- Define the metric before you print. Primary: reviews posted per 1,000 exposures. Secondary: scan rate, scan-to-review rate, cost per review. Raw scan counts are a diagnostic, never the goal.
- Set the baseline and the target. Run one variant for two weeks to establish the baseline, then aim for a 25–30% relative lift. Chasing a 5% difference on 60 conversions is noise, not a result.
- Change one variable per test. Frame copy, code size, placement, and channel should be tested separately. Two variables max, or you will not know which one moved the number.
- Split the audience cleanly. Alternate variant A and B by matched location, week, or day of week, and print only two versions. Give each cell its own code with a unique URL and UTM. Rotating by week neutralizes seasonality, but never launch during a promotion, a sale, or a holiday week.
- Run it long enough. Minimum six weeks and 100 conversions per arm. If the baseline interaction rate is low, extend to eight weeks rather than calling a winner early.
- Instrument it properly. Log impressions where you can estimate them, log scans separately from review submissions, and make sure the Google Business Profile panel reflects the same period. A spike in scans with no spike in reviews means the landing page is the problem, not the print.
- Agree on the decision rule before launch. Ship the winner only if the lift clears your threshold and holds in the same direction across at least two consecutive blocks. If the result is inconclusive, test placement next; placement beats copy for print and it is cheaper to change.
| Baseline interaction rate | Required exposures per arm | Minimum test length |
|---|---|---|
| 2% | ~3,000 | 8–10 weeks |
| 5% | ~1,200 | 6–8 weeks |
| 10% | ~600 | 6 weeks |
| 20% | ~300 | 4 weeks |
Print preflight checklist. 1 inch minimum size, or 3+ inches for signage. Dark modules on a light background at 7:1 contrast. Four-module quiet zone. Level Q error correction if a logo sits in the center. 300 dpi or vector export. Dynamic code pointing at a domain you own. One unique code per test cell. Test-print at actual size and scan it from the intended distance, in the intended light, on the oldest phone you can find.
The bottom line: printed QR codes and shortlinks are not competitors, they are two distribution systems for the same ask. Print wins where the customer is physically present and the moment is emotional. Shortlinks win where you have a list and the customer is at their phone. Get the print spec right, frame the code so strangers act on it, and use a dynamic code so your next campaign does not require a reprint. Then let a six-week test tell you where to put your money. The free GR Review widget gives you a branded review destination to point those codes at, and the free alternative to paid review widgets covers the display side.