Google Review Card Design And Copy Guide

Why design is a conversion decision, not a brand decision

Every review card dies the same quiet death. It is born in a design review where everyone admires the kerning, ships to the counter looking gorgeous, and then sits there ignored while customers walk straight past it — because not one of them can tell, at a glance, what the card actually wants them to do. The post-mortem always finds the same cause: the design team treated the card as a micro brand asset, the operations team treated it as a tool for getting reviews, and the design team won, because a beautiful card photographs better in a review meeting than a card that merely works. Both framings are right; they simply point to different layouts, and the prettier one tends to win the argument in the room while quietly losing it on the counter.

• A review card lives in a moment of compressed attention: three to eight seconds between a customer finishing service and choosing what to do next. Branding only works if the action is understood first. A card that takes 12 seconds to parse loses most customers to the next thing demanding their attention. The pickup alert, the queue behind them, the phone notification, the partner at the door.
• The card is held in one hand and read at arm's length, often in mixed lighting (warm lobby lamps at 2700K, cool countertops at 4000K, outdoor patios in full daylight at 6500K, or half-lit dining rooms during brunch). Type that looks crisp in a Figma mockup on a calibrated display can become illegible on a printed card under incandescent lighting through reading glasses. Every design decision must survive the lighting it actually meets.
• Conversion-first design does not mean ugly design. The best cards read as premium because the action is obvious. Clarity feels confident, ambiguity feels amateur. The 'beautiful cards with no call to action' problem is the single most common design-review outcome in operator-facing luxury brands, and it costs real review volume that would have been available with ten minutes of layout surgery.
• Audience skew matters: a retirement-community pharmacy card and a hip coffee shop card both benefit from conversion-first design, but their execution is different. Older audiences need larger type, higher contrast and fewer abstract icons. Younger audiences tolerate cleaner minimalism but are faster to bounce if the action is not obvious within a second.
• Design and operations should own the card jointly. Design-only ownership produces beautiful cards that do not convert; operations-only ownership produces clear cards that do not feel like the brand. Joint ownership with design setting the visual system and operations holding a veto on legibility and clarity is the structure that ships good work.

The layout hierarchy that actually works

A review card has roughly three bands of attention. Getting the hierarchy right is the difference between a tap rate that feels natural and one that needs a staff prompt every time, and the hierarchy is robust across verticals. It works the same way for a café as it does for a hotel front desk, a dental practice or a car dealership service drive.

• Top band: promise and action: a short line that tells the customer what happens when they tap. 'Leave us a Google review. Tap or scan' is clearer than 'We'd love to hear from you.' The action verb is non-negotiable, because the human reading-at-a-glance reflex latches onto the verb before it parses anything else. Weak openers ('Thank you for visiting', 'Your feedback matters') occupy prime eye-path real estate with no conversion value.
• Middle band: visual affordance: an NFC tap icon with a phone silhouette, plus a QR code sized at a minimum of 2 cm square (ideally 2.5 cm for mixed audiences). Both affordances together teach the customer how to use the card without reading instructions. The visual alone tells them it is tappable and scannable. Cards with text-only instructions and no affordance icons convert 30-40% lower.
• Bottom band: brand and reassurance: logo, business name, and one reason to bother ('help a local café stay open', 'a 30-second review helps us a lot', 'your note helps future guests'). Brand lives here because it only matters once the customer has decided to engage. Inverting the hierarchy (brand at top, action at bottom) is the single most common mistake in cards designed by agency teams unfamiliar with conversion-focused layout.
• Avoid edge case: QR codes below 2 cm are unreliable from most phone cameras at natural card-reading distance. Test at 40 cm, not 10 cm — the 10 cm test passes everything and tells you nothing about real-world scanning behaviour. Below-2-cm QR codes also fail more often in dim lobby light and on warm-coloured backgrounds that reduce camera contrast.
• Avoid edge case: NFC tap icons placed near the corner often invite taps on the wrong face of the card. Centre the tap target visually, and ideally place the antenna where the icon suggests. If the icon is centre and the antenna is top-right, customers will tap the centre and fail, then give up. Antenna-to-icon alignment is a physical design decision that Figma mockups cannot show.
• Depth of field: the action line and affordance icons should be at the top layer of the visual stack, unobstructed by any decorative element. Watermarks, background gradients and photo overlays behind the action always reduce conversion because they compete for eye attention at the exact moment the card is being parsed.
• Rotation tolerance: the card will be read in both portrait and landscape orientations depending on placement. The primary action line should read correctly in the dominant placement orientation, but the secondary affordance (QR code) should be rotation-neutral. Dual-orientation designs that work both ways are slightly more complex but survive a wider range of countertop realities.

Copy that survives real customer attention

The copy on a review card has to be readable by a tired guest, an elderly customer, a tourist whose first language is not your brand's language, and a distracted parent holding a child who is actively trying to leave. That constraint quietly rules out most of the clever phrasing a creative team wants to try, and the most durable copy comes from systematically subtracting words rather than adding them. On a review card, the writer's real skill is deleting their own favourite line without flinching.

• Keep the main line under eight words. 'Scan or tap to leave a Google review' works; 'Your voice helps our little family-owned shop thrive in difficult times' does not. Longer copy is read less often and acted on less often. Every additional word is a 5-10% drop in the probability that a customer reaches the action. The word-count constraint is a feature, not a limitation.
• Use an imperative verb. Tap. Scan. Leave. Share. Not 'consider leaving' or 'we appreciate if you could'. Imperative verbs work across reading levels, work through a second language, and read as confident rather than apologetic. Conditional phrasing ('if you wouldn't mind', 'only if you have time') signals that the ask is unimportant and cuts conversion by up to half.
• Name the platform. 'Google review' converts better than 'review' because customers recognise the brand and trust the destination. They know they are not about to be routed to a private survey, a newsletter signup or a data-harvesting form. Generic 'leave a review' copy always raises the 'where does this go?' question, and the three-second decision window rarely survives that question.
• Keep the secondary line practical, not emotional. 'Takes about 30 seconds' beats 'means the world to us', because the practical line answers the silent question ('how long will this take?') that every customer evaluates before engaging. Emotional copy feels like it should work but tests consistently worse, and it ages badly. The same emotional line that feels warm in January reads as desperate in November.
• Avoid jargon. 'Tap here' is clearer than 'Engage NFC prompt'. 'Scan the code' beats 'Scan the QR' with less tech-literate audiences. Technical vocabulary belongs in product datasheets and engineer-facing documentation, not on customer-facing cards. Plain verbs plus the right affordance icon carry more information than any amount of technical copy.
• Internationalise carefully. In markets with mixed language customers (tourist-heavy cities, border regions, international airports), a short bilingual line is better than a dominant-language line with a smaller translation. Stacking two 5-word lines with equal weight outperforms stacking one 8-word line in the dominant language with a 4-word line in secondary type beneath it.
• Read-aloud test: if the copy does not read naturally out loud, it will not work when a staff member needs to echo it during a verbal prompt. Copy that a staff member cannot say with a straight face never makes it onto the card in practice. The team silently drops the line and the programme loses the verbal reinforcement that depends on it.

Visual system: icons, type and colour contrast

The visual system is where most review cards fall apart in real placements. A design that looks crisp in RGB on a calibrated design-studio display can lose its affordance once it is printed on a reflective laminate under a warm incandescent bulb, because the physical environment of the card is always harsher than the review environment of the design.

• NFC icon: use the Wi-Fi-style arc plus a phone silhouette, or the N-Mark universal NFC indicator with a 'tap' verb next to it. A generic contactless symbol alone is too abstract for many customers, particularly older audiences and customers from markets where contactless payments are less normalised. The icon's job is to say 'put your phone here' without a reading step.
• QR code: minimum 2 cm square (2.5 cm for older audiences or lower-light placements), quiet zone of at least 4 mm, high contrast between code and background. Reverse-contrast codes (light on dark) work but need 20% larger sizing to scan reliably because many phone cameras under-expose dark-background codes. Embedded logos in QR codes are fine up to about 10% of the code area; more than that breaks error-correction reliability.
• Type: sans-serif, minimum 10 pt for the main line, minimum 7 pt for secondary — 11 pt and 8 pt are safer if the card is commonly read by older audiences. Serifs and decorative faces read as 'fancy branding' and dilute the action; humanist sans-serifs (Inter, Proxima Nova, Avenir) carry both clarity and brand warmth. Avoid ultra-thin weights below 400 — they disappear on coated laminates.
• Colour contrast: target WCAG AA (4.5:1) between headline copy and card background, and push to AAA (7:1) if the audience skews older or if the card will be read in low-light placements (spa corridors, dim restaurants, lobby bars). Many premium brand palettes (cream on gold, silver on white, tint-on-tint) fail this test; fix the card even if the broader brand keeps the palette, because the card has to work in a way the corporate brochure does not.
• Finish interaction: gloss lamination creates glare under lobby spotlights, ceiling halogens and bright windows. The card becomes unreadable at exactly the angle it is most often picked up. Matte or soft-touch laminate reads more reliably in mixed lighting and also feels more premium in hand, which is a nice side effect of the functional choice.
• Dark mode bias: phones are often viewed in auto-dark mode, and customers who check the QR code destination on their phone see it rendered in their phone's current theme. Cards with light text on dark backgrounds still read fine because the physical card is unaffected by the phone's mode; cards with low-contrast mid-tones become unreadable because the already-marginal contrast collapses under off-angle glare.
• Print variance: the same CMYK spec produces visibly different results across print vendors. Establish a Pantone match for the primary brand colour and a specified paper stock, and verify the first batch from any new vendor against the master sample. Print-to-print drift is the silent reason a third reorder looks subtly wrong against cards from the original batch.

Balancing brand identity against the action

The hardest conversation in review-card design is about how much brand can live on the card without hiding the action. These are the principles that tend to survive that conversation, and they are worth memorising because every review-card design review tries to re-litigate them from scratch.

• Brand lock-up stays on one face. Usually the front. The back face carries the tap or scan affordance with less brand overlay. Two-sided cards let the brand present formally on one side and the programme work on the other, and the face-up orientation can default to whichever side matches the placement (counter mats face the affordance up; handed cards lead with the brand).
• Pattern elements (background textures, gradient washes, watermarks, brand patterns) belong on the edges, not behind the action. A decorative element that forces the action copy down to 8 pt is the wrong trade, even if the decorative element is the signature brand asset. The card is not a brochure; it is an instruction surface.
• Colour: the brand's accent colour is a great home for the NFC icon and the QR frame. A small accent-coloured ring around the QR code, for instance, or a tint behind the tap-here line. Using the accent colour to highlight the action aligns brand and conversion instead of fighting them, and it solves the 'the card doesn't feel like us' critique without moving the action.
• Photography: if the card carries a photograph, reserve it for the back face. Photographs on the action face almost always reduce tap rate because the human eye lands on the photo first, processes the scene, and only then looks for instruction. A photo also adds visual noise that works against QR scan reliability under low-contrast printing.
• Debadging: resist the urge to remove the 'Google' wordmark for design cleanliness or regulatory squeamishness. Customers recognise Google trust more than they recognise your logo; hiding it costs more than it gains, and the 'Google' wordmark is available under Google's brand-asset licensing for exactly this use. Design directors sometimes remove it out of habit, and conversion loses every time.
• Material and finish as brand carrier: substrate choice (metal, wood veneer, soft-touch PVC, Mohawk paperboard with letterpress detail) carries brand identity without taking visual-plane real estate. Premium materials lift brand perception while leaving the front face clean for action-focused layout. This is where luxury brands should invest their brand-identity budget on a card.
• Legibility trumps brand identity in every edge case. If a design decision between 'more brand' and 'more legible' is genuinely close, choose legible. The programme owner's job is to make sure this is the default, because design reviews routinely tilt the other way.

Proof review and sample validation

The proof round is where design decisions meet printed reality. Running it well turns design into decision faster than any mockup can, and skipping or shortcutting it is the single biggest reason operators end up reprinting cards within 90 days — defects that were invisible on screen become obvious on a counter.

• Print two to four variants: the current favourite, one more conversion-first variant (larger action line, smaller brand), one more brand-first variant (prominent brand, tighter action). Compare in the real placement, not on a design review call. Placement reveals things a conference-room review cannot. Sightlines, reflection angles, cohabiting with receipt printers and tip-jar signage.
• Test QR scan at 40 cm on three phones: a modern iPhone (current generation), a modern Android (Samsung or Google Pixel), and a three-year-old mid-range Android with an older camera module. If any of them misses, sizing or contrast needs to change. Also test with a customer wearing reading glasses. The card will meet that customer regularly, and the scan reliability varies.
• Test NFC tap with the card face down on a bag, face up on a counter, and held at waist height over a handbag strap. Antenna location matters: some cards only tap reliably when held in a specific orientation, and customers who fail to tap on the first try often give up rather than rotating the card. Four-corner antenna coverage is ideal; centred antenna with a matching icon is second best.
• Read-aloud test: hand the card to three people who have never seen it and ask them to describe what they would do next. If they do not say 'tap' or 'scan' inside two seconds, redesign the action line. This is the cheapest, fastest test in the proof round and it catches the 'the design is beautiful but what do I do?' problem before printing 5,000 units.
• Lighting test: review the card under warm white (3000K like a lobby lamp), cool white (4000K like an office ceiling fluorescent), daylight (6500K like a sunlit patio), and mixed under-cabinet halogen. Colour contrast that passes under one lighting condition can fail under another. Cream-on-gold often fails under warm white, deep navy on black often fails under daylight glare.
• Durability test: leave a card on a busy counter for 30 days in the real placement. Check for scuffing, ink rub-off, QR-code degradation, NFC-antenna failure from repeated contact. Laminated cards last; unlaminated cards often show visible wear at 20 days. Substrate choice is a durability decision.
• Printing tolerance: the proof batch should check registration (alignment of front-to-back print), colour fidelity against the Pantone master, and the QR-code module clarity under 4× magnification. Any misalignment or blur in the proof will be worse at volume.

Technical production specs — prepress files, colour profiles, dielines and NFC antenna cutaway

A beautifully designed card can still fail in production if the prepress file is wrong. The print vendor needs a file that survives their CTP (computer-to-plate) workflow, their press calibration and their lamination line without colour drift, registration shifts or antenna interference. Getting the technical specs right the first time saves two to four weeks of reprints and avoids the 'why does this not match our website?' crisis at the first delivery.

• File format: PDF/X-4 is the modern standard (supports transparency, layers, ICC-managed colour), PDF/X-1a is the safer fallback for older presses (flattened, CMYK-only, embedded fonts). Adobe Illustrator and InDesign export both cleanly; avoid Figma-exported PDFs for production because Figma's PDF export does not preserve print-critical metadata (trim box, bleed box, colour intent).
• Dimensions and bleed: standard credit-card is 85.60 × 53.98 mm (CR80 ISO/IEC 7810). Add 3 mm bleed on each side (trim to 91.60 × 59.98 mm total). Safety zone 3 mm inside the trim edge for all critical text and graphic elements; anything inside the safety zone is guaranteed to land on the final card. Print-ready file delivered at the bled size with crop marks at 3 mm offset.
• Colour profile: ISO Coated v2 (FOGRA39) for European printers, GRACoL 2013 CRPC6 for North American, Japan Color 2011 for Japanese printers. Provide an embedded ICC profile in the PDF; a CMYK file with no profile renders unpredictably because every press assumes a different colour intent in the absence of a specified profile.
• Pantone matching: the primary brand colour almost always needs a Pantone spot match because CMYK simulation drifts 5–15% between presses. Specify both a Pantone solid-coated reference (e.g. PMS 286 C for a typical brand blue) and a CMYK fallback for any press that does not run spot colour.
• Raster vs vector: logos, icons and wordmarks should be vector (Illustrator .ai or embedded SVG in the PDF). Photography and continuous-tone artwork is raster at 300 dpi at 1:1 scale at the final printed size. Raster at less than 300 dpi pixellates visibly on a printed card; vector at any resolution stays crisp because it is re-rasterised at press resolution.
• Overprint settings: set black text to overprint (so black ink prints on top of underlying colours rather than knocking them out, which avoids white halos on misregistered presses). Rich black for large filled areas is C60 M40 Y40 K100 (not K100 alone, which looks grey); avoid rich black for small text because registration shifts show up as colour fringes.
• Laminate and finish specifications: specify the laminate type (matte, soft-touch, gloss, spot UV), thickness (7 mil for typical PVC, 10 mil for premium), and any finish effects (foil stamp colour, emboss depth in thousandths of an inch, spot UV mask as a separate die file).
• NFC antenna dieline: the NFC antenna (typically a copper or aluminium loop around the card perimeter or a spiral in the top-third) must be visible on the dieline so the print file avoids printing heavy metallic inks over the antenna, which detunes the resonance. Most NFC chip vendors (NXP NTAG 216, ST M24SR, Infineon SLE66CLX) provide dieline templates for standard antenna geometries.
• Proof approval workflow: a hard proof (actual printed card from the actual press on the actual substrate) should be approved before the full print run. A digital proof (PDF on a calibrated display) is a preview only; colour on a press never matches colour on a display, and the proof is the only reliable way to see what will ship.

A/B testing, sample sizing and conversion benchmarks across verticals

Card design decisions often stall at 'the team prefers A, the owner prefers B'. An A/B test in the actual placement settles the argument with data, and the measurement infrastructure is cheap to build because the review redirect already logs taps. The discipline is in sample sizing, interpretation and avoiding the common statistical mistakes that make A/B tests declare winners that are not real.

• Test structure: print 200-500 cards of each variant. Split placement evenly (both variants at the counter simultaneously, or alternating weeks if placement space is limited). Run for at least 4 weeks (restaurant, salon, dental) to 8 weeks (low-frequency verticals like auto, luxury hotel). The URL redirect carries a variant parameter (?v=a, ?v=b) so tap counts are attributable.
• Minimum detectable effect (MDE): a 30% relative lift (e.g. 4% vs 5.2% conversion) needs about 800 total taps per variant at 80% power and 5% significance. Smaller lifts (10-15%) need 3,000-5,000 taps per variant. Larger than most cafés or single-location businesses can generate in a reasonable test period. Multi-location groups have the volume to detect these, independents usually do not.
• Sample-size calculator reference: the Optimizely, VWO or Evan Miller calculators give rigorous numbers given your baseline conversion rate and desired MDE. A 5% baseline conversion with a desired 20% relative lift at 80% power needs roughly 1,500 taps per variant — 4 weeks at a medium-busy café.
• Statistical significance: require p < 0.05 (or the Bayesian equivalent, 95% posterior probability) before declaring a winner. Declaring a winner at 60% confidence after 100 taps per variant is the most common mistake and produces false positives the team then amplifies in subsequent design decisions.
• Café benchmark: typical card-driven review conversion 3-7% of transactions. Baseline (no card) is often 0.3-0.8%, so a well-designed card delivers 5-10× baseline. Cafés with baristas who verbally prompt regularly hit the top of the 3-7% range; cafés with silent placement-only cards hit the bottom.
• Dental and specialist healthcare benchmark: 5-12% of checkouts. Higher than café because the relationship is deeper and the checkout has more dwell time. Pediatric dentistry skews higher (parents review readily); specialist practices (ortho debond, implant case close-out) can hit 18-25% conversion at milestone moments.
• Hotel benchmark: 4-8% of checkouts (full-service), 2-4% (select-service), 10-15% (luxury with concierge-delivered prompt). OTA-booked guests have lower rates because they often review through the OTA instead of Google. Loyalty guests convert at 1.8-2.2× walk-in guests.
• Fitness benchmark: 2-5% of check-ins for self-serve turnstile stickers, 8-15% of class attendees for instructor-led prompts at cool-down, 15-25% for PT-milestone asks. Member-tenure segmentation matters: 0-60 day members convert 1.5-2× long-tenured members.
• Salon benchmark: 7-15% of services for stylist-led ask at reveal, 3-8% for reception-led ask at checkout. Volume-format salons hit the bottom of the range; luxury colour salons hit the top. Spa services convert slightly below hair (5-10%) but produce longer, more detailed reviews.
• Automotive benchmark: 6-10% of sales deliveries (F&I + delivery combined), 4-7% of service visits (advisor handoff). EV deliveries convert higher because the delivery walkthrough is longer and the relationship is warmer. BDC follow-up SMS adds 20-35% incremental volume on top of the physical card alone.

Step-by-step: design and proof a card from blank slate to print-ready

The fastest path from 'we want a review card' to a printed batch on the counter is a six-step workflow that pre-empts the rework most teams discover only after their first proof or first 30 days of live service. Treat each step as a gate. Skipping a gate to save a week typically costs three weeks downstream because the failure surfaces only after print, when reprinting is the only fix.

• Step 1 — Capture the per-location Google review URL: in Google Business Profile (business.google.com), open the location, click 'Read reviews' or use the 'Get more reviews' card on the dashboard, copy the short URL (g.page/r/...). This URL is the anchor for everything downstream. Wrap it in a brand-owned redirect (review.brandname.com/location-slug) so the printed URL stays stable even if Google's short-link format changes. Verify the redirect resolves from a phone outside your network before any artwork is exported.
• Step 2 — Lock the layout hierarchy: in Figma, Adobe Illustrator or Affinity Designer, set up an 85.60 × 53.98 mm CR80 artboard with 3 mm bleed. Place the action line ('Scan or tap to leave a Google review') in the top band at 10-11 pt sans-serif. Place the NFC icon and 2.0-2.5 cm QR code in the middle band with 4 mm quiet zone. Place the brand mark and one-line reassurance in the bottom band. Do not invert the hierarchy even if the design review meeting prefers brand-on-top.
• Step 3 — Run the contrast and legibility test: open the WebAIM contrast checker (webaim.org/resources/contrastchecker/) and verify the headline against the background reads at WCAG AA 4.5:1 minimum, AAA 7:1 for senior-skewing audiences. Print at 1:1 scale on a desktop colour printer, hold the print at 40 cm under a 3000K warm bulb, and confirm a person wearing reading glasses can read the action line without leaning in. If they cannot, increase type size before exporting the press file.
• Step 4 — Export press-ready PDF/X-4: in Illustrator or InDesign, File > Export > PDF/X-4:2008, embed the ICC profile (FOGRA39 for European printers, GRACoL 2013 CRPC6 for North American), set black text to overprint, include 3 mm bleed and crop marks at 3 mm offset. Verify the file in Adobe Acrobat Pro Output Preview or callas pdfToolbox before sending to the print vendor. PDF/X-4 catches missing fonts, RGB images and unflattened transparency that PDF/X-1a silently drops.
• Step 5 — Order a 2-4 variant proof batch: send the press-ready file plus at least one alternate (one more conversion-first, one more brand-first) to the print vendor. Order 50-100 units of each variant, not the full run. Test in the real placement under real lighting for two weeks. Read-aloud test with three customers who have never seen the card; QR scan test on a current iPhone, current Android and a 3-year-old budget Android; NFC tap test face-up, face-down and on-bag.
• Step 6 — Approve the hard proof and place the volume order: a hard proof (actual printed card from the actual press on the actual substrate) is the only reliable colour and tactile reference. Approve in writing against the Pantone master sample. Place the volume order at 110-120% of forecast first-quarter need. Not the typical 150-200% over-order that produces dead stock when the artwork drifts in the year-one refresh.

Pre-print checklist — verify these 12 items before the volume order

The most expensive review-card mistakes are the ones that survive the proof round and only surface after 5,000 cards are sitting in a warehouse. The checklist below is the structured version of what experienced design and operations teams check before signing off the volume order. Run through it as a yes/no list with the print vendor on a 30-minute call; the cost is zero and the failure modes it catches are unrecoverable after print.

• Per-location Google review URL captured, wrapped in a brand-owned redirect, and verified to resolve from a phone outside the office network within 2 seconds.
• Layout hierarchy locked: action line at top, NFC + QR affordances in middle, brand at bottom. Action line under 8 words. Imperative verb. Names 'Google' explicitly.
• QR code at minimum 2.0 cm square (2.5 cm if audience skews 55+), 4 mm quiet zone, embedded logo (if any) under 10% of code area. Tested on three phone models at 40 cm.
• Type size: headline at 10-11 pt sans-serif, secondary at 7-8 pt. Avoided ultra-thin weights below 400. Read-aloud tested with three first-time viewers in 2 seconds.
• Contrast: headline-to-background passes WCAG AA 4.5:1 (AAA 7:1 for senior audiences). Verified in WebAIM contrast checker against the actual hex values, not the brand-deck approximation.
• Press file: PDF/X-4 (or PDF/X-1a fallback), embedded ICC profile (FOGRA39 EU / GRACoL NA), 3 mm bleed, 3 mm safety zone, vector logos, 300 dpi raster artwork at 1:1 scale.
• Pantone match for the primary brand colour, with CMYK fallback for non-spot presses. Hard proof on the actual substrate at the actual press, approved in writing against the Pantone master sample.
• Substrate and laminate specified: PVC (standard), soft-touch PVC (mid-tier), wood veneer or metal-look (premium). Matte or soft-touch laminate to avoid glare; gloss only if the brand demands it and the placement is glare-controlled.
• NFC chip family chosen and antenna dieline visible in the press file: NTAG 213 (144 bytes user memory) for static URL only, NTAG 215 (504 bytes user memory) for richer NDEF, NTAG 216 (888 bytes user memory) for multi-record payloads. Avoid heavy metallic ink over the antenna loop.
• Per-location URL programmed onto the NFC chip, encoded in NDEF Type 2 format, locked to read-only. Tap-test 5 random cards from the proof batch on iPhone (iOS 17+) and Android (Android 13+).
• On-metal mitigation if the placement surface is steel or aluminium: ferrite-laminated label, 3-5 mm non-metallic spacer, or moved to a non-metal adjacent surface. Documented in the placement plan, not assumed.
• Version number printed in 4-5 pt type on the back of the card (e.g. 'v1.0 Q2-2026') so future audits can identify drift across reorders. Cheap insurance that avoids the 'which batch is this from?' question two years in.

Common mistakes to avoid — seven patterns that consistently kill conversion

The mistake patterns below are the ones we see repeatedly across operator reviews of cards that under-perform. Each one is invisible in the design studio and obvious only in the live placement. Memorise them and the design review meeting becomes much shorter, because most of the bad ideas the meeting will surface are versions of one of the seven below.

• Brand identity dominates the action: the most common mistake. The brand mark sits at the top of the card, the action line is squeezed underneath in 8 pt, and customers register 'logo' before they register 'tap'. Conversion drops 30-50% versus an action-first layout, and the brand impression does not improve because customers never engaged. Fix: brand at the bottom, action at the top, no exceptions.
• Generic 'leave a review' copy without naming Google: customers ask the silent question 'where does this go?' and the three-second decision window rarely survives the question. 'Google review' converts 15-30% better than 'review' alone in comparison tests. Fix: always name Google, always include the imperative verb (tap, scan, leave).
• QR code under 1.5 cm: fails on older phones, in dim lighting, on warm-coloured backgrounds. The 10 cm test in the studio passes everything; the 40 cm test under warm lobby light is the reality. Fix: 2.0 cm minimum, 2.5 cm safer for senior audiences, with 4 mm quiet zone.
• Decorative elements behind the action: watermarks, gradient washes, photographs and brand patterns competing for eye attention at the exact moment the card is being parsed. Eyes land on the visual richness, then have to hunt for the instruction. Fix: keep the action band visually clean, push decoration to edges or to the back face.
• Antenna-to-icon misalignment: the NFC tap icon sits in the centre of the card while the antenna runs in a top-corner spiral. Customers tap the centre, fail, and give up. Most cards do not document antenna geometry; the design team assumes any spot works. Fix: confirm antenna location with the chip vendor or NFC inlay supplier and align the icon to the antenna physical location.
• Gloss laminate under bright placements: gloss creates glare at the angle the card is most often picked up, making the action line and QR code unreadable for the critical 1-2 seconds. Fix: matte or soft-touch laminate as default; gloss only when the placement is glare-controlled (low ambient light, no direct overhead spots).
• No version number on the back: when a third reorder looks subtly off against the original master sample, nobody can tell which batch is which. Drift accumulates across reorders and the brand visually fragments. Fix: 4-5 pt version mark on the back face from the first batch.

Multi-location and franchise consistency

A single-store review card decision is a design problem. A multi-location card programme is a system problem, and system problems need more rules than a single card does. Designers who are comfortable shipping one perfect card sometimes under-invest in the template governance that keeps fifty cards cohesive. Across a network, consistency is rarely a question of talent; it is a locked template and someone with the authority to tell a well-meaning franchisee that no, they may not 'just tweak' it.

• Template kit: a locked hierarchy with editable location name, review URL (encoded and printed) and optional accent colour. Franchisees customise inside the template; they do not redesign the template. The editable-zone boundaries should be visible in the template file itself, so a local print shop cannot accidentally edit outside them. Change control on the template lives with central brand, not with any location.
• QR URL per location: each card points to the specific Google Business Profile for its location via a per-location redirect URL. Central URL-shortening or tracking wrappers are fine (and actually helpful for measurement) as long as they use a brand-owned domain; generic brand URLs that point to the brand page instead of the location page are a programme failure and waste every scan.
• Language packs: for international brands, publish the template in the three to five most common customer languages (English, Spanish, Mandarin, Arabic, French cover most markets). Local teams pick the language pack, not the layout, because untrained teams making typographic decisions in a script they do not read consistently fragments the visual system.
• Central approval: the template itself requires central brand approval before print, and every new language pack requires central approval. Local customisations (address line, phone number, local review URL) do not require approval because they live inside the pre-approved editable zones. The approval gate is for template changes, not location content.
• Measurement baseline: every location reports review volume lift from the first 90 days of the card programme, compared to a matched 90-day window the prior year. Laggard locations get coaching on placement, staff prompt and timing. Not a new card design. Redesigning the card is almost never the fix for a location that under-performs the network median; the fix is usually operational.
• Substrate tiering: a multi-location programme benefits from a published substrate-tier matrix (flagship / standard / express) that maps location format to card substrate. Tier decisions are made centrally; individual locations do not substitute a premium substrate without approval. The tier matrix solves 80% of 'can we have a nicer card?' conversations before they start.
• Artwork version discipline: every printed batch carries a version number in small type on the back. V3.2, Q2 2026. When a location asks 'which version of the card are we on?', the answer is auditable rather than vague. Version discipline also supports the quarterly refresh process by making drift visible.

FAQ