Conference Badge Design & Printing: A Complete Guide

A conference badge does three jobs for eight hours straight: it identifies a person at conversation distance, it grants access at scan distance, and it starts conversations that wouldn't otherwise happen. Most badge guides are listicles of taste. This is a production guide: sizes, legibility math, QR engineering, stock and printers, and the pre-print-vs-on-demand decision — the things that determine whether the badge works.

Sizes: pick one that fits the holder you can actually buy

The de-facto standards, and where each fits:

Three practical rules beat any aesthetic preference: match the size to holders and lanyards you can source in quantity (an exotic size is a supply-chain problem wearing a design award); portrait hangs better on a single-clip lanyard (landscape badges seesaw); and double-clip or double-sided — a single-point badge spends half its life backwards, which is why the most useful "design" decision is printing the name on both sides.

Legibility: the actual math behind the font-size debates

Badge guides contradict each other (28pt! 72pt!) because they skip the underlying rule. Signage practice uses roughly 25 mm (1 in) of letter height per 3 m (10 ft) of reliable reading distance. Apply it to how badges are actually read:

• Conversation distance (1.5–3 m): you need ~8–12 mm cap height for the first name → roughly 30–40 pt. This is the number that matters.
• Across-the-room distance (4–6 m): you'd need 30–50 mm letters — a 72 pt first name (~25 mm) gets you to ~3–4 m, at the cost of the entire badge face. Worth it only for name-first networking events.

So the hierarchy, in type sizes that fit a B7 badge: first name 30–40 pt bold, last name 16–20 pt, company 12–16 pt, role/title 10–12 pt. First name biggest because that's what conversation uses; company above title because "who do you work for" is the second question, not "what's your title".

The rest of the legibility checklist:

• High contrast, always: dark text on light ground. Text over photography or gradients fails at a distance of exactly one lobby.
• Role color-coding needs shape redundancy. Speaker/staff/attendee/press color bands work — but ~1 in 12 men has some color-vision deficiency, so pair color with a word or icon on the band, and pick colorblind-safe pairs (blue/orange beats red/green).
• Arabic and bilingual events: set names in a typeface designed for Arabic (Tajawal, Cairo, IBM Plex Sans Arabic) rather than a Latin face's bolted-on glyphs, and test mixed-direction names (Latin company, Arabic name) on one physical proof before the run.
• What to leave off: email, phone, full titles, anything sensitive. The badge is public by definition — contact exchange is what the QR is for.

The QR code: engineering, not decoration

The QR earns its badge space only if it scans first-time under bad lighting on a moving lanyard. Four specs decide that:

1. Size: minimum 20 × 20 mm printed, and the rule of thumb for distance is code width ≈ scan distance ÷ 10 — a 25 mm code scans comfortably at ~25 cm, exactly the kiosk-and-handheld range. Under 20 mm you're gambling on camera quality.
2. Error correction: level M (≈15% restoration) as the default, Q (≈25%) if badges live a rough life — lanyard wear, rain, exhibit-hall grime. QR error correction has four levels (L≈7%, M≈15%, Q≈25%, H≈30% — per DENSO WAVE, the format's inventor); higher levels survive more damage but add modules, so at fixed print size each module shrinks. M is the standard trade-off; jump to Q only with print size to spare.
3. Quiet zone: 4 modules of empty margin on all sides. The most common badge-QR failure is decorative — a designer trims the white margin or lays the code over a brand pattern, and scan rates quietly collapse.
4. Payload: encode a short opaque token, not a data dump. Shorter payloads → lower QR versions → bigger modules at the same print size → faster scans. A per-registration token that the system resolves also means the badge leaks nothing if photographed — unlike badges that encode a vCard with the attendee's phone and email in plaintext.

Print production: the specs your printer wishes you knew

• Artwork at 300 DPI minimum, in CMYK for offset/inkjet runs. RGB artwork shifts color in print — the brand navy that turns purple on 400 badges is a classic.
• 3 mm bleed on all sides plus a 3–5 mm safe zone inside the trim — badge holders cover edges, and holder windows vary by a millimeter or two.
• Stock: 300–350 gsm coated cardstock is the workhorse (stiff enough not to curl on a lanyard). Synthetic stock (Teslin/PVC-class) for outdoor, multi-day or wet environments. Direct-thermal label stock only for on-demand text badges — it fades in sunlight and heat, so it's a day-badge material.
• Finish: matte laminate or holders beat gloss — gloss glares under stage lighting and camera flashes, which is also a QR-scan problem.

Printer decision, by class:

Pre-print, print-on-demand, or hybrid (the right answer for most conferences)

• Pre-print everything when registration closes early and no-shows are low. Cheapest per badge, best brand quality — but every no-show is a printed badge in the bin, every walk-in is a Sharpie badge, and every typo is a reprint queue.
• Print everything on demand when walk-in share is high or data changes late. Zero waste, always correct — but you've moved printing into the arrival peak: an 8-second print inside a 30-second check-in cycle caps each lane's throughput, so you buy more printers and lanes to hold the line (the check-in math covers exactly this).
• The hybrid most conferences should run: pre-print branded blanks — background artwork, event name, sponsor band — and print only the variable data (name, company, role, QR) at check-in on a transparent overlay or directly onto the blank. Brand quality of offset, correctness of on-demand, and the on-demand step is fast because it's laying down text, not flooding a page with color.

How this looks in practice

This hybrid is exactly how badge production works in Novex: the badge designer is a drag-and-drop template editor with millimeter-stored positions — size presets (B7 88×125 default, up to custom sizes), name/title/company/QR/logo fields, per-field sizes and alignment, Arabic-capable fonts, and background artwork. The same template then drives all three production paths: the emailed eBadge PDF each confirmed attendee gets (with a per-registration QR token), the bulk pre-print that renders one multi-page PDF for the whole event in a single run, and the kiosk print at the door, which automatically strips the background and logo and prints just name, title, company and QR — the transparent-overlay pattern — onto your pre-printed blanks. Template changes propagate to every badge produced after the change, which is what makes the T-14 badge lock date real rather than aspirational.

Day-of badge logistics (the part that shows up in reviews)

• Lock badge data T-14 to T-7 for pre-print runs; late registrations after lock go to the on-demand station by design, not by panic.
• Order 3–5% spare stock — spoilage, reprints, VIP whims. For pre-printed badges, sort by first name (people give you their first name) unless check-in is QR-driven, in which case sorting doesn't matter — another quiet argument for QR.
• Run a reprint station separate from the main lanes: one printer, one staffer, handles typos, lost badges and title vanity without stalling a queue.
• Proof one physical badge before the full run — stock, margins, QR scan test on three different phones, lanyard hang test. Five minutes that regularly saves four figures.
• Recycling bins at the exits on the last day: lanyards and holders are reusable next year; attendees will hand them back if a bin makes it effortless.