QR codes in marketing: when they work, when they are a gimmick

Three numbers set expectations.

• Pew reported in May that 35 percent of US adults own a smartphone. Among adults under 45 with household income over $75,000, it is well over half.
• comScore reported that 14 million Americans scanned a QR code in June, which is 6.2 percent of mobile users. Most scans came from magazines, newspapers and product packaging, and most happened at home.
• Neither the iPhone nor most Android phones ship with a reader. People have to find and install an app such as RedLaser, ScanLife or QR Droid before a code does anything.

So the audience that can scan is a minority, skews younger and more affluent, and has to have taken a step most of your customers have not. That is not a reason to avoid codes. It is a reason to use them only where the payoff is worth that friction, and to never make the code the only way to reach something.

The common thread in placements that produce scans is that the person is standing still, has a reason to want more, and has time.

Other uses that are earning scans: business cards that save a contact to the phone, in-store displays linking to demonstration videos, and museum and venue labels that add audio or detail.

• On websites. The visitor is already online. A link does the same thing without an app.
• In email. Same problem, plus most people read email on the device that would do the scanning.
• On billboards and vehicle wraps. Nobody is scanning from a moving car at 60 miles an hour, and if they try you have a different problem.
• Where there is no signal. Subway platforms, parking garages, the inside of many big-box stores.
• Linking to your desktop home page. The number one failure. The scan works, a full-size site loads over 3G, and the person pinches and zooms for ten seconds before giving up.
• With no reason to scan. A naked code with no call to action tells the person nothing about what they will get.
• Too small or too distorted. Codes under about an inch square, printed on glossy curved surfaces, or stretched in a layout program often will not decode.

A practical rule: if you cannot write a four-word promise next to the code, such as "Scan for 20% off" or "Watch it assembled," you do not have a use for one.

1. Build the mobile landing page first. One screen, one message, one button, under 100 KB, with no Flash. Test it on an iPhone and an Android phone over cellular, not Wi-Fi.
2. Use a short URL you control. Fewer characters mean a simpler code that scans faster and prints smaller. A short link also lets you change the destination after the print run and gives you click counts.
3. Add campaign tags. Put Google Analytics parameters on the long URL behind the short link (utm_source=print, utm_medium=qr, utm_campaign=fall-catalog) so scans appear as their own source.
4. Print at the right size. Minimum one inch square for close scanning, larger for anything a person will hold at arm's length. Keep the quiet zone, the white border, intact.
5. Keep contrast high. Dark code on a light background. Reversed or colored codes can work but test them on three phones first.
6. Tell people what they get and that they need a reader. "Scan with your phone's QR reader for the full video" covers both.
7. Give an alternative. Print the short URL under the code so the 65 percent without a smartphone, and everyone without an app, can still get there.
8. Test the printed proof, not the PDF. Scan the actual press proof with multiple apps before approving the run.

Scan rates on print are low in absolute terms. On a magazine ad reaching 500,000 readers, a few thousand scans is a good result. On packaging, a scan rate of 1 to 3 percent of units sold is strong. On a direct mail piece with a personalized offer, 2 to 5 percent of recipients scanning would be excellent.

Judge the code the way you judge any response device: by cost per response compared with the alternatives. The code itself costs nothing to add, so the real question is whether the mobile page and the offer behind it are worth building. If the same page can serve a short URL in print, a link in email and a code on packaging, the economics usually work.

Record three things per campaign: scans by source, what the person did on the page, and whether they converted. Short-link counts give you the first. Analytics gives you the rest, as long as the tags are in place before printing.

• A code on a trade-show banner that links to a PDF brochure the phone cannot open.
• A code generated with a free tool that links to a long URL that was later changed, so every printed piece is now dead.
• A restaurant that put a code on the menu linking to its Facebook Page, which asks the diner to log in before showing anything.
• A direct mail piece whose code links to the same page as the printed URL, with no tags, so nobody can tell which worked.
• A code printed in brand orange on a kraft-paper bag that no reader could decode.

We treat a QR code as a bridge between print and a mobile experience, which makes it a connected brand, growth, web and technology job. Discover asks where your customers meet your print and what they would want at that moment. Plan picks the placements that pass the test above and defines the offer behind each. Build produces the mobile pages, the short links with tracking, and print-ready codes tested on real devices. Improve reads scan and conversion data by placement and moves the next print budget toward what responded.

A representative engagement is on our real estate work: a brokerage whose yard signs and listing flyers carried codes that opened a mobile page for each property, with photos, a floor plan and a one-tap call to the agent. Each listing had its own short link and tags, so the brokerage could see which signs produced calls. Scans were a small share of overall traffic, but the leads from them were among the most qualified the firm received.