QR codes for direct mail: personalisation, tracking, and the 2026 USPS discount

Direct mail never really went away, and pairing it with a QR code has changed what a mail campaign can prove. A postcard used to be a one-way message: printed, posted, and then a guess at whether anyone acted on it. A QR code turns the same card into a two-way channel, because every scan is a data point you can see, and a dynamic code lets you change the destination without reprinting a single piece.

This guide covers what agencies actually need when they put a QR code on a mail piece: generating a personalised code for every name on the list without hand-building thousands of links, tracking a mail campaign so the report holds up, and what the USPS's 2026 Integrated Technology Promotion really requires, because the version of that discount circulating in marketing blogs isn't quite the real one.

PURLs and QR codes are the same idea, two entry mechanisms

A personalised URL, or PURL, is a unique web address printed on a mail piece, usually built around the recipient's name (janesmith.yourclient.com). Because a human has to type it, a PURL is conventionally kept to around 50 characters or fewer, short enough to read off a postcard and enter correctly on a phone keyboard. A personalised QR code does the same job through a camera instead, so it carries no visible length limit and can encode a long, opaque tracking link.

Mechanically, both usually resolve to the same kind of tracked landing page. A dynamic QR platform encodes a short, managed URL; scanning it triggers a server-side redirect that looks up the current destination and logs the scan (device type, approximate location and timestamp) before sending the visitor on. That lookup is what a static, one-off code can't do: change the destination after the mailer is already in the postal system, or route different segments of the same run to different pages without touching the printed artwork. Our guide to dynamic versus static QR codes covers that distinction further.

Most personalised mail campaigns that use one increasingly print both, the PURL as a fallback for anyone who'd rather type than scan, while the QR code carries most responses on mobile. What decides whether either performs isn't the entry mechanism, it's the landing page: fast, relevant to the specific piece someone is holding, built to convert on a phone. Our guide to QR code landing pages that convert covers that half of the campaign.

Generating a unique QR code for every name on the list

Variable data printing, or VDP, is the print-industry term for a job where common design elements stay fixed across every copy while specific fields change from copy to copy. In its simplest form it's a mail merge: the same postcard, a different name and offer pulled from each row of a spreadsheet. A personalised QR code fits in as just another variable field. A VDP engine reads a data source (usually a CSV from the client's CRM) and calls a barcode function once per record, so each piece gets a QR code generated from that row's own tracked URL rather than one code shared across the whole run.

Building thousands of individually tracked URLs by hand is what actually blocks agencies from doing this well, which is why the batch needs generating programmatically. Our guide to generating QR codes in bulk covers producing a CSV of unique, trackable codes at the volume a real mail run needs.

Decide the personalisation grain before generating anything. A code can be unique per individual recipient, useful when the offer or landing page genuinely differs person to person (a renewal date, an account number), or unique per segment or mail wave, simpler and usually enough when the offer is the same for everyone in that batch. Per-recipient codes cost more to generate and reconcile, and only earn that cost if something downstream uses the granularity, such as pre-filling a form. If the report only needs "how many people from segment B scanned," segment-level codes with consistent UTM tagging do the same job for less overhead.

Tracking a mail campaign so the numbers hold up

A single print run, one poster, one flyer version, is straightforward: one code, one set of parameters, one number at the end. A personalised, multi-wave mail campaign is a different shape of problem, because the same campaign might run three creatives across four segments in two waves, and the client will ask which combination actually drove the response. If you're tracking a single print run rather than a segmented mail campaign, our guide to tracking a print campaign with QR codes covers that simpler case; this is about the added structure a mail-merge campaign needs.

The fix is a naming convention applied before a single code is generated, not reverse-engineered from the data afterwards. Every code needs to carry the client, the campaign, the mail wave or drop date, and the segment or list source, so a report can be sliced by any of those without re-tagging anything. Our guide to QR code UTM parameters sets out a naming scheme that scales across clients; apply the same discipline to a mail campaign's wave and segment fields rather than inventing a one-off scheme per client.

It's worth being precise with clients about what a scan actually captures, because "we track the QR code" gets over-promised in this industry. A scan logs the device type, an approximate location from IP address (city level, not a GPS pin), a timestamp, and which code was used. It doesn't identify the person unless the landing page itself captures that, a form, a login, a pre-filled parameter tied to their record. For a mail campaign, that pre-fill is usually the real value of going per-recipient: the code lets the page greet someone by name, and analytics attach the scan to that record on the back end, not because the code "knows" who scanned it.

Once live, the ROI conversation is the same as any other channel, with the mail cost added to the denominator: cost per piece (design, print, list, postage) against the value of conversions the scan data can attribute. Our guide to calculating QR code campaign ROI walks through that maths, including the response direct mail always gets even without a code, people who see the mailer and search for the brand later rather than scan on the spot.

The 2026 USPS discount, and the part most guides get wrong

The USPS runs periodic postage discounts for mailers who combine print with an interactive element, and the current one is the 2026 Integrated Technology Promotion, confirmed live under a published guidebook rather than a proposal. The headline is a 5% base postage discount, with two optional 1% add-ons (an Informed Delivery campaign, a sustainability element) for a maximum of 7%. A separate 0.5% incentive exists for electronic postage-statement submission, but that's a parallel programme, not part of the 7% ceiling.

The eligibility condition most blogs skip: the promotion doesn't reward having a QR code, it rewards using one of eight named technologies as the interactive trigger, each with its own content bar. The eight are augmented reality, mixed reality, virtual reality, a voice assistant, video-in-print, NFC, mobile shopping, and AI. A QR code is the most common way to trigger several of these, most practically mobile shopping, which requires the destination to be a complete, mobile-optimised experience through to purchase completion, not an informational page or a generic homepage. A QR code that opens a plain "learn more" page, with no purchase path, doesn't meet that bar even though it's a working, trackable code.

The mechanics, sourced from the live guidebook:

• Eligible mail: First-Class Mail letters, cards and flats, USPS Marketing Mail, and Nonprofit USPS Marketing Mail letters and flats. Periodicals, Bound Printed Matter and Media Mail don't qualify.
• Registration: the mail owner's CRID must be registered for the promotion through Incentive Programs on the Business Customer Gateway. There's no auto-enrolment.
• Pre-approval: each mailpiece design needs separate approval through USPS's My Products Portal, which the guidebook estimates at around four business days, so this has to happen well before the print date.
• Enrolment window: a mailer picks any six consecutive months within 2026, with the CRID enrolled at least two hours before the qualifying mailing goes out. Registration itself is open from mid-November 2025 through the end of 2026.
• Placement: for mobile shopping, the trigger can't sit on a reply card, order form, detached label or perforated section, and must measure at least half an inch square, with directional copy printed at least as prominently as the primary marketing message.

A mailer running real volume with a genuine mobile-shopping destination can claim 5 to 7% off postage, which adds up at scale, but the planning has to start with the destination page, not the QR code: build the mobile-optimised, purchase-capable experience first, register the CRID and get pre-approval with real lead time, then treat the code as the trigger that unlocks a discount you'd otherwise leave on the table.

Sizing the code so it survives the post

The promotion's own minimum, half an inch square for the mobile shopping trigger, is a compliance floor, not a design target: the smallest size USPS will accept for the discount, not the smallest size that scans reliably on a folded, handled piece of mail. Our general guide to QR code size for print sets out the quiet zone, contrast and minimum-size rules that apply regardless of the mail piece, and those are the numbers to design against; the USPS figure is a lower bound to clear, not a target to hit exactly.

The other mail-specific detail is proximity: keep the code clear of the address block, the postal indicia, and the barcode clear zone USPS uses for sorting. A code that sits too close to those elements risks a processing problem at the post office and a scan failure at once, since the quiet zone a scanner needs to lock onto the pattern can be interrupted by nearby ink just as easily as by low contrast.

What direct mail with a QR code actually converts at

This is the section where most guides on this topic quietly make things up, so it's worth being direct. A widely repeated claim, that roughly 44% of consumers have scanned a QR code from direct mail, plus a companion claim that personalised URLs lift response by around 50%, both trace back to blog posts citing each other in a circle with no working link to an original study underneath either number. Neither is exactly a fabrication, they're just untraceable, and repeating them as fact is what makes a post read like every other one on this topic.

What does have a real, citable source is the ANA's 2023 Response Rate Report, which measured a 15.6% response rate for house file (existing customer) direct mail and 10.8% for prospect file (cold list) mail, from a self-reported survey of 250 marketers. The report is upfront that its sample is small and self-selected, and explicitly cautions against treating its figures as an industry benchmark, which is itself useful context: even the best public number on this topic comes with a footnote, and a QR-specific scan-rate figure with no footnote at all deserves more suspicion, not less.

The honest position for an agency to take with a client is to stop looking for an industry average and build the client's own baseline instead. Track response and scan rate against that client's past mail campaigns, segmented the same way each time, and use that trend as the benchmark rather than a number lifted from someone else's blog post. Our guide to what counts as a good QR code scan rate goes into why a single blanket percentage is the wrong question for any QR channel, mail included.

Running direct mail QR campaigns across multiple clients

Everything above scales differently once an agency is running personalised mail for more than one client, mostly because the failure mode changes from "did this campaign work" to "did we mix up two clients' data." A naming convention that includes the client alongside the campaign, wave and segment fields matters even more here, since it's the only thing standing between a clean report and an embarrassing one where client A's scan data ends up in client B's deck.

Structuring work by a dedicated workspace per client enforces that separation at the platform level rather than relying on everyone remembering the convention correctly on a busy print deadline. For a direct mail specialism, set the workspace and naming structure up once, as a template, before the first campaign for a new client rather than after the second time a report needs manual untangling.

Frequently asked questions

What is a personalised QR code in direct mail?

A code generated for an individual recipient or a segment of a mail run, rather than one shared code on every piece. It's produced during variable data printing: a unique tracked URL from a data source is rendered as a fresh QR symbol for each piece, in the same pass that personalises the name and address.

What's the difference between a QR code and a PURL?

Both are entry points to the same tracked landing page. A PURL is a short, typed address, conventionally under about 50 characters. A QR code is scanned with a camera, so it carries no visible length limit. Many campaigns print both, the PURL as a fallback for anyone who'd rather type than scan.

How do you generate a unique QR code for every recipient in a mail run?

Programmatically: export the list as a CSV, generate a unique tracked URL per row with a bulk QR code generator, and hand that file to the print vendor as the data source for a variable data printing merge.

What percentage of people scan a QR code from direct mail?

There's no reliable, sourced answer as a single percentage. The commonly cited 44% figure traces back to blog posts citing each other with no verifiable study behind it. The best public direct mail benchmark, the ANA's 2023 Response Rate Report, measured overall response rates (not QR scans specifically) of 15.6% for house file mail and 10.8% for prospect file mail, from a small, self-reported sample the report itself cautions against treating as a standard.

Can you get a USPS discount for using QR codes on direct mail?

Only indirectly. The 2026 Integrated Technology Promotion offers a 5% postage discount, up to 7% with add-ons, but it rewards using one of eight named interactive technologies as a mailpiece trigger, not simply printing a QR code. Most agencies would qualify under the mobile shopping category, which requires the destination to be a complete, mobile-optimised buying experience through to purchase completion, plus CRID registration and mailpiece pre-approval ahead of the print date.

What size should a QR code be on a postcard or mailer?

The USPS promotion's own minimum, for mobile shopping, is half an inch square, but that's a compliance floor, not a reliable scanning size. For a piece that will be folded or handled, follow general QR print-sizing rules for quiet zone, contrast and minimum dimension instead.

Can you change a QR code's destination after the mailer has already been sent?

Yes, if the code is dynamic, encoding a managed short URL rather than a static code with the destination baked into the symbol. Its destination lives in a database the scan looks up, so it can be updated, redirected or paused after delivery, without reprinting anything.

Should each segment get its own QR code, or one code with different UTM parameters?

For most campaigns, a unique code per segment or mail wave, tagged consistently for client, campaign, wave and segment, gives clean reporting without the overhead of a code per individual recipient. Reserve per-recipient codes for when the landing page itself uses the recipient's identity, such as pre-filling a form.

What data does a QR code scan on a mailer actually capture?

Which code was used, the device type, an approximate location from IP address (city level, not a GPS pinpoint), and a timestamp. It doesn't identify the individual recipient unless the landing page itself captures that, typically because the code carries an identifier that pre-fills a form.

Do you need both a QR code and a PURL, or does one replace the other?

Neither replaces the other, they're two entry points to the same tracked destination. A recipient can scan with a phone camera or type the shorter PURL by hand, and the choice doesn't affect what gets tracked once they land on the page.

Does putting a logo in a direct mail QR code hurt its reliability?

It can, if the error correction level and logo coverage aren't matched correctly. It's solvable rather than a reason to avoid branding the code; our guide to adding a logo without breaking the scan covers the limits that keep a branded code reliable, which matters more on mail since a printed piece can't be corrected after it's out the door.

The short version

A QR code on a direct mail piece earns its place for two reasons: it lets an agency personalise the response path per recipient or segment through ordinary variable data printing, and it turns a one-way mailer into something with real, reportable scan data attached. Generate the codes in bulk from a tracked URL per row, tag them with a naming convention that includes the client, wave and segment before the first piece is printed, and build the ROI report on that data rather than on borrowed industry averages that don't survive a source check. If postage savings matter, treat the USPS 2026 promotion as a destination-first requirement: build a genuinely mobile-optimised landing page, register the CRID and get pre-approval with real lead time, rather than assuming any QR code on the piece will qualify. Set the naming and workspace structure up once, before the first campaign for a client, and the report at the end writes itself instead of getting rebuilt from scratch.