Direct Mail Attribution for B2B: QR Codes, PURLs, Matchbacks, and Holdouts

Introduction: if you can’t measure it, you can’t scale it

Direct mail has a reputation for being “hard to attribute.”

That’s only true when measurement is an afterthought.

In B2B, you can measure direct mail well enough to make confident decisions by combining:

• direct response tracking (QR codes, trackable URLs, PURLs)
• CRM matchbacks (pipeline outcomes by cohort)
• incrementality tests (holdouts and lift)

This guide is a B2B playbook you can implement without building a data warehouse first.

Step 1: build solid foundations (the boring part that matters)

Attribution falls apart when your data is fragmented.

Minimum foundation:

• a CRM that is the source of truth for pipeline
• a way to log mail touches (date, segment, template)
• a landing page / redirect system for QR and URL tracking

If you want a quick overview of tracking mechanisms, see: How to track direct mail marketing campaigns.

Step 2: decide what you’re trying to measure (B2B outcomes)

Pick 1–2 primary outcomes per motion.

Outbound / ABM:

• replies
• meetings booked
• opportunities created
• stage velocity

Customer programs:

• renewal saves
• expansion rate
• exec sponsor engagement

Define your event taxonomy (so reporting is consistent)

At minimum, standardize:

• mail_campaign_id (one per motion/experiment)
• mail_template_id (creative/copy version)
• mail_segment (tier/persona/industry)
• mail_touch_date (when it was sent)

Optional but high value:

• trigger_reason (demo completed, opp stalled, renewal window, event follow-up)
• holdout_group (test/control)

If those events start inside HubSpot or Salesforce, pair this taxonomy with the CRM-triggered direct mail workflow guide so the send trigger, campaign ID, and follow-up task use the same fields.

If you track events in analytics as well, use a simple schema:

{
  "event": "direct_mail_sent",
  "mail_campaign_id": "abm_tier1_q1_2026",
  "mail_template_id": "note_v3",
  "mail_segment": "tier1_cro",
  "trigger_reason": "post_demo",
  "contact_id": "crm_123",
  "account_id": "crm_acc_987",
  "sent_at": "2026-04-06"
}

Step 3: direct response tracking (QR codes and trackable URLs)

Trackable URLs (baseline)

Use a unique URL per campaign:

• include UTMs
• route through a redirect you control (so you can update destinations later)

QR codes (reduce friction)

QR codes improve response by:

• making the action easy on mobile
• creating a measurable “scan” event

Best practice:

• generate a unique QR per campaign (or per segment)
• use UTMs to label source/medium/campaign

Redirect links: keep control of destinations

Prefer QR/short links that go to a redirect you control, then forward to the final page.

Why:

• you can change landing pages later without reprinting
• you can log scans/visits consistently
• you can standardize UTMs even when teams forget

Example UTM convention:

• utm_source=direct_mail
• utm_medium=handwritten (or printed)
• utm_campaign={{mail_campaign_id}}
• utm_content={{mail_template_id}}

Step 4: PURLs (personalized URLs) when you need identity

PURLs help when you want:

• a clean link between offline touch and a known contact
• personalized landing experiences

Common approach:

• example.com/{{slug}} routes to a landing page
• page recognizes the contact/account and can pre-fill forms

Trade-offs:

• requires careful privacy handling
• more implementation work than generic QR

A safer alternative to PURLs: “identify later”

In many B2B funnels, you don’t need identity at scan time.

You can:

• attribute scans/visits to a campaign/segment via UTMs
• capture identity when they book a meeting / fill a form
• match back to the mailed cohort using CRM records

This avoids overly “personal” URLs while still keeping measurement useful.

Step 5: matchbacks (pipeline-first attribution)

Not everyone scans a code.

In B2B, many of the best outcomes show up as:

• replies
• meetings booked
• stage movement

Matchbacks attribute impact by cohort:

• segment A received mail on date X
• segment B did not
• compare outcomes over an attribution window (e.g., 14/30/60 days)

This works best when you log mail touches in the CRM and keep your segmentation clean.

Matchback methodology (simple, defensible)

1. Define a cohort: “accounts mailed between date A and date B”
2. Define a comparison cohort: similar accounts not mailed in that window
3. Choose an attribution window (14/30/60/90 days)
4. Compare outcomes:

• meeting rate
• opp creation rate
• stage progression speed
• win rate (if cycle allows)

Keep the cohorts stable during the window—don’t change targeting rules mid-test.

Step 6: attribution windows (choose them intentionally)

Attribution windows depend on:

• your sales cycle length
• how quickly your audience typically responds
• the “shelf life” of your mail piece

Common B2B starting points:

• 14 days: replies and meeting booking lift
• 30 days: opportunity creation and stage movement
• 60–90 days: win rate and revenue impact (for longer cycles)

Step 7: incrementality (holdouts and lift)

If you want true confidence, use a holdout:

• randomly exclude a subset of accounts from receiving mail
• compare outcomes vs the mailed cohort

Holdouts are the closest thing to “proof” you’ll get in channel measurement.

Practical tips:

• hold out at the account level (not contact) for ABM
• keep criteria identical across cohorts
• run long enough to capture your outcome window

How to run a holdout without overcomplicating it

• Randomize within a segment (e.g., Tier 1 CRO persona)
• Start with a small holdout (10–20%) so sales still gets coverage
• Lock the rules for the duration of the test
• Use pipeline outcomes as the truth (not just scan metrics)

When you report results, separate:

• direct response (scans/visits)
• downstream impact (meetings, opps, wins)

Step 8: reporting (what to show weekly and monthly)

Weekly:

• mail sent volume by segment
• response signals (scans, landing visits)
• early funnel outcomes (replies, meetings)

Monthly:

• opportunity creation rate by cohort
• stage velocity
• win rate (if cycle allows)
• renewal/expansion lift (for customer programs)

A simple scoreboard template (example)

Track by segment and cohort:

• mailed accounts
• meetings booked rate
• opp creation rate
• pipeline created per account
• lift vs holdout (%)

Step 9: common attribution traps

• judging mail only by QR scans
• changing targeting mid-test
• no frequency caps (results get noisy)
• no control group (hard to prove lift)
• mismatched windows (declaring winners too early)
• “multi-touch” attribution without clear rules (creates false certainty)

Conclusion

B2B direct mail attribution doesn’t need to be perfect.

It needs to be good enough to answer the only question that matters: Is this creating incremental pipeline and revenue?

Build foundations, track direct response, measure pipeline matchbacks, and use holdouts when you need proof.

For the campaign workflow, review Scribeless as a measurable handwritten direct mail service.

Want help setting up tracking and lift measurement for your direct mail program? Book a campaign consult