Measuring offline campaigns with links and codes

Offline is not unmeasurable, it is unattributed by default

The reason offline spend feels unaccountable is that nothing carries an identifier from the physical world into your analytics. A QR code or a spoken short link is exactly that identifier. Once each placement has its own, you have the same kind of data you have for digital channels, with different biases.

One identifier per placement, decided before production

Not per campaign, per placement: this billboard, that magazine, the insert in the March mailing. Retrofitting is impossible once the artwork is printed, so the link map has to exist before the design files are finalised. This is the step that most often gets skipped and most often makes the whole exercise pointless.

Accept that response rate understates impact

Someone who sees a billboard and searches for your brand two days later never touches your identifier. Scans and typed links measure immediate response, which is a subset of total effect and a badly biased one — formats that are easy to act on immediately will always look better. Use the identifier for comparing like with like, not for judging whether a channel works at all.

Look at the indirect signals alongside the direct ones

Branded search volume, direct traffic and store visits in the exposed region during the flight window all move if the campaign is working. None of them is clean, but together with scan data they give a defensible picture. Establish the baseline before the campaign starts, because you cannot reconstruct it afterwards.

Design the physical asset for response, not just for recall

A code too small for its distance, placed above eye level, or without a line explaining what it does, will underperform for reasons that have nothing to do with the media buy. When a placement shows near-zero scans, check the physical execution before you conclude the channel failed.

Keep the destination alive far longer than the campaign

Packaging sits in cupboards; magazines sit in waiting rooms. Plan the destination for a two-year horizon, and set an expiry behaviour that offers something useful rather than a 404. This costs nothing at setup and preserves a slow trickle of traffic for years.

Report it in the same table as digital, with a footnote

Put offline placements in the same channel comparison as email and social, with an explicit note that response-based measurement understates offline. Excluding it from the table entirely is how offline budgets get cut for the wrong reason.