QR Codes and Phone Numbers on Transit Ads: Tracking Response Properly

The most common structural failure we see in auto campaigns is not creative and not fleet size. It is a well-planned fleet with the main office number on the hood — the same number printed on the website, the shopfront, and every digital ad. Every call it produces is attributed to "phone," which means it is attributed to nothing. This article covers how to set the tracking up properly, before the first hood is installed.

Why Does a Transit Ad Need a Dedicated Response Mechanism at All?

Because without one, the campaign gets zero credit for the business it creates. A response mechanism is anything that exists only on the autos: a virtual phone number, a campaign QR code, a WhatsApp keyword, a promo code. Its job is not to increase response — its job is to make response countable. Those are different goals and they are sometimes in tension: a memorable vanity number may pull better than a virtual one, and a QR code takes panel space away from the message that actually drives the call.

The cost of getting this wrong is not the tracking budget. It is that a campaign generating 80–150 leads a month per 100 autos — the range we see for local-intent businesses in our analysis of what auto advertising actually delivers — looks like a campaign generating zero, and gets cancelled at month two.

QR Code or Phone Number: Which Works Better on an Auto Rickshaw?

The phone number, in most cases. The reasoning is physical rather than a matter of preference. A phone number is readable at the distance auto creative is designed for — our 30-metre rule for auto ad design exists because that is the realistic viewing distance in Indian traffic. A ten-digit number in high-contrast type is legible there, and more importantly it is retainable: people remember it for the thirty seconds it takes to reach a pocket, and they can act on it later.

A QR code cannot do any of that. It requires the viewer to be within a couple of metres, to have their phone out and unlocked, and to hold reasonably still. In moving traffic that combination almost never occurs. Where it does occur is specific and predictable:

• At traffic signals, from the vehicle immediately behind the auto.
• At auto stands and parking ranks, where the vehicle is stationary and people approach on foot.
• From inside the auto — the passenger is a captive audience for interior and rear-panel creative for the length of a ride.

How Large Does a QR Code Have to Be on a Moving Vehicle?

Use the signage rule of thumb: maximum scan distance is about ten times the code's width. A 10 cm code scans from roughly 1 metre, a 20 cm code from 2 metres, a 30 cm code from 3. That ratio, not aesthetics, should decide whether a QR code goes on the panel at all.

Format choice and rates are set out in full on our hood branding and 24×6" strip sticker pages. Four production details then decide whether the code scans in the field, and all four are free to get right:

• Keep the quiet zone. The QR specification requires a clear margin of four modules on every side. Designers routinely crop it to save space, and the code stops scanning.
• Dark code on a light background. Inverted codes — light on dark — fail on a meaningful share of phone cameras. If your brand palette is dark, put the code in a white box.
• Stay off folds and curves. A rexin hood has panel edges and creases. A code that wraps a corner is a code that will not resolve.
• Test the printed proof outdoors, at the real size, in daylight, from 2 metres, on both an Android and an iPhone — not on-screen at 400% zoom.

What Should the QR Code Actually Point To?

A dynamic QR code resolving to a UTM-tagged landing page you control. Three parts, each of which matters.

Dynamic, not static. A static code encodes the destination URL permanently — if the landing page moves or the offer changes, every printed hood is dead, and re-printing 100 hoods is a real cost. A dynamic code encodes a short redirect you control, so the destination can change mid-campaign. It also counts scans, which a static code cannot do at all.

Your own short domain, not a free shortener. Free link shorteners get retired, rate-limited, and occasionally flagged. A redirect on your own domain will outlive the campaign.

A UTM-tagged, mobile-first destination. Something like utm_source=auto&utm_medium=ooh&utm_campaign=city-month&utm_content=zone. The utm_content slot is the important one: give each zone its own value and scans arrive pre-segmented by neighbourhood — exactly the granularity that makes the month-3 reallocation decision in our zone-targeting guide possible. Land on a page that loads fast on a mid-range Android over 4G and asks for one thing.

How Do You Set Up Call Tracking Properly?

Virtual number services are inexpensive and the setup takes an afternoon. What matters is the discipline around them:

1. 1One number per campaign at minimum; one per zone if the fleet spans more than two zones. Zone-level numbers are the cheapest attribution upgrade available, and they are the difference between knowing the campaign worked and knowing which neighbourhood to move autos into.
2. 2Print it only on the autos. The moment it appears on a website footer or a Google Ads extension, it stops being evidence.
3. 3Route it to a real, staffed line with the same answering script as your main number. A tracked number that rings out converts nothing.
4. 4Log every call with its zone tag into the CRM — the number is only useful if the source survives into your pipeline data.
5. 5Keep it live for at least 60 days after the campaign ends. Auto advertising builds recall over months; calls continue arriving after the hoods come off, and they belong to the campaign.

How Do You Measure the Response Tracking Never Sees?

This is where most transit campaigns are undervalued. Attributable response — scans plus tracked calls — is a floor, not a total. Someone sees your hood four times over three weeks, searches your brand name on the fifth, and converts through organic search. Nothing in your tracking setup will ever credit the auto. Two methods close that gap without much effort:

A branded-search and direct-traffic baseline. For the four weeks before launch, record brand-name search volume and direct sessions, filtered to the campaign cities in Search Console and analytics. Then watch the delta from week 3 onwards, when frequency starts compounding. It is not a clean experiment, but a fleet-sized change in branded search in one city and not in others is a strong signal.

A matched-market comparison. If you run two comparable cities or two comparable zones, fleet one and hold the other. The difference in enquiries, walk-ins, and branded search between them is the closest thing to a controlled read that offline media allows. Zone-level numbers make this nearly free to run. For fleets of 100+ autos, a 200-respondent aided-recall survey in the target zones — once before launch, once at week 10–12 — measures the layer that neither scans nor calls capture at all.

What Does a Complete Tracking Setup Look Like?

Everything below should exist before the first hood is installed. None of it can be retrofitted once the fleet is on the road.

• One dedicated virtual number per zone, printed only on the autos, routed to a staffed line, logged into the CRM with its zone tag.
• One dynamic QR code per zone — hood and 24×18" formats only — at 20 cm or larger, dark-on-light, quiet zone intact, off folds, proofed outdoors on two phones.
• A single mobile-first landing page, UTM-tagged with the zone in utm_content.
• A mandatory source field on every enquiry form and walk-in log.
• Four weeks of pre-launch branded-search and direct-traffic baseline, recorded per city.
• A calendar entry at month 3 to review cost per lead by zone and reallocate autos.

Budget the tracking against the media. A 100-auto hood campaign is ₹58,000 a month in media at 2026 rates, plus one-time printing and metro union charges as set out in the 2026 cost guide. Virtual numbers and a dynamic QR service cost a low four-figure sum a month against that. There is no version of this arithmetic where skipping tracking is the rational choice.

What Are the Most Common Tracking Mistakes?

• Three response routes on one panel — number, QR, website, and a social handle. Each one splits attention and none gets used properly. Pick one primary, at most one secondary.
• The main office number on the hood. Free to avoid, and it invalidates the entire measurement.
• A QR code printed at 8 cm because that is what was left after the logo. An unscannable code is worse than no code: it occupies space the message needed.
• A static QR to a URL that later changes, stranding every printed panel.
• Judging the campaign on scans alone. Scans are the smallest and most physically constrained slice of response on a moving medium.
• Setting up tracking in week two. The pre-launch baseline is unrecoverable once the fleet is live.
• Killing the tracked number the day the campaign ends, and losing the tail of late responders that recall-driven media reliably produces.

The number on the panel decides whether the campaign gets a second year. We have watched fleets that were clearly working get cancelled at month two because nobody could prove where the calls came from.