Skip to content

Guides

How to make QR codes for real estate signs

A for-sale sign is printed to be reused, and the house behind it is sold in weeks. That mismatch is why the QR code on the panel is so often a rented redirect, repointed from one listing to the next. The fix is not a better redirect; it is admitting the sign carries two different things with two different lifespans and printing them separately.

The five-minute version

  1. Give each listing a permanent path on a domain you own: yourname.com/p/12-elm-street. While it is for sale it shows (or forwards to) the listing; after completion it says sold and shows similar homes and how to reach you.
  2. Put that code on the rider, the window card and the flyer-box sheet for that listing. It belongs to the house and is thrown away with the listing.
  3. Put a second, permanent code on the main panel that opens you: your listings page or your contact card. It is printed once and never changes.
  4. Size for a person on the pavement: 10 to 15 cm across, mounted at eye level, plain dark code on a light matte background.
  5. Generate the office’s riders in one pass from a spreadsheet, one row per listing. Scan the proof from the pavement, in sun and at dusk, before the sign company prints.

Why agents buy dynamic codes, and what they actually need

The argument for a dynamic code on a sign sounds airtight. The panel costs real money and lasts years; the listing lasts six weeks; so the code on the panel must be something you can repoint at the next house. The static-vs-dynamic guide concedes exactly this case, a printed code whose destination has to change, and it is fair to say a for-sale sign is the most common example of it in the wild.

Watch what the repointing does to the people who scanned it. A buyer photographs the sign on Saturday, scans the photo on Sunday, and lands on a bungalow across town because the house went under offer on Friday and the code now serves your next instruction. A saved link in a buyer’s notes opens a different property a month later. Every scan of a repointed code is a small lie, and the buyers most likely to notice are the ones who cared enough to scan twice.

What the sign is really carrying is two things. The listing, which changes every few weeks and should never be repointed, only retired. The agent, which does not change and should never be retired. One rented redirect is being asked to be both, and it is bad at both. Print them as two static codes: a per-listing code on the rider, a permanent agent code on the panel.

Once the listing has its own throwaway rider, the reason for the dynamic code disappears, and with it the subscription that can lapse while a hundred riders are planted around town. The expiry guide covers what that lapse looks like from the pavement: a sign that says scan me and a phone that says nothing.

What the listing code should open

  • A path on your own domain, not the portal. Encoding the Rightmove, Zillow or Realtor.com URL hands the lead to the portal and dies when the portal delists the property, which is the day you most want the page to still work. yourname.com/p/12-elm-street can forward to the portal while the house is live and do something useful afterwards. The packaging guide makes the long version of the own-the-domain argument.
  • The listing, not a form. The person scanning is standing outside the house; give them the price, photos and floor plan on the first screen and put “book a viewing” under it. A form that gates the photos behind a phone number is a bounce, and most of the people you gate are neighbours.
  • A graceful ending. When the sale completes the path should say sold, show two or three similar homes and your contact details. That page earns enquiries for months from a sign you have already taken down and a photo a buyer forgot about. The coupon guide makes the same point about offers that end: a sentence beats a dead redirect.
  • If you have no website of your own, the portal link is the honest fallback, on the reasoning the restaurant-tables guide uses for ordering platforms: the portal is a landlord you have already signed with, so encoding its URL adds no new dependency. A QR vendor’s redirect on top of it would.

What the agent code should open

The permanent code on the main panel is the one most agencies never print, because the panel has the agent’s phone number on it already. A phone number is a barrier at 7 pm on a Sunday; a code that opens a page with every current listing from this office and a tap-to-call is not. Encode yourname.com/agent/jane or the office listings page. If you would rather the scan save you to the buyer’s phone, a vCard in the pattern is the business-card guide’s answer and it works on a sign too, with the caveat that the sign then carries your details for as long as it is in service.

The office is a batch

An office with forty active listings has forty rider codes to make, and each one is wanted on three or four pieces: the rider, the window card, the flyer-box sheet, the brochure back page. That is a spreadsheet job, and the spreadsheet you need is the one you already have.

  • One row per listing. A url column built by formula from the listing’s slug: ="https://yourname.com/p/"&B2. A label column with the short address, which becomes the caption under the code, so a rider in the van can be matched to a house without scanning it.
  • Generate the batch in one pass, static, and file the PNGs by address. A new instruction is one new row and one new code; the existing forty are untouched.
  • Because the codes are static, a reprint is pixel-identical. A rider lost to a storm is the same PNG sent to the sign company again, not a support ticket with a vendor.
  • Same code on every piece for a listing. One URL, one code, scanned from the rider or the brochure or the window, so the page can count where scans come from with a ?src=rider parameter per piece if you want that number, and you will have it without a vendor dashboard.

Sizes that scan from the pavement

The rule in the print-size guide is that a code should be about one tenth of the distance it is scanned from. A sign is the one place in this cluster where that distance is measured in metres, and the answer changes depending on who you are designing for.

  • The person on the pavement stands 1 to 1.5 m from the sign. 10 to 15 cm across is comfortable, 8 cm is the floor. A standard rider (15 × 60 cm) takes a 12 cm code with room for “Scan for price and photos” beside it. Design for this person; they are the one who walked up.
  • The car at the kerb is 2.5 to 3 m away and the code needs to be 25 to 30 cm, which is most of a rider and therefore belongs on the main panel. Nobody scans from a moving car; they stop, or they get out, and if they get out the pavement size applies.
  • Height. Mount the code between 1.2 and 1.6 m off the ground, at the eye level of the person standing in front of it, not at the top of a 2 m post where the phone has to tilt into the sky. Face it to the pavement, not the road.
  • Short payload. A short own-domain path keeps the code sparse and the modules big, which is what survives distance, sun and a phone with a scratched lens. The portal URL with its tracking parameters is often 80 characters and doubles the module count.

Weather, sun and the sign company

  • Matte over gloss. A laminated gloss panel in afternoon sun is a mirror, and the camera sees the sky. Ask for matte laminate or a matte vinyl, and put the code on the side of the sign that is in shade at the busiest viewing hour if you have the choice.
  • Dark on light, and plain. The agency red on white is fine; white on the agency red is a negative and many phones refuse it. Skip the logo in the middle and the rounded modules; spend the error-correction headroom on mud and fading. The six failure causes are in the not-scanning guide.
  • UV-stable ink. An inkjet rider fades in a summer, and it fades unevenly, which is worse for a scanner than a uniform grey. Riders are cheap; specify solvent or UV ink and replace one that has gone pale rather than hoping the code still has the contrast.
  • Send a vector or a large PNG. A 1024 px PNG at 12 cm is over 200 DPI, which is plenty; an SVG is better for the panel size. A 200 px thumbnail pulled from a listing page and blown up to 12 cm is a soft-edged square that scans at arm’s length and fails at one metre.

Before the sign goes in the ground

  • Scan the proof from 1.5 m, on a cheap phone, in full sun and again at dusk. Then scan it at an angle, because the pavement is not always square to the sign.
  • Open the page on mobile data with wifi off. A listing page that loads in 12 seconds over 4G is a page that is not read on the pavement.
  • Set one test listing to sold and scan its rider. You should see a sentence, similar homes and a way to reach you, not a portal’s “listing removed” page.
  • Match rider to house before the van leaves. The caption under the code is there so the driver can read it; a rider on the wrong house scans perfectly and lies.
  • Keep the CSV with the PNGs. The office’s signage is now a spreadsheet, and the next forty listings are forty rows.