Retail Window Display: Built for Three Seconds

Whoever built your window spent forty minutes standing two feet from the glass, inside the shop, looking out. They stepped back, squared up the middle, adjusted a fold. It looked right.

Almost nobody sees a retail window display that way. Your customers get it from eight to twelve feet out, from the side, at about four feet a second, with maybe a quarter of their attention — the rest is on a phone, the curb, and the person coming the other way. A window arranged by a stationary viewer at arm's length is built for an audience of one, and that one already works there.

How do you know if a retail window display is working?

Count capture rate: people who come in, divided by people who walk past. Pick a twenty-minute slot, tally passers-by on your side of the street and entries through your door, then repeat that same slot on the same weekday for three weeks to learn your normal range. Change one thing in the window and count again. A shift smaller than your week-to-week swing means nothing.

What to count, and when to believe it

The trap is believing the first count. Rain, a delivery van parked across your glass, a half day at the school up the road: any of those move the number more than a window change will. Write the weather beside each week. If your three baselines come out at 4%, 6% and 5%, a fourth week at 6.5% has told you nothing.

Hand-counting entries gets abandoned around week two. Transaction counts are the cheap substitute: they miss browsers, but they are recorded already and consistent week to week, so a same-slot comparison holds. Whatever till you ring on will have that in its sales history — and if it reports per location, as VoVi does, a two-shop owner can put the same Tuesday slot on two streets side by side.

The leading edge does the work

Walk out of your own shop, go forty feet up the pavement, turn around and come back at your normal pace. Then do it from the other direction. What arrives first is the vertical edge of the window, foreshortened to maybe eighteen inches of visible depth, the rest still hidden behind its own frame.

That leading edge is where the best thing goes. Most shops put the hero item dead centre, the last part of the window to come into view, landing when the passer-by is already level with the door. Centre is the position of honour if you are standing still on the opposite pavement. Your customers are not on the opposite pavement.

You have two leading edges, because people come from both ends, and they are rarely equal. A cart is the extreme version of this: a kiosk has no front door at all, only approach angles. If a station sits one way and a parking lot the other, stand at the door for ten minutes at your busiest hour and tally direction of travel. Load the busier edge, and put the second-best thing on the other.

Design for four feet a second

Adults on flat ground walk at roughly 3.6 to 5.4 feet per second. A twelve-foot window at that pace sits in front of someone for about three seconds, arithmetic rather than a rule of thumb: two and a quarter at a commuter clip, three and a third at a Sunday one, most of it peripheral. That is room for one idea per window: one thing a person could name after they have gone past. Add eleven supporting items and they dilute it to none.

Then lettering. The US Access Board's summary of variable-message-sign legibility research puts the index for average younger and older observers at around 35 feet of viewing distance per inch of character height, and about 17 ft/in for the index that covers roughly 85 per cent of older observers. A legibility index marks the threshold: the point at which characters can just be read. Take the 17 ft/in figure and from twenty-five feet up the pavement, where reading starts, that is about an inch and a half. Treat it as a floor and roughly triple it: four inches for anything that has to survive a glance in motion, six for the one word you want someone to leave with.

Contrast does more than size. A dark coat against a dark shop interior has no silhouette from ten feet out; against a plain pale board it reads as a shape before anyone works out what it is. And keep goods up off the window floor: a walking adult's sightline lands at chest height, so anything below two feet is shown to knees. Painted plywood boxes make a plinth for the price of the paint.

Put the price in the glass, and make it big

An unpriced window reads as expensive. Someone who cannot see a number supplies their own, and it is higher than yours, because that is the safe guess about a shop you do not know. Put prices in, big enough to read from the pavement, on what you want to sell.

Two arguments come back at me. The first is that pricing cheapens the goods, and above a certain level that is true: if your average sale runs to four figures, a price card costs you more than it earns. The second is that competitors will see your prices. They can also walk in.

One honest anchor works hardest: a real item at a real number, priced in the middle of your range, because a cheap anchor brings you people who want the cheap thing. If you have an offer, put the number in the glass and leave the word “sale” out of it. By October, “sale” is wallpaper on most shopping streets.

In daylight your glass is a mirror

On a bright afternoon at street level, the passer-by is looking at the building opposite. Outdoor illuminance runs to 10,000–25,000 lux in full daylight and around 1,000 lux on an overcast day, while office lighting sits between 320 and 500 lux. Your display is losing that fight.

Fixes, in the order that pays. Shade the glass with an awning, a canopy or a deep reveal: less daylight on the pane, less reflection to see through. Then aim your light at the goods. Board the back of the window in something solid and pale. Last, stand on the pavement mid-afternoon, find the patch of glass holding the building opposite, and move your hero item out of it.

Most of that is aiming, not buying. The same holds inside the shop: re-aiming the fixtures you already own beats adding more, and the window is where your light has to beat the sun.

Rotate on your repeat-visit interval

You change the window monthly because the month ended, not because anyone stopped seeing it. The right interval comes from how often the same person goes past you. A shop on a commuter route gets the same faces twice a day, ten times a week, and by day four your regulars are walking past without a glance. A destination shop in a market town sees the same customer every six weeks and can hold a window for a month.

Your sales history has the median gap between a customer's first and second visit. Change the window at roughly half that gap: regulars back every three weeks means something moves every ten days. Get that right and people come in at speed, and then the first few feet inside are the next problem: the space just inside the door is a decompression zone. Swapping the leading-edge item and the price card makes a familiar eye stop and re-read, and takes fifteen minutes before opening. Run that fifteen-minute swap forty times a year and it will out-earn the full seasonal rebuild most shops save three weekends for.

Common questions

What is a good capture rate for a small shop?

There is no honest benchmark to hand you. A shop on a transit route and a boutique people visit deliberately will produce numbers an order of magnitude apart, and neither is doing anything wrong. Compare your Tuesday to your own Tuesday.

Should the item in the window be findable inside the shop?

Yes, and within a few seconds of someone walking in. A customer who came in for the thing in the glass and cannot see it in the first ten feet usually leaves without asking. Put it, or the same line, in direct sight of the door, and pull it the day it sells out.

What do I do if my window is shared or covered by a shutter?

With a shared window, stop fighting for the whole pane and take the leading edge on your side, since your share of the traffic meets it first anyway. With a shutter, treat it as your night-time window: paint or wrap it with your name, hours and one product image. That beats bare steel overnight and costs less than glass graphics.