The Shelf-Stopping Box Formula
Design the first, second and third thing the eye lands on — and aim for pick-it-up, not buy-it.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 7
- Confidence
- 92%
Elan identifies two places games fall apart, and the first is right off the bat: the box sucks or the name sucks. His caricature is the endless supply of games called something like the legendary folklore of Gorgonzel — non-descriptive, with cover art that conveys no gameplay and no compelling narrative about the experience you are about to have. His fix is a deliberate three-stop visual formula on the front whose goal is only to get the box picked up and turned over, and a back-of-box structure dominated by a three-pane how-to-play whose goal is purchase. Everything contestable — the coyote's hat, earring, wink, the saffron and lime green — was resolved by testing rather than taste, which is what Tim insisted on throughout.
Origin
Tim brings decades of the same practice: for The 4-Hour Workweek he split-tested roughly ten title and subtitle contenders using Google AdWords, bidding on related keywords and measuring clicks to under-construction pages, then physically placed different covers on a same-sized book in the new non-fiction section of Borders on University Avenue in Palo Alto and counted pickups with a bouncer's clicker. Elan's parallel discipline comes from a decade of fighting for attention in the games aisle.
Core principles
- 01Seventy percent of games sell in person, so the box has to do the selling with nobody standing next to it.
- 02A non-descriptive name and a picture that does not describe gameplay give a product zero chance of success, however good it is.
- 03Shelves now hold five hundred games rather than five, so a box that does not say pick me, pick me will simply not be picked.
- 04Overly complicated packaging with no dominant element reads as noise, and the eye moves to the next box.
- 05The front of the box has one job and it is not the sale: the target reaction is 'I'm going to pick this up and turn it over'.
- 06The front follows an ordered formula — you must know the first, second and third thing the eye will land on, in that order, exactly as in old-school print advertising.
- 07The back has more real estate and a different job: now the goal is purchase, and the reader will accept a short story if they can see it ends.
- 08The dominant element on the back should be how to play, compressed into three visual panes.
- 09Every contested design decision should be tested rather than argued, on real demographic panels.
- 10Vocabulary on the box is part of the design — words like sabotage are chosen as carefully as colours.
How to run it
- 1
Fix the name first
Elan treats a bad name as a terminal condition — those games have zero chance of success. Coyote's name took roughly 573 candidates and lots of testing; Hurry Up Chicken Butt took a thousand names, narrowed to ten Elan liked, then read aloud to his daughter and her friends.
Pro tip The kids did not even want to hear the rest of the names after Hurry Up Chicken Butt — an unambiguous signal beats a close call.
Watch out A name that only means something to people who already know the product is a non-descriptive name.
- 2
Establish one dominant element
Overly complicated packaging with too much going on and no dominant element leaves the browser asking what they are supposed to look at, and the answer is nothing — they look at the next box.
Pro tip Coyote's dominant element is the trickster coyote's head, cut off at the snout, with hat, earring and wink.
Watch out Every additional competing element reduces the probability of a stop.
- 3
Order the three eye stops deliberately
Elan says that when you look at the Coyote box, they know the first thing you will look at, the second, and the third — it is a formula, crafted very carefully. Tim identifies it as exactly what works in old-school print advertising: header image, headline, subheading, then text.
Pro tip Write the intended sequence down and verify it with fresh eyes before printing.
Watch out If you get the order wrong the eye does not know how to track and bounces around — and with better options next door, off they go.
- 4
Aim the front at pick-up, not purchase
Elan is explicit that the target reaction is not 'I'm going to buy that game'. It is 'I'm going to pick up this game and turn it over'. That is the entire job of the front.
Pro tip Once someone commits to picking it up, you have earned the right to a longer message.
Watch out Front panels that try to close the sale usually overload and lose the stop.
- 5
Use the back to teach and then close
The back has more real estate and a different task; you still know the first thing they will read but care less about the second. The dominant element is how to play, in three panes: put cards on the table and take turns performing them; play cards to help or sabotage other players; mess up and you're out, last player standing wins.
Pro tip A browser will read a short story as long as they can see the end of it and know it is not too long.
Watch out The word choice matters at this size — sabotage was very carefully chosen.
- 6
Test everything contestable
Use demographic panels rather than opinion. Tim used PickFu, Intellivy and Sticky Beak to serve variants — hat or no hat, earring or none, both eyes open, eyes looking off to the side — to fine-sliced audiences and got a strong statistical signal from thousands of votes.
Pro tip Tim's stance whenever Elan asked his opinion was: I do have an opinion, but I don't even want to tell you — let's test it.
Watch out Elan's praise for that stance is instructive: the winner was so clear it was not even close, which no internal debate would have surfaced.
- 7
Add an in-store advocate if you can
Coyote's back carries a line — not sure if you should buy this game, give us a few seconds to convince you — and a QR code linking to a fifteen-second video of Elan personally pitching it. It came after two rejected ideas: installing a telephone in every retail location, and hiring a person to stand in every store.
Pro tip It costs almost no real estate, so the downside is negligible — once again, test it, see what happens.
Watch out Elan is honest that he does not expect many people to scan it; the value is in the few who do getting a friend's recommendation.
In the wild
Tim describes the box as a slightly enlarged deck-of-cards box, with a top in an autumnal orange almost like a saffron Buddhist robe, and a lime green. He says it is impossible to overstate how much time everybody spent on it — he remembers photographing particular leaves at particular times of year with light coming through them to identify the hexadecimal and Pantone numbers for that exact green, then working through colour theory and the colour wheel.
→ A colour scheme developed with the retailers in mind, since retailers run annual colour themes and matching them makes a buyer's yes easier.
For his first book, Tim bid on subject-related keywords in Google AdWords so that roughly ten title and subtitle contenders were automatically split-tested against each other, with each variant pointing at a different URL that led only to an under-construction page — he did not care about conversion, only about sufficient interest to click. For the cover, he put different covers on a book of the same dimensions on the new non-fiction shelf at Borders on University Avenue in Palo Alto and, with a bouncer's counter, tracked pickups during peak hours over a couple of days.
→ Both the title and the cover of The 4-Hour Workweek chosen on measured behaviour rather than editorial taste.
Elan's reasoning was that if he went to the store with you, he could convince you to buy his favourite game because he would pitch the hell out of it — but he is not in the store with you. His first idea was asking retailers to install a telephone in every location so shoppers could call and talk to somebody; they said no. His second was hiring a person to stand in every store; they said sort of, but you pay for all of it, which was not going to work. Option three was a QR code linking to a fifteen-second video of him saying, listen, you're holding a game in your hands, let me try to convince you.
→ A first-ever experiment on Coyote's box that gives a browsing shopper the experience of a friend telling them to trust him.
Common mistakes
The non-descriptive fantasy name
Elan's example is the endless run of games called things like the legendary folklore of Gorgonzel — names he cannot remember and that tell a browser nothing. Such games will sell ten copies to the ten people who bought volumes one through six, and that is the ceiling.
Packaging with no dominant element
When there is so much going on that the shopper cannot tell what to look at first, the eye does not choose — it leaves. Tim and Elan agree the result is simply noise, and with better options right next door the browser is gone.
Trying to close on the front panel
Elan is explicit that 'I'm going to buy that game' is not the goal of the front. Designing the front as a sales page overloads the three-stop hierarchy and loses the pick-up that would have earned you the back of the box.
Settling design arguments by opinion
Elan had opinions about the hat, the earring and the wink; Tim's answer every time was let's test it. When the panel returned, the winner was not even close — a margin no internal debate would have produced, and one that only exists because nobody anchored the group first.
Is it for you?
Best for
Physical product teams selling through mass retail, and authors or creators choosing covers and titles.
Not ideal for
Purely digital or direct-to-consumer products where the purchase decision happens through search, reviews or referral rather than a shelf.
From the transcript
“picking the wrong name, picking a non-escriptive name, putting a picture on the box that does not describe gameplay at all”
“when you look at this box, we know the first thing you're going to look at and we know the second thing you're going to…”
“and your reaction should not be, i'm going to buy that game. that's not our goal. your reaction is, i'm going to pick up this…”
“and to your credit, your answer was always, let's test it”
From the episode
#821: My Two-Year Secret Project, COYOTE — The Strategies and Tactics for Building a Bestseller from Nothing with Elan Lee of Exploding Kittens