The Name Nobody Submitted
A naming contest, 200,000 fake votes, and the word nobody thought to submit.
On day four of my naming contest, first place had 200,000 votes; second had four for the same name, misspelled.
I’d built BrandForge for four B2B wholesale platforms merging under one holding company. Their users needed different blends of custom publishing, order complexity (furniture versus gift), and reports. They were tired of new names and redesigns, with plenty of expectations those changes hadn’t met. We couldn’t just pick a day and announce one name for all four; it had to help them move toward the new company without sounding like another warning their software was about to change.
I went to the employees for ideas because they knew the history and heard the customer complaints. They could submit names, explain them, vote, and win cash. Four coworkers found their way around my lax security and put a slightly different spin on the voting part.
More cash mostly bought me more spam. Meanwhile, most people visited once, put in their ideas, and went back to work. Familiar names filled the top slots, where they could keep getting votes while the rest barely got seen.
I gave the joke names a room of their own, grouped similar entries, and put new names first. The spammers started producing useful ideas. By day nine, though, I’d still managed to end up with a dead contest.
So I made it rain names.
In Idea Storm, people threw names left or right and voted on the ones they kept. New features gave them a reason to return, while Test of Mettle used pop culture and sarcasm to push their thinking past the same old naming habits. Over twelve days, about 55 people played for almost 1,000 minutes. All of them submitted more than one name.
We paid the winner $500. Their name and most of the top 15 were nowhere near what we chose. The word that mattered was in the notes under the entries, where staff kept using it for the reps, brands, and retailers we served. Nobody had thought to put it in a name.
I’d spent three weeks getting people to submit answers, and the useful bit was what they wrote to explain them. The employees knew the problem better than their proposed names could express.
Between the hacked votes and my own steering, the leaderboard couldn’t tell me what people would choose left alone. It had given me something better to work with: their reasons. The trick was reading below the part everyone was voting on.
That’s the short version.
The remaining 8 minutes cover the naming constraints, the fixes that failed, and how games got people back into the contest.
One of the four platforms made the naming job harder. It had pioneered the gift market and gone by several names over the years, so you couldn’t talk to a customer about functionality or UX without hearing why the new platform couldn’t do what the old one did. A legacy name like that becomes a standing comparison, and it put a hard gate on what the new name could be.
Merging two words was the tried-and-true move, and two of the four companies had already made it, so those words would have to be very different from theirs. There was also a strong pull toward a cool, modern, single-word name like the category’s billion-dollar competitor. It’s a great name, and it took a fortune to build a brand around it. We needed something our next-generation products could eventually define, while giving four customer bases enough comfort that they’d move toward it rather than flee.
I’m leaving the name out of this account. You won’t miss it much, because nearly everything worth telling happened before the room came around to it.
Before all the glitz and glam of the final identity package and the stakeholder presentations, I approached this the way I approach most product design problems: who are the experts, and what do they think it should be called?
For once, the experts weren’t customers; they were employees. About 30% of the people across the four brands had worked for some iteration of them for half their careers or more. That can leave experts naive about the ecosystem outside their own platforms, but many had left during past mergers and shakeups and come back. Most were passionate about the work, and a lot of them sat less than an arm’s length from real customer feedback and complaints.
So I built an app called BrandForge and promoted it through email and Teams. Employees could submit a name, add a tagline or the inspiration behind it, vote on everyone else’s, and watch the leaderboard. The winning name came with a cash prize. To the naked eye, it looked and acted like a naming contest. Its real purpose was to generate data for a second tool, one nobody knew about. The experiment hinged on participation, but more than that, it hinged on getting a wide range of names into the forge.
I set a deliberately low goal of at least five names and ten votes a day across about 200 participants, and built a private area with analytics and management tools to capture participation and grade the data as it came in. For three weeks, I kept things moving with a fairly sarcastic, pun-laden voice. The first few days were busy. By the fifth night, we were under goal.
By the end of the first week, I had three problems wearing the same disguise: a disappointing leaderboard.
The first was clustering. A lot of names shared the same constructions, especially names starting with “Market” or “Rep,” and they were collecting enough votes to take over the top 15. The more original names never got the face time they needed to be voted on.
I’ve watched the same thing happen to products for twenty years. People spend years running the workflows a product gives them, and eventually both they and the designer struggle to see past the prison of what exists. These employees were naming the company out of the parts they already knew. They were the right experts on the problem. That didn’t make them reliable prophets of the solution.
The second was spam. Four employees had gotten around my admittedly lax security using their browsers’ developer consoles, boosting, hijacking, and isolating specific names until the contest looked like an election held exclusively inside a botnet. Hence day four.
The third was that nobody came back. People had their ideas, submitted them, maybe voted on a few others, and went back to work. I had given them somewhere to put the two names they’d thought of on day one, and very little reason to visit on day two.
So I sent an even more enticing email, with a couple of bonus cash incentives for participation, and mostly succeeded in making the spammers work harder and longer. Leadership was frustrated with the spam and the thin participation; the expectation had been that staff would keep it professional, fun or not.
Early in my career, I watched users ignore a CTA while clicking a seemingly empty void next to it. You don’t fight an inclination like that. You move the button into the void and work from there, and you keep the voice, because it can still draw the eye to the button.
The spammers were my empty void. Technically, they were participating at an exponentially higher rate than anyone else. With some digging, I found out the four of them were fairly vocal members of the largest legacy brand, people who could speak to a big group. If I could deal with their spam in ways that didn’t read as punishment, I might get more out of them, both their own ideas and their ability to turn one-time visitors into return users.
I started with what the leaderboard showed people.
I created a “wacky” category that only I could assign. It was off by default and appeared as a single wacky emoticon in the filter bar. Anyone could toggle it on, and it stayed on for anyone who did. The spammers’ names didn’t get deleted; they got a room of their own, one tap away.
I manually grouped names that were similar or built in very similar ways, so each group appeared as one listing you could expand to see the names beneath it. Votes rolled up into the group’s total. With about 15 slots visible on most devices, a couple of groups could absorb a large share of the increasingly similar suggestions and leave room for something else. The toolbar let you turn grouping off. No one clicked it.
And I changed the default sort to all names, newest first. Toggling to Top sorted by votes instead. New names got seen before they got buried, and they gave visitors something to react to.
The spammers became useful, producing several new top names. Grouping let contributors see the well-worn paths quickly and stop pouring creativity into them, which opened up directions like “Omni.” What none of it did was bring people back. The changes made a visit better; they didn’t produce a second one. Despite more messages showcasing interesting names alongside a sarcastic take on whichever spammed name was leading, by day nine participation was dead.
My first few ideas were shit at boosting participation.
I already had a trivia concept in development, but I knew trivia wasn’t enough to lead a comeback on its own. So I built a game called Idea Storm. Instead of rain or hail or snow, names people had suggested fell from the sky, and you knocked or threw them to the left or right edge of the screen: right for names you liked, left for names that were silly or that you didn’t. At the end of an average four-minute storm, the game-over screen showed the names you’d kept, and you could vote on them right there.
By the end of the first week there were over 40 valid submissions, ungrouped and not goofy, and nearly all of them had one vote or none, because a leaderboard only shows you the top. Idea Storm put every name in front of people and gave them something to do with each one.
I announced new features in daily drops, which gave people a reason to come back each day, and in the last week I added a special artifact that, thrown to either edge, opened a minigame where you essentially made up nonsense names. By day eleven, participation was high. Over the twelve days the game ran, it logged almost 1,000 minutes of play across about 55 people. All of them submitted more than one name, and 14 submitted three or more.
The trivia, Test of Mettle, ran alongside it, and it’s where most of my steering happened. I wanted variety and original concepts in the data, which meant pushing people to think differently, and sarcastically. Each question blended pop culture and movies with a behavior I wanted to change.
Scrap Metal #1 steered names far away from anything rep-centric:
The “Rep” Obsession
We need a Swiss Army Knife. What happens if we forge a knife that only says “Reps.com”?
A) We accidentally crash the Gibson. B) We alienate Retailers & Brands. C) We become Scruffy Nerf Herders.
B. NAILED IT! Rep-centric names are brittle. They break when you try to sell to a Retailer or a Brand. A Swiss Army Knife must be useful to everyone. Labeling it “Rep-Only” dulls the blade for 66% of our market.
Scrap Metal #2 steered around “Omni” names rather than away from them:
The “Omni” Illusion
Voices from the forge called out for “Omni” names... but why is “Omni” a weak alloy?
A) It transforms us into the T-Virus. B) It blends into the “Beige” of Tech. C) It creates a RISC Architecture bottleneck.
B. SOLID HIT! “Omni” is the most commoditized prefix in B2B. It’s not a brand; it’s a description. Our audit found OmniFlow and OmniCore are already taken. Using “Omni” is like forging a sword out of tin — it bends under pressure.
The game and the trivia let people have fun while exposing them to names most of them had never seen, and they let me steer their creativity with just enough of a hand to stay invisible, interesting, and useful. I never told anyone what to submit. I just made some directions more fun to think about than others.
That also means I had no business treating the leaderboard as a clean measurement of what employees independently preferred. I had changed what people saw, what they could do, and which constraints they spent time thinking about, on top of the votes four people had already made a mess of. The participation numbers told me the experience had come back to life; choosing a company name still required reading what people had actually contributed.
When the contest wrapped, we paid the winner $500. By then, the winners cared far more about joining the Brand Council, where they could keep promoting their names, than about the money. But the winning name, and honestly most of the top 15, weren’t remotely close to the name we eventually chose.
Every submission could include a tagline, the inspiration behind it, or both. The first line of either appeared under the name, and almost everyone filled one in. Toward the end, people went back to annotate earlier submissions too, explaining what a name meant to them and what they thought the company ought to hold together.
In those notes, people kept describing everyone we serve (reps, brands, and retailers) with one collective word. Nobody had used it in a name anywhere.
The submissions were people’s solutions. The notes were their explanations of the problem, and the explanations turned out to be worth more. I’ve run into that in product work more times than I can count: users describe problems far better than they prescribe solutions. You can count votes all day and never find a word nobody submitted. You find it by reading.
The highest number on the page belonged to a joke. The most useful word wasn’t on the leaderboard at all.