Filed Under Nothing
Buyers sort you into a category before they evaluate anything.
Dean Kamen unveiled the Segway on December 3, 2001, on Good Morning America, where Diane Sawyer said on air, “I’m tempted to say, that’s it?” The coverage that followed was about the machine: the gyroscopes, the rider who leans forward to go forward. What nobody could say was what it was. A vehicle, a scooter, a toy, a piece of medical equipment. The Segway sold for $4,000 to $5,500, which lands in the gap between two budgets: too much to buy the way you buy a toy, too little to be worth the procurement process that buys municipal vehicles. A buyer who can’t say what kind of thing you are has no way to decide whether she wants one.
The buyer picks your shelf with or without you
Al Ries and Jack Trout turned the problem into a law in The 22 Immutable Laws of Marketing: a company that can’t be first in a category should set up a new category it can be first in. Their concern was memory, which brand comes to mind when the category comes up. The shelf does more than that. It settles which budget pays for you, which title has the authority to sign, and what a fair price looks like, and it settles all three before anyone evaluates a feature.
A shopper who picks up an unfamiliar bottle in a grocery store decides what it is before she reads the label. Soda, juice, supplement, medicine. A CIO who meets a new technology does the same thing. Is it a CRM, a SIEM, an ERP, or something she has no shelf for yet? She’s trying to spend as little effort on the decision as she can, so she’ll place you somewhere whether the fit is good or not. When your category is undefined or keeps moving because you keep rewriting the homepage, no shelf fits.
The distance between two shelves is the distance between two budgets. A product read as a developer tool gets bought on a team’s discretionary line by a director who can sign alone. The same product read as infrastructure goes to a capital review, waits for the annual cycle, and needs a sign-off from the CFO’s office. Nothing about the software changed. What changed is the price it can carry, the number of weeks it takes to close, and the number of people who have to be convinced.
I’ve argued that a buyer who hasn’t been handed a way to read a new thing will supply one herself, reaching for whatever costs her the least thought. An undefined category makes the cheapest available thought I don’t know what this is.
You have two ways out. Pick one deliberately, before launch, rather than finding out at the end of a bad year which one you’ve been half-making.
Borrow a shelf
The first way is to anchor inside a large existing category and take a niche within it. “An AI-driven text-generation engine” gives a buyer nothing to hold on to. “A CRM for freelancers” lands instantly because buyers already know roughly what a CRM costs, which budget it comes out of, and who signs for it. You inherit the parent category’s stability and spend your scarce time explaining your niche instead of on teaching the market what kind of thing you are.
What you inherit along with the shelf is its ceiling. The parent category carries a price the market has already settled on, and charging above it means arguing against an anchor you picked for yourself. You also get graded on the incumbent’s feature list, because a buyer who has filed you under CRM will evaluate you as a CRM, including the twenty things you deliberately didn’t build. Borrowing is the safer path, and it’s the right one when the parent category is mature and your real innovation is the segment rather than the kind of thing.
Salesforce is usually told as a category-creation story, but the more interesting part is what it left alone. At the turn of the millennium sales leaders already knew what CRM was, what Siebel charged for it, and who owned the budget line, and Salesforce asked them to relearn none of it. What it renamed was the delivery model, with the word “software” crossed out in the logo. The familiar shelf carried the price anchor and the buyer along with it, so the whole argument narrowed to how the thing got installed.
I’ve borrowed a shelf myself. When we called Mesosphere’s product a Data Center Operating System, we weren’t asking anyone to learn a new kind of thing. Everyone already knew what an operating system was and knew you couldn’t run a computer without one, and we took fifty years of accumulated understanding for free. We invented the name and borrowed the shelf.
Build a shelf
The second is to build the product and its category together. Higher risk, higher ceiling, and more work than it sounds, because a category isn’t something you can ship. It exists only once other people are using it: when buyers have words for the space and when a company has a line in its budget to spend against. It gets built alongside the product, by you, and it never shows up in a release note. Three jobs come with it, and none of them is engineering.
Standardize the vocabulary so that the market has words for the space when you aren’t in the room. Define what winning looks like, because if you don’t set the measure the incumbents will apply theirs, and theirs was built to make their product win. And make the problem legible before you sell the answer. “Data gravity” is a problem; “container orchestration” is a solution. Categories are built on problems, because a buyer has to agree the problem exists before she’ll accept that a class of product addresses it.
Red Bull is the cleanest example outside software. Dietrich Mateschitz launched it in Austria in 1987 into a European market that had soft drinks and nothing else. The drink itself was not new: Red Bull is a reformulation of Krating Daeng, on sale in Thailand since 1976. What Mateschitz built from scratch was the shelf, and most of that work was vocabulary rather than product: what an energy drink is for, when you drink one, what it’s reasonable to pay. It took years, and the shelf they built now holds every competitor they have.
The Segway is the same path attempted with none of the three jobs done. Kamen built the product and expected the category to arrive on its own, and what vocabulary work happened went to legislatures rather than to buyers. The company had a legal class within three months of the unveiling and never got a budget line. It sold roughly 23,500 units in the U.S. between March 2002 and September 2006. Building a category and building a product are separate jobs, and finishing one of them earns no partial credit on the other.
Your early customers will hide the problem from you
Visionaries and early adopters buy interesting technology on its merits, and they’re happy to be the only company doing something. When you can’t tell them what kind of thing you are, they’ll work it out for themselves and enjoy the puzzle. They’ll take your call, sift through the architecture, and sign. Geoffrey Moore named the wall in Crossing the Chasm in 1991: the early majority buys known solutions to known problems in categories that already exist from vendors their peers can name. Your early wins are real revenue and they’re also camouflage because they came from the one group whose buying behavior doesn’t test the thing that’s broken.
You hit the wall when you run out of visionaries and start selling to the early majority. The damage shows up late in the sales process, far from the category decision that caused it. Cycles get longer. Deals need three more meetings. A champion who loves the product stops returning calls after he takes it to his boss, and the loss reason gets written down as “no decision” or “budget.” What happened is that a buyer who can’t categorize you can’t defend you internally, so ambiguity converts into perceived risk somewhere, usually in a room you weren’t in.
The work nobody schedules
Category work needs investment and stamina because the number that would justify it arrives after the budget cycle that has to fund it. A roadmap review happens every two weeks with a list you can check off. Evangelism has no weekly number, produces nothing demoable, and takes quarters to show up anywhere a board can see it.
Category work is also cheap while it’s still on paper and expensive afterward. Before launch, settling what kind of company you are costs an argument among the founders. After two years with no shelf, it costs a repositioning, with a price change, a sales team retrained off a pitch they’ve internalized, and a customer base that has to be told the thing they bought is now something else.
You can run a test this week without anyone’s permission. Call three customers who’ve been live for a quarter and ask each one how she described you to her boss when she asked for the money. You’re listening for one sentence, without your product name in it, naming a shelf her boss already had. If you get three different sentences, or a sentence that takes a paragraph, you’re already paying for it in cycle time and in losses booked as “no decision,” and the bill goes up every quarter you leave it alone.