Thursday, February 12, 2009

What is a market? (a guide for hackers)

(This post was inspired by a conversation with Nivi from Venture Hacks, but is otherwise not his fault)

There has been a proliferation of frameworks and metaphors lately that are designed to help startups avoid the all-too-common fatal mistake of failing to find a market. To wit: achieving product/market fit, getting customer validation, making something people want, things that matter, and of course the many excellent books on the topic, of which I'll mention just two of the best, Crossing the Chasm and The Innovator's Dilemma.

Having worked with dozens of founders over the past few months, I think I can safely say, without naming any names, that most of us are not too clear on what we're talking about when it comes to markets. Is a market a set of paying customers? Are there different types, or are they all similar? If so, how do we think about successful companies that don't charge money for their product? Are advertisers customers, and are they part of a market? Founders are constantly being barraged by incoherent and contradictory advice. For example, focus on finding a big market, but also, don't try to compete against large incumbents who already have a large market. But also, remember that eBay started out with pez dispensers but also remember that Google never advertised and I've heard that Facebook got it right from the start.

Few people want to admit that they don't understand what other people are talking about. That's especially true when jargon is flying and the stakes are high. For people who went to business school, I don't really know what to say. But I work with a lot of hackers-turned-founders who I do relate to. There's no earthly reason we should expect a programmer will have picked up a good understanding of market dynamics along the way, while they were busy figuring out how to grok partial template specialization.

Here's my attempt to explain market types using a metaphor most of us should be able to understand. It takes advantage of the idea, which I owe to Clayton Christensen, that customers buying a product are really "hiring" it to do a specific "job" for them. It's as if every customer, whether they are an enterprise, a small business, or an individual consumer, is actually an employer who wants to get things done. "People don't want to buy a quarter-inch drill. They want a quarter-inch hole!"

So imagine you're applying for a programming job. You're smart and have l33t skills. And yet, depending on who you talk to and what jobs you apply for, you may find an easier or harder time getting the job. Those differences are actually predictable, and fall into categories, and those categories are called market types. There are four:
  1. Existing market. This is applying for an open job req. An employer is trying to hire someone for a specific job, and they think they know what that job is. You might be smarter than the other applicants, but there are zillions of them. Because the employer has so many choices, they can afford to be very picky. This is the world of incomprehensible job postings, jargon, and HR keyword-based screening. It's important to understand how these companies see you: your resume looks like crap, your references are poor or nonexistent, and you don't have the mandated 10 years of experience in J2EE ERP CRM WTF. Tough sell. In the world of startups, this is like trying to sell a product to a very demanding customer who needs to see a lot of features before buying.

  2. Resegmented market (low cost variety). Here's a situation where you offer to do the job for 1/10th the cost. Instead of being paid $120k/yr, you are willing to do it for $12k/yr. This is how outsourcers get jobs they aren't otherwise "qualified" for. Keep in mind that a moderate price savings won't get it done; you can't call up traditional HR and say "I'll do the job for $105k" because that's not a meaningful sum to most companies, especially compared to the cost of making the wrong hire. In fact, you can't go through traditional channels at all. Recruiters and HR departments rarely recommend outsourcing - more likely, someone inside the company needs to get work done cheaply and circumvents the established hiring process to use an outside (cheap) vendor.

    Here your challenge is to convince that manager that you really can do the job at such low cost. Now your crappy resume and low-grade references become a strength: you're obviously not wasting money on marketing. Why do you think outsources have been happy to generate all this PR the past few years about people losing their jobs to India? Although it generates some political backlash, it establishes tremendous credibility among people who are desperately trying to save money. They must be thinking "if it's costing people their jobs, it must really work. Maybe I should try that..."

  3. Resegmented market (niche variety). This involves changing the job description. If the company is looking for programmers, you convince them they absolutely need Ruby programmers. Now, there's nobody with 10 years experience in doing ETL QVC in Ruby, so now your 3 years is starting to look pretty good. And now all those "qualified" candidates who you were competing against in scenario #1 are starting to look unqualified, because although they may be experts in something, they don't know Ruby, and you've convinced the client that Ruby is the be-all-end-all of programming languages. Remember Java?

  4. New market - this is like applying to a company that does not have an open req for programmers. Your challenge is to convince them to hire you anyway, even though they don't know what they need. Now even 10 years of experience is probably not good enough, because they have no idea how many years experience you ought to have. Without a basis for comparison, you first have to drive home the need, then you will have an easy time making the sale (after all, you're the person who they trust to bring it to their attention). A possible approach is to "steal" another job category. For example, "If you hire me you can free up 10 of your staff in department X, because I will write software that ..." Toughest part of the sale is to get agreement that you really can deliver benefits that they didn't previously know were possible.

    Here's the good and the bad news about new markets: you don't have any competition. When you call on a customer to try and get th job, it's unlikely that somebody else got there first. Unfortunately, that also means the csutomer is unlikely to know what you're talking about. Be prepared for an extended slog.
Think back to the confusing and contradictory advice I mentioned earlier. The #1 best way to cope with advice like that is to know your market type. That way, when someone say something like "you only have a few niche customers, so you can't be in a very big market" you can reply and (as a bonus) know what you're talking about. How about something like "don't worry! We're busy resegmenting the XYZ market with a disruptive low-price offering. Although that currently means we only can serve these few low-end customers, as our product improves, we'll eventually move up-market and kick the incumbents out."

They may not understand what you're talking about, but - don't worry! - they'll probably be too embarrassed to say anything.


  1. One definition of a market I like comes from Geoffrey Moore's Crossing the Chasm: "a high tech market is a set of actual or potential customers, for a given set of products or services, who have a common set of needs or wants, and who reference each other when making a buying decision."

    I think the last line "reference each other when making a buying decision" is key to understanding the distinction between what's required in different markets.

  2. If you're open to alternative metaphors with which to explain market-types, Porter's Five Forces, etc, try the video game console wars. Hackers tend to understand why one console trumps another. PS2? Trendy image, DVD and a successful pre-cursor. Wii? Innovative, Auntie-friendly controller, etc. Twenty years from now, operating systems might also work as a good metaphor.

    However, attempting to slot a market stereotype matrix on to the environment in which your start-up operates isn't always going to be particularly fruitful. It's more important that you just understand why your product exists, and how it manages to do so in relation to those (unique) start-ups which would just love to take your customers. Simply learn about your competition, use their products, and track their activities - and don't worry too much about an apparent lack of business buzz word usage in the day-to-day running of your business. Every start-up is different, so no market opportunity is going to be the same.

  3. great analogies. i'm three chapters into 4 steps to the epiphany and you've nicely described blank's four market types in more familiar terms.

  4. I think it's critical technologists understand market segments, even more than market types. I recently took a stab at explaining segments, based on Moore's definition Sean mentions above.

    One can see how it ripples through the entire business plan.