Targeting Matters!

The other day, Cindy Alvarez (product manager at KISSmetrics and author of a great blog), wrote a post called “Anybody, As Long As It’s Not You” , saying:

“Who should give you feedback on your early-stage product mockups/demo? Anybody, as long as it’s not you. OK, sure, there’s probably an “ideal” audience to show your product to.  But it probably doesn’t matter.”

demographicsOn Twitter, I commented about the importance of demographics / targeting, and Cindy responded that many products are not very specialized, so anybody’s feedback is a useful first step.  I won’t disagree with her — it *is* a useful first step. Still, I need to bang on a few points.  I touched on this the other day with my thoughts on product-market fit, and find myself hung up on the line “it probably doesn’t matter“.

First, most startups cannot design for ‘everyman”. Instead, you usually have a particular problem (or form of entertainment) in mind, which usually means you have a particular kind of customer in mind, even if broadly defined.  This is especially true with enterprise products (how big? what industry verticals? how high in the organization?), but it applies to a ton of consumer applications as well.

You might be creating a product that the whole world could use, but the reality is that you are going to start with a group of early adopters of one kind or another.  As a startup, your marketing resources are going to be limited so you need to focus your energies, otherwise you will waste time with an ineffective (and possibly expensive) shotgun approach or simply play a “hit and hope” game praying for viral adoption and media buzz. #notgood!

So yes, talk to lots of people, but categorize and filter your feedback based on what kind of people they are.  You are not just testing your product, you are also trying to answer the questions: who are my primary customers? who are my very first customers?

If you can even roughly answer these questions, you can figure out how these people learn about new products and structure your customer acquisition strategies accordingly.  Yes, it all needs to be tested and measured, but you want to pick a smart place to start. #focuswin!

I’ll pick startup examples out of a hat: Stardoll needed to test their product with teenage girls, not 40 year old men.  Smartbear needed programmers, not marketers.  Gilt Group needed fashionable urban women, not midwest farmers. At the start, Facebook cared about college kids not baby boomers. KISSMetrics wants Internet entrepreneurs, not retirees.  Foursquare needs social media geeks and young urbanites.  There are exceptions to all of this, but I bet you find it harder to think of them than the opposite.

As Steve Blank says, “get outside the building” and test your assumptions for success, but do not focus exclusively on product and forget about customer acquisition.  Your need to pivot might come from the problem you are trying to solve OR your product design OR your expected demographic. You need to test all of the above!

PS. where I see demographics being a little less important is usability testing. Yes, you want to test your primary customer, but watching *anyone* get confused as they try to use your software is painfully illuminating.

Final note: I’m not trying to put words in Cindy’s mouth because she *was* just talking about a first step, not ignoring this stuff.  I just felt the need to pound on the topic, perhaps not unlike the apes at the start of 2001: A Space Odyssey.

UPDATE:

I’m interested in gathering other opinions on the topic, posting them here, and seeing if I can’t advance my own thoughts.

  • Eric Ries: my $.02: targeting is something you discover, not something you decide. it’s important, but not if it keeps you in the bldg” andI’ve literally been the situation of: To customer: “get out of my way, I’m trying to talk to my target market””
  • Jason Cohen of A Smart Bear: “maybe you don’t know the perfect customer, but you can cut out a lot of folks, and that’s important.  It’s a similar argument to listening to feedback from people using the tool for free versus paying for.  With freemium you have many more people using for free than not, and usually that means their feedback overwhelms the others.  But frequently those willing to pay have different needs, and those are in the end the most important ones.”
  • Cindy Alvarez’s response, “Wait until your idea makes sense, then start targeting
  • Great post Giff.

    Feedback from the wrong people can be disastrous. As a market researcher by training, I am sometimes appalled by the lack of rigor that is put into creating the right controls for information gathering in the lean startup (and also customer development) methodologies.

    Much like technology has made it much cheaper to launch a startup, it is much easier to do proper market research (concept testing, etc.) Find your target market via adwords or facebook ads…. setup a survey monkey questionnaire… great actionable data for less than $100.

    The whole notion that you get feedback from anyone… or just go talk to a beta customer without any care taken to the right approach is foolhardy.

    • Joel, have you written up your recommendations for lean startups in terms of more rigor / better controls?

      • It’s on my to-do list… hopefully in the next week or two I can post something.

  • Great post Giff.

    Feedback from the wrong people can be disastrous. As a market researcher by training, I am sometimes appalled by the lack of rigor that is put into creating the right controls for information gathering in the lean startup (and also customer development) methodologies.

    Much like technology has made it much cheaper to launch a startup, it is much easier to do proper market research (concept testing, etc.) Find your target market via adwords or facebook ads…. setup a survey monkey questionnaire… great actionable data for less than $100.

    The whole notion that you get feedback from anyone… or just go talk to a beta customer without any care taken to the right approach is foolhardy.

  • I think Eric’s right to fear analysis-paralysis, buying think some sort of niche targeting makes sense to focus your design. I think it’s better to think in terms of behavior and psychology than demographics. And, as Eric says, be ready to pivot

  • I think Eric's right to fear analysis-paralysis, buying think some sort of niche targeting makes sense to focus your design. I think it's better to think in terms of behavior and psychology than demographics. And, as Eric says, be ready to pivot

  • Joel, have you written up your recommendations for lean startups in terms of more rigor / better controls?

  • It's on my to-do list… hopefully in the next week or two I can post something.

  • syedrayhan

    Giff, thanks for starting this conversation on a timely topic that I am focusing on right now. So, having read your thoughts and others here, I am thinking you discover your target market, but you can start with a broader universe based on the type of software you are building. In my case, that would be anyone doing IT projects (not just software), and then talk to anyone who fits this description. The first batch might need to be random within that universe. Then based on the result, we might start formulating more specific hypothesis about a specific group within that initial universe. Eventually, we should discover our target market.

    What do you think?

    • I think we’re on the same page — start by casting a wide net, within reason.

      The goal is to understand not just what product elements really resonate, but with *whom* and, as Jason Cohen stressed, who is willing to pay? It’s about being ruthless with yourself in terms of testing assumptions and keeping an open mind.

      In the case of Aprizi, which is a consumer play, I have my suspicions as to the primary makeup of my customer base, but I’ve been talking to lots of folks. As I get customer feedback, I am keeping note of what I’m hearing across different categories such as gender, age, location (urban, suburban, country), and personality/interests. I’m looking for commonalities that help us focus. With that stage done, I have some educated guesses based on what people *say*. The next stage is to really watch what they *do* when the alpha is up!

  • syedrayhan

    Giff, thanks for starting this conversation on a timely topic that I am focusing on right now. So, having read your thoughts and others here, I am thinking you discover your target market, but you can start with a broader universe based on the type of software you are building. In my case, that would be anyone doing IT projects (not just software), and then talk to anyone who fits this description. The first batch might need to be random within that universe. Then based on the result, we might start formulating more specific hypothesis about a specific group within that initial universe. Eventually, we should discover our target market.

    What do you think?

  • I think we're on the same page — start by casting a wide net, within reason.

    The goal is to understand not just what product elements really resonate, but with *whom* and, as Jason Cohen stressed, who is willing to pay? It's about being ruthless with yourself in terms of testing assumptions and keeping an open mind.

    In the case of Aprizi, which is a consumer play, I have my suspicions as to the primary makeup of my customer base, but I've been talking to lots of folks. As I get customer feedback, I am keeping note of what I'm hearing across different categories such as gender, age, location (urban, suburban, country), and personality/interests. I'm looking for commonalities that help us focus. With that stage done, I have some educated guesses based on what people *say*. The next stage is to really watch what they *do* when the alpha is up!

  • Steve Krug has a chapter in his new book that sums up my opinion: “Recruit loosely and grade on a curve”. How loosely you recruit depends on what you’re testing.

    For usability testing something like your home page, anyone outside the building qualifies. Discovering customers requires being more selective while keeping your eyes and ears open. I started my <a href="http://www.boxcloud.com"first product with graphic designers in mind but ended up services small business owners that struggled with FTP.

    • Thanks Ash, I’ve been meaning to check out Krug’s work (and I’m enjoying your startup blogging – keep it up).

  • ashmaurya

    Steve Krug has a chapter in his new book that sums up my opinion: “Recruit loosely and grade on a curve”. How loosely you recruit depends on what you're testing.

    For usability testing something like your home page, anyone outside the building qualifies. Discovering customers requires being more selective while keeping your eyes and ears open. I started my <a href=”http://www.boxcloud.com”first product with graphic designers in mind but ended up services small business owners that struggled with FTP.

  • Thanks Ash, I've been meaning to check out Krug's work (and I'm enjoying your startup blogging – keep it up).

  • Hi Giff,
    I’m with you on this one and unless I’m missing something, I’m a bit surprised this point is in question. IMO, target market is one of the most critical assumptions you test during customer discovery. I adhere to Moore’s defintion of market segment, which I wrote about for VentureBeat.

    1st step of Customer Development is testing your core C-P-S assumptions, customer-problem-solution. When one defines “customer,” one is guessing what the target is. So one “discovers” the target, in that, as you say, one “tests the assumption.”

    I’ve written several times on the subject of segmentation, since I think it’s one of the least understood marketing principles. (E.g., segmentation != demographics.) The segmentation exercise I outline here is perhaps overkill, but may be useful in your thinking.

    • You might be right — maybe this is obvious stuff, but I wonder if it doesn’t get a bit lost when people are talking about product-market fit. We talked about it the other day — you can’t just be fixated on 40% “very disappointed” without also really thinking about who is answering the question. I also think that you should be brainstorming about customer acquisition tactics from day 1, even if you don’t yet spend much money on it.

      Perhaps my impression is merely because of the need to prioritize the message. It *is* more important to first get people “out of the building”, or to shake bad habits of spending money to scale before proof points are really there (whether from entrepeneur ambition or VC pressure).

      I’m going to read the post you link to. I do think demographics are an important thing to think about for consumers – age, location, income level, etc – but certainly don’t believe that it is the only way you should be trying to understand customer commonalities.

      • I didn’t mean to imply it was obvious – but rather I’m surprised there’s not consensus. Demographics are definitely important, but not (necessarily) the determining factor of a market segment.

        Let me know what you think about the post. I think you’ll find it agrees with your opinion that you need to think about (and weigh) who is answering the question and what acquisition tactics might be.

        • (for anyone reading, Brant’s post is here: http://bit.ly/9VnXsf)
          I do agree that it’s smart to think about word of mouth (references) when thinking about a segment. I haven’t had time to make sense of your formula and see how I could use it effectively — might knock on your door for a more concrete example of it in practice.

          On a separate note, something we’re grappling with when it comes to this whole targeting question is on the UI design front: making an educated guess as to male/female split and age range in terms of our first UI look. (obviously it will iterate)

  • Hi Giff,
    I'm with you on this one and unless I'm missing something, I'm a bit surprised this is point in question. IMO, target market is one of the most critical assumptions you test during customer discovery. I adhere to Moore's defintion of market segment, which I wrote about for VentureBeat.

    1st step of Customer Development is testing your core C-P-S assumptions, customer-problem-solution. When one defines “customer,” one is guessing what the target is. So one “discovers” the target, in that, as you say, by “testing the assumption.”

    I've written several times on the subject of segmentation, since I think it's one of the least understood marketing principles. (E.g., segmentation != demographics.) The segmentation exercise I outline here is perhaps overkill, but may be useful in your thinking.

  • You might be right — maybe this is obvious stuff, but I wonder if it doesn't get a bit lost when people are talking about product-market fit. We talked about it the other day — you can't just be fixated on 40% “very disappointed” without also really thinking about who is answering the question. I also think that you should be brainstorming about customer acquisition tactics from day 1, even if you don't yet spend much money on it.

    Perhaps my impression is merely because of the need to prioritize the message. It *is* more important to first get people “out of the building”, or to shake bad habits of spending money to scale before proof points are really there (whether from entrepeneur ambition or VC pressure).

    I'm going to read the post you link to. I do think demographics are an important thing to think about for consumers – age, location, income level, etc – but certainly don't believe that it is the only way you should be trying to understand customer commonalities.

  • I didn't mean to imply it was obvious – but rather I'm surprised there's not consensus. Demographics are definitely important, but not (necessarily) the determining factor of a market segment.

    Let me know what you think about the post. I think you'll find it agrees with your opinion that you need to think about (and weigh) who is answering the question and what acquisition tactics might be.

  • (for anyone reading, Brant's post is here: http://bit.ly/9VnXsf)
    I do agree that it's smart to think about word of mouth (references) when thinking about a segment. I haven't had time to make sense of your formula and see how I could use it effectively — might knock on your door for a more concrete example of it in practice.

    On a separate note, something we're grappling with when it comes to this whole targeting question is on the UI design front: making an educated guess as to male/female split and age range in terms of our first UI look. (obviously it will iterate)

  • all right, all right already. I only write one blog post a week, so part 2 of my customer dev interviews series is going to have to wait.

    http://www.cindyalvarez.com/testing/wait-until-your-idea-makes-sense-then-start-targeting

    • thanks for taking the time to write more on the topic, Cindy. Heading over there now and have appended my post with a link to yours.

  • all right, all right already. I only write one blog post a week, so part 2 of my customer dev interviews series is going to have to wait.

    http://www.cindyalvarez.com/testing/wait-until-

  • thanks for taking the time to write more on the topic, Cindy. Heading over there now and have appended my post with a link to yours.