Nov 11

Discovering the One Sentence Behind Your Company

All truths are easy to understand once they are discovered; the point is to discover them. — Galileo Galilei

An IM popped up on my screen. “Hey Michelle, want to see Heroku?”

“What’s that?”

“Watch this,” messaged my friend.

I sat at my desk at Rackspace in anticipation of…anything. A few minutes later, he sent me a URL. On the page was the following message:

“Hey Michelle. I just deployed this Sinatra app for you on Heroku. Isn’t it cool?”

My head exploded. This wasn’t just “the cloud” — this was the cloud made simple. Here Amazon made the cloud scalable, and some group of guys in a YC startup made it so easy to manage for developers, it seemed like magic voodoo or something. Scary for us in 2009, but so awesome.

I instantly went to my boss and said “We should buy this company” and received a response that will remain confidential. Heroku impressed me so much, I moved to the Bay Area to work there.

Basically, they were scary as hell…

Why was I willing to stake my reputation with my boss who bought almost all of our acquisitions to show him Heroku? It wasn’t because they worked hard. I literally had no idea what technology was under the hood at that time. It was because I saw a product that solved something no one else was solving. Heroku made it simple for developers to deploy, manage, and scale apps. Combine this billion dollar problem with a starting price of free, and Heroku was scary as hell to me.

Product messaging strategy: cracking your purpose in one sentence

Sure enough, Heroku solved this pain point after dealing with it with dozens of high profile clients. Paul Graham invested in them after seeing the same issue of Rails deployment with countless startups.

Creating a pitch deck? Working on website copy? Drafting up a blog post about a feature release? Ensure you’re doing the proper discovery work with this checklist:

  • Check your ego.
  • Talk to your paying customers. Understand their needs. Study the market of what’s out there by reading analyst reports and talking to experts. Know your customer and be honest about the size of the market you can reasonably serve*.
  • Ask yourself, “What is the problem my product/feature solves for people? How does it solve it in a way the market currently does not solve?”

  • No one, not bloggers, customers, investors, not even good employees, care what your product is. Nobody cares about your fluffy buzzwords or simply the size of the industry you are going after. Your team only matters when it comes to one thing — what problem can they solve better than anyone else for the market they serve? Why does it exist, and for whom? Understanding your company’s “reason for being” in one sentence will save you and your employees a lot of grief not only in marketing — it simplifies product, hiring, and other decisions within your company as you scale.

    *hint: this *is* scary. It means you have to say no to a lot of people and money. By defining your market clearly, you will in better position to find more people who will easily say “yes”. Hey, even Amazon started off as a bookseller…

    Resist the urge to unleash the “Ship it Squirrel”

    Sometimes, you just gotta give this little guy a break.

    So go get a haircut and grab your business cards. Go straight to people in your target market, and listen…a lot. Take time and ask questions to define their pain points. THEN practice your pitch. Practice it hundreds of times, or at least dozens of times for feature releases. Look people straight in the eyes, listen to the tone of their voice, and track if they follow up to receive your product. You can’t really serve people if you aren’t even willing to have a conversation with them. The marketing part is much simpler once you get the people part — it’s just putting the message out there at a broader scale.

    Jul 02

    The “If a Feature Falls in the Forest…” Test

    It’s not uncommon among startups to maintain a blissful ignorance or a general disdain for Sales and Marketing. With so many tools and theories about how to run a marketing org, who can blame them? It’s much easier to focus on code. No customer meetings, no complex analytics tools to evaluate.

    It’s about listening, people

    Dog-Grooming-Big-EarsI have met some salespeople that upon initial inspection, seemed as dumb as a box of rocks. Then I heard them on calls and discovered how many questions they asked each customer. I saw how intently they listened, reframed a customer’s needs, and then patiently explain how a product could meet those needs. They followed up when it made sense. And shocker, these people could close. It’s not voodoo — it’s about listening.

    It doesn’t matter how technical you are as a founder. If you can’t name your customers’ top five pain points, you could be coding features no one is using. It’s possible that customers just don’t know about these features, but it’s also possible that customers never needed them in the first place. I guarantee a good salesperson can both name pain points and tell you how savvy customers are with your product in seconds — can you?

    The “If a Feature Falls in the Forest” Test

    treefallsinforestHere’s a test that will likely scare you. Every startup should do it. Heck, every tech company should do it.

    1. Install Google Analytics across both your website and your documentation. Let this collect data for at least a few months.
    2. Find the pages in your documentation and website that mention the features that differentiate your product from other solutions on the market.
    3. Go to “Behavior” -> Site Content -> All Pages.
    4. Search for these pages across your online presence. Look at the percentage of visitors who have even seen these pages, much less actually read about these features and then purchased your product.

    Is this percentage less than two percent? Even half a percent? Think about it — have you spent thousands of hours coding features that prospective customers don’t even know about?

    Playing to an audience

    Sales and Marketing is not about shoving a message down customers’ throats. Solid sales and marketing people listen more than they speak or write. They know which features customers want and know about. Good sales and marketing people don’t dictate what a Product team does, but they can and should serve as allies. After all, why even bother building features if no one actually needs or knows about them?

    Feb 06

    The Humility Ratio — a Litmus Test for Persuasive Copy

    Humility is not thinking less of yourself, it’s thinking of yourself less.

    — Rick Warren

    It can be easy to be tied up in thoughts of your own company, especially if you are a startup. You’ve lived, breathed, and sacrificed for this place, potentially for years. Startups are also hard. Here’s the difficult part about selling that vision though:

    Customers are fairly indifferent to how smart or noble you are, and how much you’ve sacrificed. They don’t have time for your mission. Customers are just busy, so they come to you for solutions to their problems. Your company exists to serve them.

    What a downer, right? You can pour your heart and soul into a product, and all they care about is whether or not it will save them a few bucks or time, or make them more productive or happy?


    A website is often your biggest opportunity to show that your company fulfills customers’ needs and/or desires. Does it do that? Here’s a good starting point to test:

    1.) Look at key pages of your website. Count how many times you see “we”, “our”, or “us” in the copy, in reference to your company.
    2.) Now, look at those same pages, and count how many times you see “you” or “your”.
    3.) Divide #1 by #2. We will call this “the Humility Ratio”.

    It turns out, “you” is one of the most persuasive words in the English language. People are more receptive when you address them instead of talking about yourself. That’s why successful companies that have a lot to brag about like Toyota to Apple still have solid Humility Ratios. Customers don’t care that these companies took painstaking care to ensure reliability — they care that their car or computer works when it is supposed to.

    So, what’s your company’s Humility Ratio?

    (pic borrowed from the eternally amazing Kathy Sierra)

    Dec 15

    Keep Austin Warm this Holiday

    OK. So the idea of Austin being July-like warm in December a sign that the apocalypse is nigh and perhaps the most depressing thought ever. Rest assure, Austin-lovers, the city has actually seen record cold temperatures this December. This is great if you hate the heat, but pretty brutal if you are new in town and don’t have any money for a coat.

    My friend Elaine Allan helps a group of refugees in town who are in this exact situation. Many of these people have never seen or felt this sort of cold in their lives. Imagine being dropped in Siberia without proper clothing, and then having to go to work or school everyday.

    If you are in Austin and have surplus coats, clothes or blankets, consider donating them to local refugees or others in need. If you don’t have extras, consider a cash donation. Simply leave a comment or email me at michelle(at) I will introduce you to Elaine. She is working tirelessly to make sure everyone who needs coats, blankets and warm clothes actually gets them.

    Not in Austin? Not to worry. It’s still good karma to donate to those who need it most, and a lot more fun than the typical holiday madness anyway.

    Happy Holidays to you and your family.


    Help kids like these in your area by donating coats and blankets.

    * Just a reminder: I don’t live in Austin anymore and haven’t for two years. I am just here visiting family. Please stop inviting me to your Austin events on Facebook. Thanks!

    Sep 03

    Leave Cancer in Our Wake by Supporting Swim Across America

    debzokissAfter four years of sickness and tremendous pain, my sister Debby Greer-Costello lost her battle against leukemia in May of 2009. She was just 42 years old.

    To this day, I struggle to make sense of Deb’s suffering. By the time her soul left her blistered, jaundiced body, she had been on oxygen, had a tracheotomy and dialysis, suffered from encephalitis that made her hallucinate, and couldn’t walk or eat. It has taken me years to accept that 1.) it was even ok to be angry about her treatment and 2.) eventually I could choose not to be angry and sad about this. After all, Deb raised tens of thousands of dollars for the Leukemia and Lymphoma Society even while struggling with her own illnesses. She was a beautiful person who cared tremendously for others her entire life. It really is stupid to dwell on anything else but that.

    So in her honor and at the coaxing of swim guru and Awesome Heroku VP of Sales Mike Pyle, I have joined the Swim Across America Salesforce team. Swim Across America is dedicated to raising money and awareness for cancer research, prevention and treatment through swimming-related events.

    Our team is committed to a swim in the San Francisco Bay on September 29. As anyone who knows me can attest, I hate cold. I’m also not a fan of sharks and underwater creepy things. The Bay apparently has both (although sharks aren’t as common as Alcatraz guards led to believe). I’m doing this even though I’m scared because this is a cakewalk compared to what the 1 in 2 men and 1 in 3 women who will get cancer in America have to face. It is my honor to help them in their fight, even in this small way.

    Cancer is so powerful, we can only fight it by cutting it out or injecting our bodies with poison to get rid of it. I’m sorry, but that’s bullshit. We need to get rid of this disease forever. Please help me do so by donating to Swim Across America at my fundraising page, and encouraging your friends and family to do the same. Awareness helps, but paying researchers to destroy cancer seems like a much sounder bet.

    If you have or have had cancer, feel free to drop me a line at michelle(at) and I will dedicate my swim to you. If you have had a loved one pass away from cancer, I will do the same for them. Please remember that you are not alone and let me know if I can help in any other way.

    Aug 29

    What A Marketer Learns After a Year of Attempted Hacking

    When Rob Spectre took a position as a Developer Evangelist and eventual Head of Devevangelism at Twilio, he learned that good products don’t exactly sell themselves. In this great post about what it is like to be in marketing after being a developer, Spectre declares, “Good marketing is a product of the same inputs as good code; long hours, sweating the details, and the judicious application of experience doing it the right way.” Here, here.

    I am in a similar boat as Rob, in that a large part of my job is not natural for me. As the Product Marketing Manager at Heroku, I felt it necessary to learn some code and actually deploy an app or two. I discovered one thing very quickly.

    Programming is f#@king hard.

    Programming is hard for me because I can’t just negotiate win/wins or write something up that convinces you that you will be more awesome simply by doing my evil bidding. I have to sit for hours on end, ignore IM and all the news and people, and just…learn. Pretty hard to do when you have the attention span of a spider monkey.

    I’m no master by any stretch, but here are some tips for marketers considering learning a thing or two about programming:

    1. Find “Your Way” and Stick with It
    There are no definitive guidebooks in open source programming. There is no one way to do something. One person will tell you you need Xcode or RVM, and someone else will tell you that those are total garbage. Every guide assumes you have a different level of programming knowledge, and they all start at different points. It’s best to stick with a few mentors and guides and just follow them all the way through.

    Chris Pine’s Learn to Program is great because it seems geared for people who have never programmed before. It doesn’t assume you know what certain terms mean already.

    2. Get Familiar with the Tools
    I’m learning Ruby and dabble with Rails, so I use a terminal, the text editor TextMate, my local host, and of course, git and Heroku.

    Every language requires you to install different things on your machine to get set up, and Windows, Mac and Linux are all different. Java developers don’t even use the terminal generally. I’d either make sure you walk through a complete tutorial or ask a friend to help you like I did.

    3. Programming Takes Discipline
    So, you finally get your machine set up and the tools you need to program, possibly in a Frankencomputer fashion if you did not follow Step #1. You finally dig in and start learning. Yay. You get confident, so you start watching “Breaking Bad” on Netflix at night instead of mastering your new language.


    Programming has fundamental concepts that make it easy to understand, but there really is a lot to know before you can even run anything, much less deploy a functional app. If you want to learn, you just have to keep at it. I can’t tell you how many times I’ve had to relearn the same things over and over.

    4. Google and Stack Overflow are Your Friends
    If you don’t know what a programming term means or how to do something, Google it. If you don’t understand what some of the terms mean in the answer, Google them. It seems super wasteful as an exercise, but there is no fast way to get around this and eventually it does sink in a little. Wikipedia is somewhat helpful, but intimidating because everything is explained from a computer scientists’ view.

    There are lots of newbie tips in Stack Overflow. You will get to know it.

    5. Learning to Code, Manage a Database, and Design at the Same Time is Not Possible
    Apps have interfaces and apps have databases. It’s hard to build all of the elements on your own.

    CSS is the style sheet language that gives apps, well, a style. It requires the memorization then subtle nudging of elements on a page. I downloaded this simple theme that consists of one page with some Javascript just to change the CSS. After poking around with it, I have no idea how designers deal with this stuff everyday. Very tedious.

    Data seems fun and useful to me though. Our developers set me up with Heroku Postgres Dataclips so I can poke around with SQL queries without breaking anything. O’Reilly has some decent books on SQL if you’d like to learn SQL concepts.

    6. Be Social! It’s Hard to Learn This Stuff on Your Own
    Google only goes so far, so real friends can help you get the basics down too.

    I hear about code at the lunch table every day and often ask coworkers dumb questions. This means I end up happily playing copyeditor or Analytics nerd for anyone who asks. I’ve also bribed people with cookies and wine for offering up a little help. Reciprocity and gratitude seem key here, so use your best judgment.

    Too shy or too far away from a crew of helpful developers? Rails Girls classes were helpful as well.

    <3 Developers by Understanding a Little About What They Actually Do

    For ten years, I’ve spent my energy keeping up the latest marketing tools, news, and blogs but couldn’t navigate a terminal in the slightest. I do hope to learn more about programming and encourage other marketers to do the same. When we understand how our products are built, we improve feedback loops between the problems customers have and the solutions developers choose to solve them. Everyone is happy and we go eat pie.

    Aug 28

    Why Geolocation Isn’t Quite “There Yet”

    Hurricane Irene Captured August 26, 2011This weekend, we cowered to think that major cities on the American east coast would be pummeled by Hurricane Irene. We watched CNN and Twitter to see how things were going.

    Geolocation tools like Foursquare could have been amazing during a natural disaster like Irene. Imagine if any check-in tagged with #Irene were pulled onto a map, and then you could see what sort of damage was done to millions of places in real time? How useful would something like that be?

    Geolocation picked up as a meme in early 2010 and then slowed down significantly for the following reasons:

    1.) Let’s face it: geolocation can be creepy. OK, so let me get this straight. You are going to give me rewards for letting the entire world know where I hang out on a regular basis? I see the benefits, but I’m not sure they outweigh the risks. This is especially true for public accounts.

    2.) Many geolocation networks got numbers, but failed to capture critical mass…anywhere. I won’t get into naming names, but I’ve seen social networks with gobs of features fail. Why? Features do help you sell a social network. However, unless our friends are using it, we don’t care. Facebook had a lot of features from the beginning. What really helped it take off was the fact that they targeted schools hard and heavy before expanding. They didn’t rush to get 10 million users all over the world. They got 30,000 of heavily concentrated users, and then got 30,000 more concentrated users. It became so popular at the “it” places (Harvard, Yale, Stanford, etc.), everyone wanted an account.

    It’s not just important that people sign up for your site. It’s important that they use it. To Foursquare’s credit, it helped them tremendously to lock down New York City and San Francisco.

    3.) Geolocation desperately needs context that can be defined through APIs. The Hurricane Irene example above is a fantastic example of a useful app that could be built quickly on geolocation APIs. What if someone wanted to build, which would allow users to check into and find spots that served organic meals? What if we wanted to call out restaurants who had bad service by checking in and leaving comments, or crowd source bike routes by checking in to spots along the way? It’s hard to add functionality to geolocation applications that will make everyone happy, but it’s relatively easy through services like Apigee or Mashery to build robust APIs that support any kind of functionality through third party applications.

    If an application like Foursquare can provide the current data and basic functionality a geolocation service would want to use, they could theoretically be the Facebook of the real world. Instead of sharing links, we would share places. We could play games in the real world the same way people play Facebook games like FarmVille and Mafia Wars. Foursquare has some action around their API and could grow significantly towards this vision in the near future.

    I’ve said this before and will say it again: the money in a social network is often in the lurkers. Make something valuable for them and things will get interesting in geolocation again very quickly.

    Aug 24

    How Can We Clone Steve Jobs?

    Tim O’Reilly’s Google+ stream tipped me off to a very interesting post on Forbes about how outsourcing both helps and hurts us. You really should read it yourself, but the long and the short of it is this:
    1.) We have to outsource where it makes sense, to keep pricing competitive, but
    2.) As soon as we do it, whomever we outsource to will most certainly approach our competitor with the same proposition, so
    3.) We end up back where we started.

    The key to solving the equation? Outsource where it makes sense, but always add innovation.

    Say what you will about Steve Jobs as a person. As a figurehead of Apple, his vision, passion and perfectionism as CEO of Apple are unmatched. Apple products are copied mercilessly and are the source of tech lust all over the world. It is why despite Michael Dell stating Apple should close shop and give money back to its shareholders, Apple is now worth more than Dell. Oh and Microsoft and Intel combined.

    Whether we tax high and offer decent public services, or tax low and allow the business sector to cover the weight, it doesn’t matter. If we don’t innovate, we will cease to be relevant.

    So I’m asking you, America: quit arguing. Just figure out how we can make more Steve Jobs(es?).

    Let’s look at what helped make Steve Jobs successful:
    1.) Republicans like Rich Perry seem to be strangling education for the sake of keeping taxes low. Steve Jobs went to a public high school in Cupertino, California. So did Larry Page and Marissa Mayer from Google. Yet today, we are cutting teachers and school funding is as low in some areas as it was in the Depression era. That doesn’t seem conducive to being competitive in a global market.

    That being said, we really need to address teachers’ unions ability to protect bad teachers from being fired. Mr. Jobs is not a fan of this policy, and he doesn’t like teaching kids to take tests either.

    2.) Apple is in Cupertino, which is in a hotbed of technology. Apple talent could come from Berkeley, Stanford, CalTech, and a host of other technical schools. So while Jobs himself got his education at Hewlett Packard with Woz, he was empowered by an educated base.

    Right now, college tuition is increasing, but college kids are not learning about technology at the rate the private sector is producing it. Ask any tech company how difficult it is to hire developers and you’ll see.

    3.) Reform patent law. I’m not a lawyer or an inventor, so I can’t confess I have a solution here. I just can’t imagine inventing anything with as many patent trolls as we see now.

    4.) Reward innovation and people who think out of the box. Foster change makers. And for God’s sake, quit watching those ridiculous mindless television shows.

    The G.I. Bill produced more engineers among the “Greatest Generation” than any other generation before or since. It helped foster engineers that put a man on the moon. Can our nation create an entire generation of Steve Jobs(es?) that innovate us out of this recession, or are we destined to care more about “Dancing with the Stars” than the X Prize?

    Jul 27

    Joi Ito Explains Creative Commons

    My good friend and multi-talented designer of this blog Shyama Golden and I were discussing how confusing the internet landscape can be to an artist or musician. Seesmic founder Loic Le Meur asks Founder of Creative Commons Joi Ito what you are actually buying each time you purchase a CD, and how Creative Commons works.

    Jul 18

    Caveats for Quantifying Online Influence

    So Omar Gallaga wrote an Austin American Statesman piece about how our lives are increasingly becoming measured by our online presences. It was a well written piece and probably opened a lot of people’s eyes about how they are being judged in a way they had not previously recognized. You should read it if you have not already.

    The piece talks about Klout, a startup that created a scoring system to measure one’s online influence. Klout uses algorithms that evaluate who you are talking to and how often in social media circles the same way Google PageRank does. Some PR and social media firms are using it to reach out to influencers, and Klout now has perks for people who have Klout scores above a certain number. To reach people with high Klout scores, you can contact the company to “get in”.

    This approach is too simplistic, like an easy button PR firms can hit so they can say to their clients, “Hey, we got you covered on this launch”. While I myself value what Klout offers, it is a factor within many factors when considering how to expand your brand’s presence.

    Let’s use this graphic/scenario on the right to show why brands really should take more care in managing their online reputation.***

    1.) Influence is never static. The influence this would wield today would be much more damaging to my brand than it would a month ago. Although Klout scores do change over time, they don’t change overnight.

    That being said, an awesome potential subject matter expert and influencer could have a low Klout score today, but could increase it over time. Is it wise to make that person feel less important than Rupert Murdoch?

    2.) A person’s influence on a subject matter depends on his or her expertise in that field. Rupert Murdoch and crew don’t seem to get new media and the blogosphere. If people judged me by News Corp’s record in new media, I could actually lose influence by association.

    3.) Influence must always be considered qualitative as well as quantitative. I don’t respect Rupert Murdoch. Say I went to a PR firm and said, “I want media influencers with Klout scores above 60 to tweet about my blog”, they could very easily return to me with someone like Rupert Murdoch. Do I want his endorsement, or the types of clients he would bring? What if he becomes my #1 nightmare customer and his followers are almost as bad? I marketed to him because he’s big instead of paying attention to whether or not his followers would benefit by my product.

    4.) PR is just as much about building community as it is “spreading the word”. When you judge people purely on their influence number or Googlibility, you get a bunch of people who are “kind of a big deal”. People who are “big deals” can be demanding as customers, and often demand completely different things. That can make a product that appeases the majority of them very difficult.

    I’m not saying you should ignore Klout completely. I just question practices like judging customers based on some sort of score instead of taking the time to evaluate how much value you can offer their friends and followers. Klout works best through a software API and paired with other metrics.

    **Relax. I don’t actually know Rupert Murdoch and I’m not tapping your phone. I do sometimes sneak in your house and rearrange your furniture though. Anywho…