Startups Unplugged

Author: Paul Orlando

  • Four simple things that make a better accelerator experience

    With a growing number of startup accelerators, bootcamps and incubators out there, here are some tips for having a good experience once you’re in a program. This comes from running two startup accelerators (three-month full-time programs) and three bootcamps (three-month part-time programs).

    Be Coachable. Disagreeing with feedback is fine and necessary at times but be open to new ideas and ways of operating. There’s no point to participate in a program if you’re not willing to try something new.

    Be Social. There are a few types of being social. One is getting to know the other startups in the program, helping each other out,  going out for drinks or dinner. Too much of that and you wake up to find you didn’t get enough work done. Not enough and you leave without getting to know new people who you might work with in the future and without getting their assistance during the program. Find the balance that works for you.

    Don’t Disappear. I’ve seen startups withhold news that is obviously bad but necessary to share, such as the loss of a co-founder or issues with money. Disappearing means that we can’t help — because we don’t know.

    Don’t Lose Momentum. After the program, no matter how disciplined you are you will probably lose momentum. We try to keep things going by continuing to talk to the startups afterwards, but it doesn’t compare to being in the thick of things during the program. Create habits that will help you continue after the program ends, such as keeping a regular group meeting and using mentors’ and advisors’ expectations to keep the pressure on.

  • How I Tried To Be a Better Startup Weekend Judge

    [I originally wrote this piece in early 2013. I just revisited it again in March 2018. I hope you enjoy it.]

    Working with many startups, I get to see how lots of people test ideas and build their startups. I also get to see how lots of people want to rush off and build in the beginning, before they know whether their ideas are any good. I’ve also seen a condensed version of this in Startup Weekend. But the truth is, even though I’ve mentored at three Startup Weekends (New York, Hong Kong and Shenzhen) and given pre-Startup Weekend talks, I’ve never been a participant in the event. And now I was to judge at the next Startup Weekend…

    So I did the only logical thing in this situation.

    I decided to hold my own one-person Startup Weekend to see what it was like. I’ll sum up the experience for you.

    It sucked. But I understand even more about why people have such difficulty doing well in these situations.

    I had a few rules for how I would participate in this self-made Startup Weekend.

    First, I would keep track of how I felt during the two days and would do more of what was enjoyable and less of what was not (sometimes bad choices that I thought participants would be likely to make). I wouldn’t ask for help with the explanation that I’m doing a test, since I think most Startup Weekend participants can’t (or shouldn’t) just ask their friends when they need customer interview subjects and instead should go out and try to find the right people to interview. I’d also have to be developer, designer and business guy without extra help (this one was more about not wanting to drag people in on a holiday weekend). Also, I couldn’t work myself like a real Startup Weekend, since, well, I had other things to do on the weekend. So, overall, this was going to be Startup Weekend-lite, but crazy also.

    In terms of being a “good” participant, I was terrible. That was clear from the start. What happens when you let yourself choose “fun” during a competitive learning process is poor results. I felt like building and wanted to postpone talking to people. This was made even more difficult for a few reasons.

    1. I had no team. That made a big difference. So much of the overall impact of the event comes from other people in attendance. Now, as a one-person experimental startup, I had to act all alone. And, I had no one there to encourage me to put the hours in.
    2. I had no other guidance along the way. This actually makes quite a bit of difference. Just being able to talk with someone else about your ideas keeps you from wasting time.
    3. It poured that weekend. As I looked at the rain streaming down outside, I have to admit I was a little relieved that the weather gave me an excuse not to go out and interview people.

    The service I was building was to help people practice their pitches — something I could have found lots of customers for at a Startup Weekend, but not during a normal weekend, especially a holiday weekend. Eleven people I contacted to interview were either unreachable or pushed me off until the following week. According to my rules, I didn’t try to explain that I was working on a self-imposed artificial deadline. Instead, I fell on the unspoken evil of Startup Weekend attendees — the online survey. How else was I to get data so quickly? I received five responses to the survey in the time that I could officially use them (other responses came in afterward).

    Further, I was trying to do all of that while also registering a domain (pitchreviewer.com), setting up hosting, email (incorrect MX records led to a couple hours delay), designing a new site and building the two services I was testing. Since I thought that early-stage startups are usually terrible at pitching themselves, I built a service to solve that. The free service I tested was a phone number to call to hear a list of questions you might get when presenting your pitch. The paid service was a grade of you pitching on video and recommendations for how to improve. I built a basic Twilio app for the first (call a phone number and hear questions like “What problem are you solving?” and “How have you tested this idea?”) and relied on PayPal and email for the second.

    None of this mattered since I felt uncomfortable about blasting out this fake startup to my contacts. I used a coupon for online ads instead. In the one and a half days I had, I managed to attract one person, I think by accident, to the site. This one person tried neither service.

    As Reid Hoffman said, “If you are not embarrassed by the first version of your product, you’ve launched too late.” I was embarrassed but also waited too long for things that didn’t matter. While I spent time designing buttons for the new site (perhaps the most fun of the process) I scrapped all but one of them when I simplified the design.

    I also couldn’t truly fulfill the requirement to do much customer development and test the business model (which I could explain, but had no data for since no one chose between the services).

    So, suggestions for you include:

    • Form a balanced team,
    • Get out of the building and interview potential customers either in-person or online as soon as possible (or set up a schedule starting Friday night of Startup Weekend, at the latest), Start testing as soon as possible — which is probably sooner than you think,
    • Remind yourself of the business model, customer development and execution criteria every few hours so you do what has value. Have one person ask “what are we trying to learn and how could we test it faster” every time there’s a new feature suggested,
    • Talk to mentors and other teams and get to know people you might work with after the weekend,
    • Check out other resources before you begin, such as this Startup Weekend Toolkit.

    Do all of that and I won’t be too hard on you as a judge.

    List of tools and services I used: PHP, CSS, Adobe Illustrator, WordPress, HostGator, GoDaddy, RetailMeNot (for coupons), Bing Ads, Google Analytics, PayPal, Twilio. Also, three glasses of a bottle of Scotch from Steve Messina and a remaining 1/2 bottle of wine from Jeffrey Broer.

  • The Recommended Daily Intake Approach to Startup Activities

    You’ve probably seen the “Recommended Daily Intake” used as a way to gauge how much of a particular nutrient or substance we should ingest in a given period. I was thinking about how startup people spend their time and ended up writing this table of “Recommended Startup Activity Intakes” for early-stage startups.

    • Reading popular tech blogs < 5 articles per week. In general, popular tech blogs preach an unhelpful message of  the “Hollywood” startup.
    • Reading articles in the industry that you’re focused on and on customers that you’re focused on > 5 times per week. You should know your industry, including the things that your customers are reading.
    • Meeting potential customers > 10 times per week. The Vitamin C of this list.
    • One-to-one coffees, drinks, meetings with other entrepreneurs, mentors or other people who can help > 3 times per week.
    • General startup events < 1 every two weeks. Less often if you usually talk to the same people each time at these events.
    • Attending events specific to your area of focus > 2 per month. In other words, these are not “startup events,” unless you are one of the few startups that sells to other startups.
    • Demo Days < 4 per year. I say this even though these are the only events I run. Go to see people who have built something talk about what they’re done. But realize that you don’t see the full picture at a demo day.
    • Pitch events < 1 per year. The processed sugar of the startup world. Go once to see how they’re run and to get a free drink. Then avoid them. The only exception is when their panel of judges includes people who have recently invested in early-stage startups in the location of the event.
    • Building your company = every damn day. OK, back to work!

  • The Two Sides of Demo Day You Might Never See

    I’ve run three different startup accelerator programs. Two of them ran demo days, but the one I currently run does not. For an explanation of why, see this post on side effects.

    Demo days can be good or bad depending on your goals and the people involved.

    I thought that this would be a good time to talk about the action takes place before and after demo day. As with many things in life, the real story is not where the attention is. There are two sides to any demo day that you might never see if you don’t know what to look for.

    The Side Before Demo Day

    Before a demo day there are at minimum months of work that most people never see. Some of this work might take place years earlier when the entrepreneurs gained skills and an outlook that led them to the problem they chose. Months or years spent putting together a team, talking to customers and understanding the market… If the startups participated in an accelerator program, typically being three months in length, they were probably also guided with workshops, interviewing customers, meeting other entrepreneurs and investors and lots of presentation practice.

    Then it all comes down to a few minutes at the front of a room.

    Before running the Hong Kong Lean Startup Bootcamp and AcceleratorHK years ago, I visited programs around the world. I visited as a mentor, spoke to participants, alumni and people who ran the programs. Most of the time I saw a lot of value generated.  I also saw a few cases of when there were misaligned incentives between the participants and the programs. The worst of these being when startups who were “behind” in their three-month growth sprint were encouraged to put on a good show at the expense of their true direction. I’ll never be involved in a program that does that. Even more recently, in the Laudato Si Challenge in Rome and the Incubator at USC, I keep learning new ways programs run and changing the model based on who is involved. This constant change is crucial.

    The Side After Demo Day

    It’s so easy to look at demo day as the end goal. For any startup presenting at a demo day, their work is just beginning. However, after the presentations and party, it’s easy to walk away thinking that the goal of demo day… is demo day. But really, if your company didn’t raise money as a result, was there another strong purpose for spending so much time on this one short presentation? (Note: I don’t believe that there is an inherent good to this type of stylized presentation prep.)

    Post-demo day memories can be short. I remember some of the first demos I saw in New York years ago, before I built my first startup. I saw some great presentations, with speakers and visuals I still remember. Checking my notes a year later I looked up these same startups and what did I find? Some successes but also lots of complete changes of direction, “domain for sale,” and blogs not updated for a year. Change, exhaustion and disappearance are expected parts of a creative process.

    Demo day is just the beginning of a long process.

    Why do demo days exist?

    With the typical startup program lasting only three months, holding a demo day at the end can take away valuable work time.

    Sometimes programs need demo days as a way to stay focused. We need them as a way to learn how to present to an audience. The startups have worked hard and have something to show for it. Demo days are a forum to meet their communities and to talk publicly, often for the first time, about their work.

    And we need demo days as a way to celebrate.

    But evaluating startups only by their demo day is like evaluating people by what they did on their first birthday.

    Keep working, guys! We’re proud of you.

  • Startup Weekend Toolkit

    Here’s a Startup Weekend Toolkit I put together after mentoring and judging. It’s a lot of info, so it’s best to read this before you go since you will be busy enough (and sleep deprived) during the weekend, but these resources will save you lots of pain.

    Startup Weekend isn’t a hackathon or a business plan competition… So work on these activities instead:

    1. Customer Development. “Get out of the building” and talk to people. Learn from them. What hypotheses do you want to test? Do your potential customers support and reinforce your assumptions? What qualifies success or failure?

    2. Business Model. Express what you learned in customer discovery: Who is your customer? What is your core value proposition? What are your key activities? What are your revenue streams? What is your cost structure? Who/what are your key partners/resources? What are your distribution channels? What is your roll-out strategy?

    3. Execution. What did your team build during the weekend?

    4. Your presentation. Not being able to express what you did is the unspoken failure point of many teams. Start to practice on Saturday, not Sunday.

    Are you really building an MVP?

    Working with lots of startups, the most common error I see people who think they are working on an MVP is that they do not understand what an MVP is. An MVP is focused on getting validated learning. Write down your hypothesis and what will qualify as success or failure for you. See more about that below. Otherwise, how do you know if you’re learning anything?

    A sample hypothesis format:
    I believe that [customer segment] has [problem / need] and [a specific number / percentage of them] will do [action / use this solution].

    Hypothesis Formation to guide you as you build an MVP and then collect data. Read this brief article for guidelines.
    Questions to consider: What are your riskiest assumptions? What MVP will enable you to test your riskiest assumptions? What will you accept as validation that you are solving your customers’ problem?

    MVP case studies:
    Think about how Dropbox, Aardvark, The Ladders etc tested their services before building them out.

    Some techniques to build an MVP:
    – Landing Page — put up a landing page that leads the visitor to believe that the product has been built out. Then test user actions — do they try to sign up, do they try to buy, etc;
    – Mockups / Wireframing — these can be on paper, using tools as simple as powerpoint (where I like to hyperlink fields on the page and connect them to other pages to make it feel like a website), or more sophisticated tools like InvisionApp or Balsamiq. Give the user the feeling of using your product before you write any code;
    – Video — make a video to demonstrate what your service does before you build it. This could be as simple as text or basic images or screenshots with a voice-over;
    – “Concierge” — use humans to emulate what the software would do later on, if you were to build it;
    – Prototype — actually build a working version.
    – “Parasite” — this is the name I use for riding on top of another network (Twitter, FB, Skype) and using the users there as test subjects rather than trying to drive new users to your new service.

    Doing Customer Discovery Interviews:

    1) Ask about situations in which potential customers might be aware of the problem you’re trying to solve and then ask questions around that. This means that if you’re trying to solve the problem of finding cheap vacation travel packages, I’d ask “Do you like traveling?” and ask them to talk about that, instead of directly asking “Do you want to find cheap flights for your vacations?” (which is a question that forces them to say “yes, of course” — not that helpful to you).
    2) How do you solve this problem currently? or What do you do to make it less painful? In the above example, they might have signed up for notifications for cheap flights from online discount travel agencies. This will tell you about the competitors who are already involved in solving this problem. Not just the existing business competitors, but also ways that people deal with this problem by themselves.
    3) Can you describe this problem to me in your own words? They may have already told you about this without you having to ask them. Directly asking this also only works if you didn’t already force the issue by asking them if they have the problem. Otherwise they’ll just repeat what you asked them already. This may also give you keywords to put on your landing page and in ads.
    4) Are there other solutions out there that you’ve tried in the past? This will make you aware of competitors or other services addressing similar problems.
    5) What do you wish you could do to solve this problem (even if it isn’t practical)? This can be really powerful, especially if you find out that something they think is not practical is now practical for you.

    Cool things you could show:

    • Number of people you spoke to;
    • What you learned that was unexpected;
    • Actual purchase orders;
    • How you hacked the system to get it all done.


    Watch out for these things. I made this list by surveying past participants:

    • Mentor whiplash (getting pulled in different directions by different mentors);
    • Replacing customer interviews with online surveys (because you run out of time or are too hesitant to find interview subjects);
    • Hesitating to “get out of the building;”
    • Underestimating the time it takes to talk to people;
    • Building too much, testing the problem too little;
    • Not taking risks, not being creative;
    • Not practicing the pitch enough;
    • Thinking too much about the judging;
    • Team fighting, worrying about IP protection, getting distracted, no team balance, team too large.


    Practice your pitch. Pitch structure ideas:

    • Introduce yourself and team;
    • Introduce the problem, Show your solution;
    • Talk about how you got there: Customer Discovery work you did, your Business Model.

    And for what happened when I tried to run my own one-person Startup Weekend before I was a judge for the first time, read this.

  • Going Gonzo to Build a Better Bootcamp

    Hunter S. Thompson practiced Gonzo journalism, a format where the reporter participates in the story. Here’s the start of the Gonzo startup advisory (what I used to run as a “bootcamp”).

    In my experience, if you’re an early-stage entrepreneur and working with someone like me to learn Customer Development techniques, you will have some common troubles. One of the most common ones is to use customer interviews as a way to learn about your target customers, their problems and what they value. It will take you time to become comfortable interviewing or even with the idea of going out to talk to people. You’ll delay doing it entirely. And when I think you’ve understood it I won’t really know how it goes out in the wild.

    I’m convinced that practicing Customer Development is a set of skills that cannot be learned without active practice, so I’m going gonzo. Salim Virani brought up the participatory idea over dinner and I’m making it a part of my work. I will go on customer interviews with startups in the early days of working together to observe, take notes and when needed, contribute to the discussions. I’ll give direct feedback right after the interviews so that the founders can improve more quickly without me needing to guess how well it’s going.

    I’ll report back later on results from going Gonzo.

  • Why Startup Lists Don’t Matter But You Read Them Anyway

    It just turned 2013 in Hong Kong and I feel like it’s 2008 in New York.

    Usually when I say that it’s because the startup growth I witnessed (and hope helped) in New York years ago reminds me of what I’ve seen in Hong Kong recently. This time, however, I’m talking about the latest in ecosystem envy. Startup reports are here again, this time in the form of Startup Genome Project and World Startup Report.

    Fact: Hong Kong made neither list. Question: does it matter?

    In the case of Startup Genome, I was one of the thousands of startups that contributed to their initial dataset in 2011. Their findings were actionable (though to be taken with a grain of salt), presented as applicable in different market types (more about that later) and changed the way we thought about survival probabilities.

    I was fascinated with their collection of data and suggestions for effectiveness of different models across market types. When they published their Startup Ecosystem Report I took it as another interesting data collection and not surprisingly, Hong Kong was not on their top 20 list.

    Likewise, when World Startup Report published their travel schedule for 2013, Hong Kong was not on the itinerary. Again, some people felt bad, seeing Hong Kong excluded. That’s on top of Dave McClure not stopping over on his way around the region recently. But really, we’re looking at this backwards.

    The goal of any city is not to get itself on a startup list. Being on a list or lobbying for a closer look doesn’t achieve anything meaningful. The goal is to do what’s best for your location and then maybe when you’re too busy to notice, someone mentions that you’re listed as a new top location. Hong Kong can’t mandate its way to global importance. The next list will be entrepreneur-led.

  • School’s Out

    A few years ago when I started working on a social voice service I thought a potential customer segment was college students. I spoke to around 120 students to test the concept. During that time I learned a lot about design and UX but also realized that this segment did not have the problems I thought. College students as a segment was invalidated but I was struck by how I was able to easily get access to people on or near campus.

    Recently, while running a Startup Weekend-like hackathon we gave teams themes to build for. Since Customer Development was required, they went out to test their hypotheses.

    Two days later I was surprised to see that most of the teams built for and tested with college students.

    Nothing in particular would have led them to this segment. But many of them had the same idea. In general, they also did more interviews and made more progress in learning than teams that chose other segments.

    But apart from competitions, should startups test their real products with college students? Among the good reasons to test with college students: startups know how to find them, accessibility is pretty easy, they exist in large numbers and they’re willing to stop and talk. They will forget you after you make mistakes interviewing them. The bad reasons: you’re probably not really building a solution for them.

    Beyond college students, I see similarities with startups that develop for any market that is easily reached compared to those who deal with difficult to reach markets. I’ve seen teams who are building something valuable spend a lot of time just trying to get in front of people. And I’ve seen other teams find creative approaches to reaching their markets.
    In an artificial environment like a fake two-day project, going with an easily reached segment can be smart. But for a startup building something real over time, it’s better to figure out who your segment is.

    The faster you do this, the faster you’ll learn if you’re on to something.

    Note: I’ve since seen this behavior repeated in multiple other startup programs. After I started to teach entrepreneurship at USC, I flipped the above around and actively encourage students to build projects for a college audience. The point being, build a Disposable Startup in order to get operational experience.

  • Startup Weekend To Dos

    This weekend there will be more than the usual number of #lean hashtags on Twitter. Startup Weekend is holding its 54-hour events around the world, including in Hong Kong. I recently gave a pre-Startup Weekend talk on “Lean Startup in 54 Minutes” and I thought I’d share the major parts of the talk here.

    This isn’t a hackathon (where you focus on building) or a business plan competition (where you focus on untested projections of fantasy). Instead, work on Customer Development, your Business Model, and building an MVP to validate your idea.

    Get familiar with the radical idea of talking to other human beings. I know it’s more comfortable in front of the glow of your laptop screen, but when you get out of the building and speak to real people, the sunlight of reality will help you learn a lot. In my experience, people wait too long to start doing this.

    Learn about the Business Model Canvas. It will help organize your thinking and open you up to new opportunities.

    Most people who say they’re building an MVP are not. They’re really just building a basic version of their product. That’s not an MVP. An MVP is focused on getting you validated learning. An MVP doesn’t need to be coded, either. You could make a video, use paper mockups, or do a “concierge” service where people perform all the behind-the-scenes among other techniques. Just remember that the point is to learn, not to just start building.

    For more on all this, I hope this toolkit will be helpful to you. Good luck!

  • Launching AcceleratorHK

    Today is the first day of AcceleratorHK, Hong Kong’s first startup accelerator. I’m happy to be the Director of the program.

    The process of getting this off the ground was straightforward but took a while:
    – Lots of discussions with people in HK’s tech community, starting December 2011
    – Lots of discussions with other accelerator programs, throughout the US, Asia/Pacific and Europe. We’ve managed to take some best practices while also building something that works in Hong Kong
    – Running Startup Bootcamp, which allowed me to test the format and content. I will continue to run Startup Bootcamp. There are seven startups in the current cohort
    – Meeting many people with similar goals for Hong Kong’s growth, including our mentors, and of course, Steve Forte, CSO at Telerik. That’s where things really moved forward

    Thanks to all who have supported these programs. I’ll blog regularly here.

Get new posts by email A few times per year. Deep dives into tech, unit economics, timing, and more.