Posted: 19 August 2016

Today’s talk will be on customer development and tech risk again. The strong emphasis placed on this shows that this is really important.

Customer development can probably be broadly split into two stages. In stage one, it’s more exploratory. You go out and meet people and ask for advice. In stage two, you try and get some validation for the thing you’re building. Basically, try and sell it! Throughout the entire process, you will meet a lot of people and you will start hearing people talking about the same thing over and over again. That’s a good thing, because you can extract the commonalities and build your product around it. Take EF for example, people reject EF for various reasons (family, stable job, not ready, no confidence to do a start-up), but people accept EF for common reasons (ambitious, want to start a company, etc.). Once again, remember that you just need the 1% who is willing to invest in you.

It’s also important to note that face to face interactions > calls > emails.

So when you meet someone, what do you actually ask them? Points to note would be not to pitch, and bad news is a learning opportunity.

Here are some questions that can guide the thought process in formulating questions for your customer:

  • What is the thing/question that would kill you first?

    For example, people can love your product a lot. “Oh it’s cool”, “This is awesome”, but when you ask them to pay for it, are they willing to? Or in my case, there are huge companies out there who are doing something similar, why can’t the customer engage them? I should expect such questions from the customer and be prepared to give a good reply.

  • Mum-test

    If you ask your mum about your product, she’s definitely going to buy it! How can I ask this person a question to prevent them from bullshitting me? Asking a question like “how’d you like this product?” is probably not that effective. Your mum’s going to say it’s great. Craft your questions such that even your mum won’t say it’s great.

When meeting the customer:

  • Talk about their life instead of your idea.
  • Ask about specifics in the past.
  • Talk less and listen more.
  • Job in the early stage is just to get people comfortable with you.

Once you have some rapport, you can then ask the questions:

  • Do you think it’s a good idea?

    This is a bad question. Opinions are worthless. This is especially true if people aren’t your customers.

  • Would you buy a product which does X?

    This is a bad question. If you’re close to the person you’re asking, it’s very easy for that person to say yes and then never get back to you again. Saying that they will buy it in the future is like a lie.

  • How much would you pay for X?

    This is a bad question. It’s sort of like a “future question”. It’s not like they are going to take out their wallet and pay you immediately. It’s a made up number and it’s not important.

  • Would you pay X for a product that did Y?

    Bad. Still quite a future question.

  • Would you pay X now for a product that did Y?


  • We can have X for you in Y time. Will you pay Z for that now?


  • Why do you bother?

    Good. Probe on underlying motive. For example, if a company is spending millions on something. Why do they even bother? Is it because his boss said so? Is it because it helps his KPIs? Is it because it’s for the better of the organization? Or what?

  • What are the implications of that?

    Great. Find out if the problem matters. For example, “making young bankers life better” or “better excel debugging software” might be a good business idea. Perhaps you’re trying to help the young banker leave at 12mn and come in at 9am instead of leave at 5am and come in at 9am. That helps his health greatly. But who gives a shit about him other than himself and his relatives and friends? Does the person paying care? What are the implications if they don’t buy it? Nothing much to the senior guy. His work still gets done. He still gets his promotion.

    On the contrary, if the implication is something greater, like the senior guy not getting work done or losing money or losing efficiency, then you have a case.

  • What else have you tried? How much did you pay for those?

    A little gray to me. Might be good, or might be bad.

  • When’s the last time you bought from a start-up?

    A little gray to me. Might be good, or might be bad.

  • When’s the last time you bought from an external vendor?

    A little gray to me. Might be good, or might be bad.

  • Talk me through the last time that happened.

    Good question. The person has to recall an anecdote and this is valuable information.

On to the second part. Why take more tech risk?

  1. Your ideas are almost inevitably bad.

    • Every idea you have come up with has been thought of before. So be sure to google and check if the problem you’re trying to solve has a solution!
    • Ask why you came up with this idea - and ask whether anyone else had the right experiences to do so.
    • One of these three things is true:

      -It’s already being done well (bad) -It’s been dismissed by domain experts for domain reasons (iffy) -It’s been dismissed by domain experts for technical reasons (good)

  2. You can avoid bad ideas by taking technical risk.

    Building on above, if it’s dismissed by domain experts for technical reasons, then take a tech risk and solve it!

    • Work on hard problems, so much so that when you have a solution, people are like wow omg I didn’t know that’s possible.
    • New problems in growing markets are good because they have very high option value.
    • Obsessions. They are good because you can see the future and build it.
  3. Don’t be scared of tech risk, because there’s no reason to.

    • Yes, some EF companies have failed because of this. But many don’t.
    • Missing the bar on technical risk isn’t fatal. You might just burn a bridge with a customer.
    • Missing the bar on market risk is
  4. You can’t control external variables, but you can control the tech variables.

    • For example, you can’t predict the market, or where investment capital goes, etc. Try and bring all those variables that you can control into your company. And one of these variables would be hiring great people and building good technology.

Some examples of bad opportunities:

  • Stuff that combines high-level ideas without low-level innovation. Blockchain for housing, blockchain for something, etc. It’s trying to mash up stuff together without really innovating. Something that is good would be like combining high-level ideas and saying “this is now possible because I have made this innovation in this area”.
  • Beware of the “X for Y” seduction. “AirBnB for something”.

I guess stop trying to find a good idea, as it’s much easier to find a hard problem than a “good idea”.

Last but not least, if you’re saying this to yourself now, you should kick yourself:

  • I’m not sure whether I can do this in 6 months.
  • Google/Facebook is doing that.
  • I’ll do that as my next start-up.

In the evening, we had a chat with Teik Guan. His first start-up was mobile payments using Palm Pilot. That was way back in time. His second start-up was a backend payment system thing. His third start-up, which was the one that was successful, wasn’t really started by him. It was by other people and he joined as the CTO. It was mostly system integration work at the beginning. The company got bored of it after awhile and decided to do some R & D. They had the idea of two factor authentication. An example use case was using smart cards to generate one time passwords (for example embed the chip in NRIC). They finished in 2004 and tried to sell to a bank. The bank wanted to do some internet banking at that time for differentiation. The bank wanted to use the 2FA as a marketing tool as well. There was a tender, and he put up a proposal and everything but the deal went to the other vendor, because their existing systems were built by that vendor. He called the VP of the vendor and asked to do the server only, and throw away all the smart card and front end stuff. And the VP agreed!

He then called the bank and told them that the server will be provided by them and the other front end stuff will be delivered by the other vendor. The bank asked for a demo in a week, and they had to basically code for a few days straight. The bank was successful in implementing this system and there was a full page feature on the newspapers. They managed to get traction and MAS was interested in this system and they presented stuff to MAS. MAS found that 2FA is important for internet banking and set these regulations in 2006.

From there on, they secured deals with other banks in region. But how do you do this in local banks (honestly I don’t know why it’s harder to do it for local banks)? They focused on enterprise systems initially (forget passwords). If you forget a password, you can use this system to get an OTP. They were desperate to get DBS as a customer. They found the CIO was going to be a speaker at a certain conference and they bought the very first booth in front. There was only one entrance. They managed to secure a demo to the CIO. They secured the deal eventually and they could leverage on the names of these big banks to get more deals.

They were eventually acquired by Gemalto for 3 reasons:

  • CitiBank Deal. They had a lot of customers.
  • IBM opens doors.
  • A good team (it’s really hard to maintain this because his guys left when they felt that they were not learning anything new).

On another note, Teik Guan also talked about how they fronted a deal and it was a nightmare. In this context, project management was even more important than the technical aspects. Teik Guan also talked about some start-up politics, and how he became the CEO.

Last but not least, start-ups can’t be business as usual. Grow fast or die trying.

Follwing the chat with Teik Guan, we had Tushar from Hubble.

Why Hubble?

  • Investment banking previously.
  • He consulted for this big retailer that had bad balance sheets because Amazon was coming into UK.
  • He saw that there’s potential for space rental to small companies, like start-ups. Trend in commercial real estate.

How did you meet your co-founder (Tom)?

  • As a non-tech person, no one wanted to speak to him.
  • He set up coffees and drinks with other people in the cohort before the program started.
  • Tom didn’t have the best CV, but they just clicked really well.
  • You need to make sure you click with the person.
  • Complementary partners are best!

What was your original idea?

  • Original idea was to automate his job. This totally sounds like the “excel for bankers thing” in a previous post.

How did they execute their idea?

  • Slightly easier for them because their customers are start-ups.
  • Cold emailing.
  • Turning up at receptions.
  • No shame.
  • They went to a company and took all photos of their buildings and then uploaded it on their website. They then approached them and said look, it’s already live! 70% accepted. super important point
  • “Don’t ask for permission, ask for forgiveness”, Richard Branson.

How was it like when they were getting investors?

  • Investors follow the herd.
  • If a big VC pulls out, other VCs will follow suit. Hubble seemed like damaged goods then.
  • They had 3 or 4 weeks of cash left and 10 employees. They had a beer, drowned their sorrows, and they tried to save it. They saved it by really reducing costs and delaying repayments. They saved it, well done!
  • While trying to save it, they didn’t want to fire anyone. Fire one person, productivity goes down. Morale goes down.

When we take a step back and look at Hubble, they are the textbook definition of hockey stick growth.