Friday, January 30, 2009

Danger of Self-Confidence Bias

Thanks to TechCrunch for publishing an article to amplify my post about biased decisions.

I stumbled upon the post "The Economy According to Mint", and I really do hope that Aaaron Patzer doesnt' really see the economy through the lens of Mint.com data.

First of all, another thanks to Mint for publishing data, not just vague qualitative assessment of the economy but real numbers. Now, here are the reasons that the "Guest Author" should've spent the time doing something more productive.

Inclusive data sources leading to self-confidence bias

1. Let's see.... (http://www.quantcast.com/mint.com)


The traffic through Mint.com just doesn't represent statistically significant set of overall population. They are well-educated, leaning towards male, ages 18+, mid-to-higher income, and no kids. The monthly visitors are around 280K in December with a majority of them being "non-regular, passers-by". (By the way, I am superposing another layer of obscurity by using Quantcast to make assumptions about the real Mint.com users.)

Just to make this clear, the state of economy represented by Mint.com usage is just that. The economy doesn't look as bad for the Mint users. Great for the company... maybe... but doesn't give any data outside its tiny world.


2. I personally am a Mint user, and I like it. I don't particularly love Mint, because it's go some technical glitches in connecting with a number of my other accounts. For example, I have a loan and savings account that several emails to customers service and numerous searches could not resolve. In the end, Mint has a very wrong view of my personal saving/spending/asset/liabilities trend.



I'm not trying to bash against Mint. They have a great product in process towards something significantly meaningful. I believe that they are on the right track towards something. Mint will probably face a moment sooner or later to make critical business decisions based on internally generated data like this. I smell self-confidence bias creeping into the office of Mint. The team at Mint.com, please surprise me!
Reblog this post [with Zemanta]

Thursday, January 29, 2009

How to Make Unbiased Decisions in Startups

A recent article in MIT Sloan Management Review "Decision 2.0: The Power of Collective Intelligence" presents a great decision framework.

In a world of startups, especially those backed by VCs, there should be a striking balance between agents' (entrepreneurs) decision making based on emotions and intuition and principals' (investors) ability to take the emotion out of the equation to exert "unbiased" advice to the company. On a second thought, emotion and biases almost ALWAYS exist in any decision making processes.

As related to a startup, the whole premise of venture financing is "Here's the money. grow your business. It's okay to make mistakes, but make as few as possible and move on quickly if there are flaws". So, why am I saying this?

STARTUPS MUST MAKE DECISIONS WITH PRUDENCE AND HAVE UNBIASED DECISION MAKING PROCESS IN THEIR DNA. Here are some of the pitfalls, as mentioned in the article, as related to venture business.

  1. Self-serving bias (seeks to confirm assuptions): Startups often deal with a very few or no information because of the nature of the business. Entrepreneurs usually make assumptions about markets, customers, financials, HR, etc. and searches for data to confirm this. If the assumption is wrong, they are just pushing themselves further away from the "right" answer. Ever had a situation where a cool product feature is exactly what you wanted and everyone else you Googled to confirm the assumption?
  2. Social interference (influenced by others): Yes, VCs follow the trend and invest in the wave of things. Entrepreneurs need to balance this out in a way that's contrarian to the society. When everyone thought search was a mature business with multiple dominant players in the late 1990s, Google made its way through, by being a contrarian.... sort of.
  3. Availability bias (satisfied with an easy solution): This is a hard problem. We entrepreneurs think anything is pretty much possible. This one is really tricky because there might be some things that are worth time and effort of doing despite the complexity. On the other hand, there are quick and easy stuff that can get done easily. I don't know what to say about this much beyond than that.
  4. Self-confidence bias (believes prematurely to have found the solution): I actually think this helps startups. Assuming that this kind of decision get reiterated many times in a short time period (like agile development process), it could be a tremendous help. Entrepreneurs, by nature and definition, are full of self-confidence. Just make sure that you know this may not be THE solution.
  5. Anchoring (explores in the vicinity of an anchor): Building personal relationship and leveraging people's knowledge in your network is often used by startups. Remember, the relationship you've built may be self-selecting. You like them. They like you. Both influence each other.... in a cyclical way. Knowledge and influence cycle through the same system infinitely without seeing the outside world. Explore more. Be ready to talk with random people to hear second thought.
  6. Belief perseverance (keeps believing despite contrary evidence): Being a contrarian and out-of-wack essentially are the same thing, depending on the outcome. Please, please, don't keep believing if a majority of people who (might) care about your product say "no". I've seen a plenty of people who keep on saying that they are right and will change the world and future will make the judgment call. If you (or people you know) fall into this category, please, please say that in the future. We got better things to do now.
This is just a partial list of decision making "frameworks" from the article. Good read, overall, so I encourage all of you to do the same exercise I just did.

Lastly, be unbiased and fair. Don't let the biased blind get in your way.
Reblog this post [with Zemanta]

Tuesday, January 27, 2009

Asking the right questions - Using Negative to Draw on Customer Insights


Let's just face it. When it comes to innovations and improving products/services, customers just do not give much valuable feedback. Wait, maybe it's an extreme position. So, let me just rephrase it: "Customers do not usually give tips on product improvements. They typically complain about a bunch of things, and companies make incremental changes to satisfy the customer needs".

When sending out a survey to gauge customer satisfaction, interest, and what not, it usually goes like this....

  • "What feature did you like the most about our product?"
  • "How much are you willing to pay for our product?"
  • "How satisfied are you?"
  • "What can we do to improve our service?"
Nothing wrong with the above questions, except..... these questions just don't provoke the innermost thoughts.

Instead, try using negative tone to give a twist, like...
  • "Why would you not use our product?"
  • "How didn't we serve your needs?"
It all gets down to people psychology where the respondents to the exact same questions with different tone can uncover a lot of interesting information. No, I'm not a psychologist, but I did a fair amount of this kind of interviews recently and wish somebody had told me this in the past.

Try this, and tell me some interesting stories.

Reblog this post [with Zemanta]

Thursday, January 15, 2009

Why get along with business people?


A typical topic that comes up in the startup economy is the need for business people. (Disclaimer: yes, I am a engineer-turned-MBA) For many small businesses, including technology startups, this is the general mentality. Yet, the consequence of building a non-business-minded startup can be enormous (in a terrible way).

Case: Two computer science PhD candidates are supervised by a professor at a world-renowned university. One day, they run into a "a-ha" moment and sees a glimpse of potential commercialization opportunity. They go full force into product development to turn technology into a usable product. It's a web business, and the mantra is, "people will follow good products". They are busy coding away. An eager MBA students steps into their garage office and give a spill on sales/marketing/finance. The co-founders say they are busy, because they are building something, and the business dude will be needed later. They politely say no to the MBA student but asks if they know any investors. Yes, cash speaks in any situations.

This is just another story of very early-stage startups. Nothing unusual about this.

Few years later: The company didn't get any money from investors. It just wasn't a VC deal, and the company's market size was just too small. It sounded and smelt like a science project than a real high-growth startup. The good news: they survived! The company increased headcount, CFBE happened, and now it's profitable. Rather than what the original business plan called for, the company generated much of the revenue from consulting. Life seems good enough, but the founders feel like there's gotta be something else to be done to grow the business. Then, they pull out the business cards that they got from the MBA student years ago to talk some strategy/marketing/finance. The MBA comes in and wastes time trying to fix things that already work pretty well.

Why am I saying all this? It's too late for them to bring in a business person. The company had already grown into a specific culture with missions that are quite difficult to change when things are profitable. The market perception of company is just that - a bunch of smart consultants. Regardless of how smart the people are, the "jazziness" of business should have begun right at the inception of company.

Takeaways:
  • "Business stuff" needs to be done at every stage of company. It can start as a simple blog (marketing/communication).
  • If you find yourself entrenched in unfavorable market perception as a small science project company, it's just too hard to change that. Stick with it, or leave.
Lastly, do you like the Lamborghini in this post? Wanna own one? Surprise: it's a replica made up with decent car parts. Remember: this is something that your business people might be able create for your company.

Monday, January 5, 2009

VC Fundraising: CEO's role

I recently ran into a whole list of questions around how VCs view on CEO's role in fundraising. First of all, raising VC money without having a dedicated CEO is a real challenge in and of itself. For now, staying away from that question, here are my thoughts that all entrepreneurs in this situation should consider to evaluate the merits and tradeoffs.

If VC still wants to invest......
  • valuation will be driven down due to the additional risks of hiring the top leadership
  • option pool needs to be carefully assessed to have something attractive enough for the incoming CEO
  • VCs want to keep the founders around. This is something that investors insist regardless of CEO's peresence.
  • VC will make recommendations (sometimes forces the company) to fill the CEO position. While this may be a big win for startup, hiring someone who fits into the culture, team dynamics, etc. can be a drag.
  • Attractiveness of the overall investment opportunity (this has to do with valuation also) deteriorates. Remember... VCs invest in A team with B idea. A team usually means quality, but it also means having the whole team in place.
  • The incoming CEO may have a different view of the business than the founders... for better or worse.
  • VC's investment thesis is very different now. It probably looks something like.. "great idea with major risks in execution. might pull the plug sooner".
VC will not want to invest, because......
  • hiring a good CEO to execute within the existing culture and idea is too difficult. It often means trying to make a hire and burn cash at the same time. In the end, the company has a smaller case reserve for the CEO to execute.
  • in this economy, finding a real talent to drop into startup is challenging.
  • it's hard to justify the economic benefits of putting their own money in to find someone that the company needs.
Either way, raising an institutional VC round with a CEO is generally a well-accepted practice. Find and convince a solid business person with vision, sales skills, and leadership to buy into your company. Have him/her be part of your team, then go out to raise money. You will save lots of time taking this step than finding a high-risk taking investor to go through the turbulent ride of CEO hiring.
Reblog this post [with Zemanta]