Crowdsourcing our application process; Feedback needed

One of the things I am very disappointed with is that I know so little about so many markets and their opportunities.

As a selection team (we have 4 internal Microsoft and 6 external members – VC’s, angel investors and experts) we know more, but still our experiences are fairly limited compared to the entrepreneurs who apply.

We have so many applicants that are much more qualified than us, that it seems like we are under qualified to make these decisions on which companies to help and which ones to fund.

I have an idea for the next batch but I’d like some feedback.

Is it better to take a “wisdom of crowds” approach as well?

Should we take say 2-3 “slots” and provide the list of companies with an overview and plan and a selected list of say 100 people vote (with justification) on which ideas and companies should be funded?

Do you think it will work?

I dont want to give all the slots up for vote since there are some companies that are of strategic value to us.

One thing that it helps is avoid many more “Why did anyone fund this company” kinds of questions.

Please let me know your feedback.

Advertisements

13 thoughts on “Crowdsourcing our application process; Feedback needed”

  1. Yes, Keeping 20%-30% of the slots open for a kickstarter like process would give a chance, to let the crowd vote on the most promising start-ups. The key in getting such a system of predictive voting right is appropriate mix of diverse & passionate set of voters, whose incentive is alined to selecting the best start-ups.

  2. If you are funding companies beyond your strategic interests what are the underlying objectives in doing so? IMHO, crowd sourcing isn’t a bad idea but the underlying objectives will help in determining whether to go with “wisdom of crowds” approach for funding such diverse opportunities.

    1. Say your objective is to fund projects that does give back to India or the Indian society, then crowd sourcing helps in identifying and prioritizing sectors that need more immediate attention, keeping the underlying objectives in mind. The next step is to identify start ups with potential in these areas and evaluate them by involving both the experts and the crowd, correlating their opinions and finding the most appropriate ones to fund.

  3. Crowdsourcing the voting directly might generate loads of fake votes I guess. I would rather suggest crowdsourcing the problem being solved by an applicant and the domain he is targeting. Say an applicant would like to develop a travel platform connecting tourists to free lance local tour guides rather than letting out the idea for vote it would be nice if you can reform the whole application and come up with a question that will actually evaluate the pain point. Say “How do you explore new places?”, “Do you hire tour guide when you travel?” and “How do you select your tour guide?” etc..

    I am no expert myself but it is just a suggestion from my end 🙂

    1. wow crowdsourcing the problem is way to go for startups to understand more better and give funding VC’s get more insight into why they have to fund.. at the end of story you are solving a problem and chunks of solutions make the problem look very easy.. Gud I would go for crowdsourcing the solution for the problem..

  4. Aligning the incentives of the voters with that of the startups AND the transparency of the process would be the absolute deciding factor of this approach’s effectiveness. But this sounds like a good experiment to run for sure!

  5. It’s a great idea Mukund. IMO, It will work if 100 folks you select to cast votes have an incentive to participate in the process as it will they will need to invest some significant time and initiative to pick the truly valuable startups. I think that’s what Praveen above said as well but I wasn’t sure

  6. This is an interesting (and somewhat ‘sexy’) idea, I wonder though if it would yield the same result, if you compare it to a random effects model (which according to some experiments, yielded better results in the stock market compared with investment advisors).

  7. Wisdom of the Crowd only works if each participant has something to gain or loose on the outcome, and if they can size their bets according to their level of conviction. Intrade prediction markets work because the ignorant don’t vote, the semi-knowledgeable make small bets, and the knowledgeable make big bets. All this aggregates into a decent estimation of outcome.

    Without skin in the game, all you get is noise.

    That said, there could be a way to give people skin in the game without money, using reputation, incentives, or some sort of disincentives instead of money. (Imagine if you made the right call all your FB friends would be messaged that you made the right call, but they would also be messaged if you made the wrong call. Something along those lines…)

  8. Yes, @witolddc that is correct. The incentives could be, the folks who voted for the winning teams be featured as advisers and commit 2 hours per week of their time, i.e. they get a chance to build relationships with the winning companies. Also each of the 100 members pool in Rs 10,000 to Rs 100K and are allowed to buy virtual stocks in the companies, so after two year this amount will be given back to the share holders, depending on best performing companies after 2 years.

    So, as a voter, I would have my incentives clear to pick a company that I would think would do well after 2 years. So this aggregations of deep knowledge of the companies and their future by the group will lead to an optimum selection. We also need to figure out a way to remove bias in the system, which would come from a small voting group.

  9. I agree with Biswa.
    As you mentioned, you find a lot of applicants who are more qualified than you, then I think you can give them the ability to vote for someone, apart from themselves of course.

    How this would help as well, is that the applicants would then start getting in touch with each other trying to get each others votes. Partnerships can be built from here.
    Depending on what kind of companies have applied, you can have applicants already forming business partners.
    I think this might also help the applicants learn the art of partnership building. 🙂

  10. Software product development is more like making movies. (Some directors “know” what will be a hit, but most make movies hoping it will become one.) From the outside, predicting a hit is just a gamble for those who try to do it. But those who actually predict, do it from their ‘Gut feel”. (Gut feel formed from their own experience, exposure and ability to connect the dots.)

    Similarly, no matter how qualified others may be, your own assessment based on your internal criteria is a much better bet than crowdsourcing.

Comments are closed.