Can innovation truly be democratized and how?

Quirky was launched in 2009 by Ben Kaufman to democratize innovation – anybody with a good idea could bring a product to market with Quirky’s support. Quirky was backed by VC powerhouses  such as Andreessen Horowitz and Kleiner Perkins Caufield & Byers as well as GE, and raised over $185 million. However, on September 22, 2015, the company filed for Chapter 11 bankruptcy as it ran out money. I’d like to take a look at what caused the demise of such a promising and innovative startup and offer suggestions for version 2.0.

How did Quirky work?

It was simple:

1.       You submit a hardware idea; it could be developed independently, or jointly with Quirky community members with various expertise who would later share in the product revenue

2.       Community members and Quirky employees would then discuss and vote weekly at a live product evaluation session called Eval

3.       Quirky would then pitch winning products to their brand partners like GE,  Mattel, Harman, Poppy or manufacture the products themselves

4.       Quirky then sold the products to big box retailers and online

5.       The innovator earned commission on each item sold

At the time of the bankruptcy, Quirky had over 1 million community members, was receiving 4,000 ideas a week, had developed more than 400 products and was in 50,000 retail stores.

So what went wrong?

A pseudo multi-sided platform that was overly concentrated on one stakeholder: Quirky’s platform comprised of 3 main groups – community members, brand partners and big box retailers. Quirky did a great job creating a large community of innovators and innovation influencers – 1 .15 million strong! However, it failed on the brand partner and retail side. Quirky had only a handful of large brand partners who picked some (likely a minority!) of innovations to manufacture. Quirky therefore had to bear the high cost of manufacturing and distributing several products with its own capital. It also failed in its selection of retail partners. It targeted only big box chain retailers who had strong buying power/leverage and so offered very low margins (retailers wanted 60-70% margin) and required large volumes to stock all their brick and mortar stores. Therefore, Quirky was unable to test concepts in stores in small volumes. It is also worth noting that some large retailers, like Walmart, are able to return unsold inventory to manufacturers, which could prove costly for failed Quirky products.

No strategic focus/market segmentation: Quirky manufactured everything – from a smart air conditioner selling for $350 to a $3.00 citrus spritzer.  It was focused on bringing any and all good ideas to market at light speed. As a result, there was no strong brand identity. This negatively impacted high value items as customers were not willing to pay the price for an unknown brand. Quirky attempted to create several sub-brands, but failed in getting buy in from consumers.  Furthermore, Quirky was so focused on volume that it did not spend time iterating and improving on products once launched.

Improvements for version 2.0

Earlier this year, Quirky tried to pivot by a) focusing more on internet connected devices for the home, anchored to its Wink app; and b) positioning itself as a supplier of ideas to large companies like GE and Harman, who will do the manufacturing and distribution. The products will carry the big company brands and the tag line ‘Powered by Quirky’.

However, investors had lost faith in the company and were unwilling to pour in more cash. Wink is now up for sale, with Flextronics as the leading potential buyer. The Quirky brand and community is being sold separately.

I would like to offer the following considerations to anyone who decides to pick up from where Quirky left off.

·       Redefine mission and become an asset lite company: Quirky should serve only as the Uber for ideas; its purpose should be to continue to filter ideas through Eval and then connect stakeholders in the value chain to figure out manufacturing and distribution. In order to ensure quality, manufacturers and retailers should also be rated by users on the platform to facilitate the selection process by innovators.

·       Refine selection of stakeholders and create a true marketplace: Quirky should look to sign up all sizes (small, medium, large) of OEMs/contract manufactures to bid for innovations it promotes. These manufacturers will have the option to purchase the IP outright and/or offer royalty to the innovator while bearing the full cost of manufacturing. However, if an innovator can bear the cost of manufacturing, he can still source a manufacturer on the platform.  On the distribution front, small, medium and big box retailers can then purchase from manufacturers or innovators on the platform. Big box retailers can have the option to bid to manufacture the products themselves or with a manufacturer on the platform as private label at their own cost.

·       Focus on simple, functional, low value items: High value items require significant investment in product development, manufacturing and marketing. It’s best left to companies like Bolt ( who can be more hands on, iterate and achieve better product market fit and provide financing. Being just a supplier of ideas to large companies is also a weak position to be in. GE, Harman could easily create their own platforms to source ideas for high value items. The next iteration should rather focus on low value items like the Quirky flexible power strip ( Very little branding/marketing is required for such products and they could easily be sold online only to reduce distribution cost.


By: Alice Agyiri

read more

Why Does Amazon Mechanical Turk Still Exist?

Amazon Mechanical Turk (MTurk) has provided inspiration for strongly-worded blog posts since its inception in 2005.[1] I will focus here on the specific question of how MTurk will persist in light of dissatisfaction among so many network participants.

Amazon launched MTurk to “crowdsource” Human Intelligence Tasks (HITs) that are relatively easy, or at least possible, for a human to complete, but challenging or impossible for a computer to complete reliably. HITs might include determining whether there is a barber shop in an image or identifying the mood of a song.

More recently, MTurk has played an important role in social science research as a convenient hive of inexpensive survey participants that, according to some studies, are sufficiently representative and reliable.[2]

MTurk is named after an 18th century ruse in which a chess master was hidden inside a contraption–a Mechanical Turk–claimed to be a chess-playing machine.

The structure of MTurk is a classic two-sided network consisting of Requesters and Workers. Requesters set prices for completion of HITs, and Workers may accept HITs after reviewing them. Workers are not penalized for not completing HITs, but receive no compensation for incomplete HITs.

Both Workers and Requesters benefit from the size of the MTurk network. Workers benefit from having many Requesters who provide many HITs from which to choose (increasing demand for labor and availability of work), and Requesters benefit from an abundance of Workers to ensure competition for HITs, driving prices down and increasing speed of HIT completion. MTurk provides value to all parties by monetizing underutilized assets–in this case, human labor–by reducing labor-market friction (transportation, recruitment, etc.) through the efficiency of a technology platform. Just like Airbnb allows someone to monetize their “underutilized” home, MTurk allows Workers to monetize their “underutilized” ability to complete HITs.

This summer, MTurk changed the structure of its commissions (the amount it charges a Requester per HIT) amid complaints from both Workers and Requesters.[3] From the MTurk blog, “These changes will help allow Amazon to continue growing the Amazon Mechanical Turk marketplace and innovating on behalf of our customers.”[4]

The commission increase is one, but not the only, reason why the current structure of MTurk invites competition from upstart competitors. Wages for Workers were very low even before the commission increase, which constrains demand for HITs and drives wages even lower.[5] Further, Workers have little recourse when Requesters provide misleading Task instructions or time estimates. So a competing service that offers more protections could be an attractive alternative. (But wouldn’t Requesters avoid a competing service that protected Workers from abuse? Stay with me.)

Requesters also see diminishing benefits of MTurk. The commission increase directly reduces the ability of social science researchers (an increasing proportion of Requesters)[6] to use MTurk as a tool to collect survey respondents, given that research funds tend to be fixed. Some researchers have decided to reduce the wages they offer, or simply recruit fewer respondents,[7] eroding the quality of MTurk for Workers.

Requesters face another type of challenge in addition to the commission increase: the inherent value of MTurk to social science survey research may be declining. This issue is more fundamentally erosive of the MTurk value proposition than commission increases and less easily remedied (though not impossible to remedy). The quality of MTurk-generated survey research is threatened by the trend toward “super users” who are skillful at answering the same types of questions; social networks such as or reddit where Workers share tips on answers and strategies; and the lack of enforcement of basic standards of social science research that ensure quality data as well as ethical practice.[8]

Other networks are forming that provide some benefits that MTurk does not. Sticky Crowd touts better rates than MTurk; sites like peopleperhour and Upwork focus on more highly skilled tasks. Notably, a popular Requester, Crowdsource, has insourced and now uses its own platform to accomplish the same end as it did with MTurk.

It seems clear that the technology of MTurk is nothing special, at least not in 2015. (That is to say, it is replicable.) MTurk may be the largest player in the space presently, but because multi-homing costs are low for both Requesters and Workers (it is convenient for either type of user to sign up for multiple services), an upstart competitor could lure Workers and Requesters by providing a superior product with better Worker protections and lower commissions. While Worker protections could force Requesters to pay more for HIT completion, it would ultimately result in a stronger, more sustainable network that would benefit both Requesters as well as Workers. Better Worker conditions would attract a more representative and reliable pool of Workers, as well as establishing accountability among all Requesters and rooting out the rotten apples that spoil the bunch.

Perhaps the reality is that survey responses as cheap as MTurk initially provided cannot yield good data in the long run and were merely the result of the platform’s novelty. So, a company that provides an MTurk-type platform at high quality could ultimately be the Facebook to MTurk’s MySpace.

Why, then, does Amazon maintain MTurk in its current format? Perhaps it is at least slightly profitable. Or the strategy is (ostensibly) frugal, if not directly profitable: if Amazon continues to use MTurk to recruit Workers for its own HITs, which was the original reason for MTurk’s existence, it makes sense to open platform for other Requesters to recruit Workers, thereby increasing Amazon the supply of Workers.

But if Amazon does not cultivate and improve the MTurk network to keep both Requesters and Workers happy, the current strategy will prove penny wise and pound foolish, and we may see the day when nobody wants to use Mechanical Turk anymore, leaving Amazon to outsource its own HITs to a competitor.


2. Buhrmester M, Kwang T, and Gosling SD. Amazon’s Mechanical Turk: A New Source of Inexpensive, Yet High-Quality, Data? Perspectives on Psychological Science 6(1) 3–5, 2011.







By: Leo Brown

read more

Profile Pictures

When I first created my Facebook profile, I spent literally hours trying to figure out which picture to use as my profile picture. I’ve changed my profile picture maybe 3 times in the last 7 years. Why? Because the last time I changed it, I spent hours trying to pick the “right” picture.

Specifically, profile pictures on dating websites

This problem (yes, I consider it a problem) is much more prevalent in the online dating space. There are a lot of reasons that I’ve never joined a dating website – pride, time commitment, the whole numbers game. But honestly, one of the big hurdles for me was the thought of picking the right profile pictures. The one picture that makes you look so awesome, that says enough about you, that piques someone’s interest enough to click on your profile and read on.

Most people, whether they admit or not, are at least a little superficial. If you don’t look good in your profile picture, you’re not going to get a lot of results. So you end up spending hours of digging through old Facebook pictures to find the three or four that make you look awesome. And, you know, cropping out your ex from those pictures.

But the worst part of the process is the lack of feedback. “I think I look good in this picture… But what will the hot blonde on the other end of the internet think? Which of these pictures is the one that’s going to convince her that I am awesome and she should click on me? (Or swipe right, for you Tinder users)”.

[If any of my relatives happen to stumble across this, of course by “dating website” I meant “”, and by “hot blonde” I mean “nice Indian girl from Tamil Nadu”. I was just trying to make this post relatable to all these white people.]

An online economy solution?

I wonder then if there is potential for an online service that would solve this problem… Imagine a website that I can go to, upload a few pictures that are candidates for my profile picture, and crowdsource this whole problem… Not like the old where your picture is compared against other users, but rather where your pictures are rated and ranked between your own candidate pictures.

You can imagine that this service could even provide additional functionality like letting me pick my target audience, so only people within my target audience see these pictures and rank/rate/comment on them.

The issue of course is how to incentivize or mobilize the “crowd” part of the equation… What would motivate someone to come to this website over and over, and help people pick the best profile picture for them to use? Monetary rewards aren’t scalable. Maybe partnering with an existing dating website as an added benefit to their users? Or maybe selling the data?

I don’t really have a good solution for this… But surely someone smarter than me can come up with a successful mobilization strategy for such a service and launch it? If you do, you can count on me as a customer. If it’s free, of course.


read more

Crowdsourcing funding for consumer loans has been a reasonably successful space with Lending Club, the leading player in the space has lent over $1B since inception in late 2008. Provisions in the recently passed JOBS Act now allow public solicitation of private companies for fund raising which opens the doors for crowdsourcing of equity funding (not just loans). The development begs the question, will there be a viable equity sourcing platform (single or multiple)? To answer this question, let’s start with looking at the source of success for Lending Club.

The business model of Lending Club is one of classic disintermediation through new media. Consumers are getting maximum 1.5% return CD’s while having to pay interest rates of 14% or higher for unsecured consumer loans with banking costs and margins taking up the spread. Lending club allows consumers to directly lend to each other. For this to work, the platform needs to have large numbers of lenders as well as borrowers and needs a way of providing security to the transactors. Because the value to both sides of the platform is large and obvious, building a network is not difficult from that perspective. The key issue to address here is the risk element, especially for the lenders. Lending club does this by obtaining credit scores and other important information from borrowers and rejecting ~90% of applicants only accepting those with higher credit scores. They then sort the borrowers by risk and assign rates while trying to educate borrowers on default rate statistics. Moreover, they mitigate lender’s risk by unbundling loans and forcing diversification of lenders for each loan.

To replicate this model for equity, we need customers on both sides of the platform. An investor can be anyone looking for another asset class to boost yields and/or to diversity their portfolio, while a typical fund-seeker would likely be a company who might otherwise seek equity from founders’ friends or angel/venture capital type investors. The question arises why this company would want to crowd-source equity funding. An answer could be that they are not a VC investment target and they don’t have friends that can fund them or they don’t want to deal with mixing business and friendships. However, another obvious answer that will strike would-be investors is that the company is either a lemon or run by shrewd owners who want to reduce their funding costs by offering unsophisticated investors a lower stake for a risky investment. While it is reasonably easy to diligence a person’s credit-worthiness, it is much harder to ascribe value to fledgling business. Would the platform take on this time and effort intensive diligence obligation? Would investors be able to trust the platform to do this as easily? Furthermore, the amount education required for investors to understand the implications of an equity investment versus debt is high: how will they have liquidity? What are the true range of outcomes for their investment? Finally, it will be difficult to ensure investors are being treated fairly in liquidity events with likely no personal recourse, one can imagine owners paying themselves handsomely while the companies go bankrupt. In short, there’s a reason the US congress has long had standing laws barring such solicitation. It is difficult to disintermediate the function performed by venture capital companies and the complex role played by personal relationships in equity investments. Hence, despite the proliferation of equity funding platforms, I don’t think space is going big anytime soon.

read more

…diverse applications of crowd-power…

Few months ago I read about a travel search company, FlightFox that had just closed an $800,000 round of funding from Y-Combinator, 500 Startups and few other private investors[1]. The website of FlightFox much as any other travel website claims, “We guarantee the lowest prices for international and multi-city flights”. However the thing that would catch your attention is the preceding line in big font, “Real people compete to find you the best flights.” – this is the USP of FlightFox! Instead of having complex algorithms matching your flight requirements, FlightFox uses the power of crowdsourcing(1) to search for your flight.

This article got me pondering over the power of crowd. Crowdsourcing that has so far been used for raising funds directly from masses, is now beginning to diversify, as in the case of FlightFox. The question is, “Can the power of crowd disrupt search?”

…the future of search…

Danny Hill, a computer scientist who was previously with MacArthur Foundation remarks, “My problem is not finding something. My problem is understanding something.” This would be possible only if search engines understood what a person was really looking for rather than just finding the keywords on a website[2]. My classmate Stephanie also points out in her blog[3], “Current search models assume all words are independent from other words.” She further discusses the recent “microsearch” algorithm that looks at relationships between words. This suggests that search companies are aware of the issue of “natural language processing” and are investing lot of time and resource to improving search algorithms. But is a better algorithm the only way to improve search results?

Before we move ahead I would like to acknowledge that over the past few months, there’s been a fundamental shift in the way I have searched for things over the Internet. Let’s say I was interested in reading about some recent issue. I noticed that instead of starting with Goggle search, I would log on to Twitter(2) and searched for the topic using hashtags; and I was doing all of this inadvertently.

But let’s take a step back and analyze my above behavior. Twitter helped me channelize my interests by allowing me to follow only those profiles that posted articles as per my interest. Increasingly it replaced my morning newspaper, that unlike Twitter was a mish-mash of stuff I wasn’t interested in reading at the first place. Twitter also provided a platform to directly engage with the authors of those articles or people who shared them. Then came the indiscernible day when I started using Twitter instead of Google for searching for topics. I knew that when I looked for topics on twitter, I would instantly see what real people were talking about. I think that is quite important to me vis-à-vis Google search which might reveal similar search results, but is quite static. In a metaphoric sense, the former provides a glimpse into the ‘conversations’ that people are having and the latter is similar to browsing a paper catalog. Given the importance of ‘people’ and ‘context’ in search, in today’s tech-speak this kind of search would fall under the realm of “social search”.

In a crude sense, can social-search be considered crowd-powered search because the crowd had sifted through several articles over the Internet and shared the interesting ones on Twitter? And if so, can we expand its application to search in general?

…crowd-powered search is here…

About 50% of all search queries that people look for over the Internet are questions they ask repeatedly. A recent study[4] conducted by Microsoft and MIT suggests that crowdsourcing or contracting people to identify answers could be used to expand the range of answers that users search for. Companies like Meddik and Blekko are already using this methodology. Both the websites make the search content more relevant to other searchers by allowing users to decide which content is important.

A crowd-powered or crowdsourced search platform would have same issues as a networked platform, where the search relevance depends on the number of people using the platform and the number of people using the platform depend on the search relevance; the classic chicken and egg problem. But that is a separate challenge altogether!

1 For the purpose of this blog, crowdsourcing and crowd-powered are used interchangeably.

2 Twitter holds tweets only from the past seven days.

[1] S. Perez, “Crowdsourced Flight Finder FlightFox Grabs $800K In Angel Funding, Joins YC’s Latest Batch,” [Online]. Available: [Accessed: 22-Sep-2012].

[2] J. Battelle, The Search – How Google and its Rivals Rewrote the Rules of Business and Transformed Our Culture. Portfolio, 2005.

[3] S. Frias, “Is Bing Google’s biggest concern in search?” 28-Sep-2012.

[4] J. Jackson, “ACM CHI: More Search Could Be Crowdsourced.” [Online]. Available: [Accessed: 22-Sep-2012].

read more