Quirky: Crowdsourcing Innovation

Quirky tried to crowdsource innovation, but fell victim to 2 fatal flaws #busmodel #opsmodel #failcity


Quirky was founded in 2009 in New York City with a mission to systematize the process of innovation, and recently filed for Chapter 11 bankruptcy (October 2015), abruptly laying off its work force of nearly 400 employees. Although various defects in its business model ultimately proved that it was an ineffective company, an analysis of its business and operating model over the past few years shows that such a conclusion is rather simplistic.

Business Model

Quirky’s business model incentivized the public to create a massive pipeline of innovative products solving everyday problems (received thousands of new product submissions per week), and use the power of crowdsourcing to refine the designs of promising products and filter out the less promising concepts. Quirky paid for this pipeline of crowd-sourced innovation by pledging future royalty rights to users that submitted and contributed to product designs. Quirky then selected the most popular product ideas from its website and tasked an internal team to manufacture and distribute the products to nationwide retailers. As revenue was generated, Quirky then paid its community a percentage of sales based on the level of their contribution. This business model had many apparent advantages in terms of value creation and value capture. In terms of value creation, their R&D process was essentially automated. With no upfront expenditures, Quirky received a steady pipeline of product ideas, that were filtered based on community popularity, with the underlying belief that popularity during the R&D process would translate to sales popularity (assuming contributors are also end consumers or at least representative of end consumers). This de-risked Quirky’s later decision to further develop products with proven momentum. In terms of value capture, the fact that Quirky was developing proprietary products prevented competitors from offering similar products and incentivized retailers to carry its new, ‘pre-proven’ products.

Operating Model

Quirky’s operating model had two parts. First, its online platform provided a 2-sided marketplace for contributors of new product designs and consumers of end products (500,000 active users). Quirky facilitated the conversion of new product designs into end products with its team of engineers, designers, marketers, and merchandisers. This model of specialization initially seemed brilliant: the people that buy products tell you what they want to buy, and the people with the skills to create products execute and give the consumers what they demand. In order to grow its operations, Quirky received $185M in funding from top venture capital firms (Andreesen Horowitz, Kleiner Perkins Caufield & Byers), and secured partnerships to crowdsource innovation for General Electric, Amazon, Mattel, Harman, and PepsiCo. From the outside, Quirky looked like it could soon become a household name.

Why it Failed

So where did it go so wrong?

There were two major flaws that became apparent throughout Quirky’s six-year sprint. First, the business model was based on the flawed assumption that crowd-sourced innovation inherently yields products that will be in high demand by end-users (thus de-risking production for Quirky). In reality, Quirky’s model was truly “2-sided”: the contributors were not representative of end-users, and thus were not reliable bellwethers of what end-users (predominantly shoppers at large retailers) would ultimately buy. This was partially due to the incentive system in place that rewarded contributors through royalty payments. For individual contributors, the bar for ‘success’ (e.g., making a few thousand bucks as a hobby) was much lower than the bar Quirky required to maintain a profitable company. As a result, the ideas being submitted were usually novelty items / trinkets (bendable power strip, water fountain for pets, fogless mirror, etc.) that you would more likely see in SkyMall, rather than in the portfolio of a Fortune 500 company.

The operational model was ineffective in correcting for this flaw in the business model. Given the low quality of contributed product ideas, Quirky should have refined both the top and bottom of the product funnel. At the top of the funnel, Quirky should have set guidelines as to characteristics it looks for in successful products (industry, market size, target product cost, etc.) to focus user submissions and thus increase the quality of product ideas. Towards the bottom of the funnel, Quirky should have taken a more active role in screening ideas and receiving end-user (and/or retailer) feedback to get a better idea of which products would likely be in high demand. This would have prevented Quirky from regularly spending $1M developing products that resulted in very little revenue.


  • https://www.crunchbase.com/organization/quirky#/entity
  • https://www.quirky.com/invent
  • http://www.inc.com/christine-lagorio/what-happened-to-quirky.html
  • https://en.wikipedia.org/wiki/Quirky


Zara: Upscale, on-demand fashion


HBS: Creating Human, Intellectual, and Financial Capital

Student comments on Quirky: Crowdsourcing Innovation

  1. Finally, a loser!

    It was pretty shocking to me, too, when Quirky failed this summer because it did seem like a great concept. To me, it seems like the first portion of the Quirky process (the idea submission of new products via inventors/contributors) is still valid. I agree with you; the issue seems to be in choosing which products to engineer and manufacture and then actually pulling off the merchandising and sales process. Did Quirky even do market research to see if their doggie water fountain would be accepted by consumers, or did they just trust that contributors’ votes would translate into sales?

    I wonder if other companies with large R&D departments can learn from the first portion of the Quirky process, somehow leveraging loyal users’ ideas and suggestions to iterate on their products, instead of developing everything in-house in the early stages of R&D. Even if they do crowdsource brainstorming, it should be up to those companies to determine which product innovation ideas will lead to increased sales & profit.

  2. Nice post! I think analyzing failures is much harder than successes–there is generally less public information available and fewer cases studies are written on them.

    I wholeheartedly agree with your analysis that the contributors to the Quirky platform were not representative of end users. Thinking about the type of people who would contribute to this platform, I am reminded of the diffusion chasm separating the innovators and early adopters from the majority of the mass market. It is interesting, however, that Threadless had a similar model for R&D yet appeared to have fared better. Is it just that Quirky scaled up too aggressively under the weight of investor expectations? Would be interested to hear your thoughts on what might have accounted for the different outcomes. Also, do you think Quirky could have found a way around this problem by, say, taking advanced orders in lieu of votes for ranking potential product designs (similar to other crowdsourcing models like Kickstarter)? Perhaps by asking their community to put down their money where their mouth is, they might have found their way around the signally problem…just a thought.

  3. Thank you for sharing Brad! I’ve been thinking along the same lines as TheOtherYiCai and I think that the success of Threadless lies in the simplicity of their product. They designed and sold t-shirts, built a community around it and did that very well. The lack of product focus at Quirky probably contributed to their demise. They competed in multiple product categories and multiple levels, all with broader consumers having limited brand awareness. Maybe if they had chosen to focus their efforts on fewer categories and re-iterated on successful products, they would have been better off.

  4. Brad, thank you for taking up the challenge of analyzing a TOM loser. I learned a lot from your post.
    Naturally, it is easier to analyze in retrospect why a certain business failed, than predicting the unfavorable outcome before it actually happens. But your post got me thinking – how can I see that a business is going in the wrong direction, and how can I correct it in real time?

    This brings me to This Week in Startups’ interview with Jeff Weiner, Linkedin’s CEO, on ‘The 5 Markers of Great Product’ (http://thisweekinstartups.com/jeff-weiner-launch-festival/). According to Weiner, great companies are those who do at least 1 thing REALLY well. Think about how Google started: one search box over a clean-white background. There was only 1 feature, and they did it REALLY well. Another example: Microsoft and Windows in its early days. Quirky, to the contrary, was all over the place. They made 50 products each year, and none of them was that great. And the reason was, that instead of implementing feedback and iterating until they became REALLY great, they simply moved on to the next product. So they never became great at anything. As a result – they did not have core customers, no market expertise was developed, and there was no brand loyalty.

    So my key takeaway is that as managers and entrepreneurs, we need to ask ourselves – what do we do REALLY well? and once we find that thing, we need to build our operating, marketing, finance and business model around that thing, so we can continue to deliver on our promise to do that thing REALLY well.

Leave a comment