Check out what we have to say!
  1. Crowdsourcing for Market Research Part 2: Getting Better Input


    In a post two weeks ago, we talked about how crowdsourcing for market research can avoid some of the inherent biases that can come with traditional research techniques.   There is another reason for why crowdsourcing is being increasingly applied for market research: it can result in better data.

    A common question we hear is “how is the quality of information, ideas and data derived from crowdsourcing better than what you might get from traditional research?”   Here are a few answers:

    More ideas: With a traditional survey, each recipient fills out the questions based on their thinking right then.  Once they have filled out the survey, they usually can’t go back to add additional thoughts that might come to them later.   In addition, since they can’t see other respondents’ replies to the survey (by design), their own thinking isn’t triggered by the thoughts of others.  How many times has a good idea come to you because of something someone else said?    Crowdsourcing provides not only a way to capture ideas both now and later, since most crowdsourcing sites live on for weeks if not months, it also enables the sharing of responses that can trigger more thoughts and ideas.

    Better ideas: With traditional surveys, each respondent puts in their own ideas, and then those ideas are rolled up and analyzed, but at no point is there collaboration that enables the improvement of those ideas.   Sometimes this is desirable and intended, but if you are looking for innovation, what you really want are the best ideas, shaped and enhanced by the collective intelligence, experience and viewpoints of the community.   In some crowdsourcing models, the submitters or “owners” of the ideas can revise and enhance their ideas based on the feedback and comments from the crowd.   In addition, through ranking or voting, you get a relative rating of how the crowd feels about a particular idea relative to the other ideas submitted.   This can result in both better input, and a way to more clearly determine market preference.

    Multi-media input: When was the last time you took a survey that allowed you to upload an image, document, hyperlink, or video to help communicate your idea?  This is becoming standard practice in crowdsourcing both for initial ideation, and increasingly for commenting and suggestions to those ideas.

    Explicit and Implicit Data: When you think of crowdsourcing, you think of the ideas, comments and votes that generally come along with that.  Those are all forms of explicit data, and by themselves can provide superior input for market research than traditional research means as discussed above.   But with crowdsourcing, you can also measure how the crowd interacts with the data itself, which can provide valuable implied insight that traditional research would miss. Implicit data can include things like how often an idea was viewed vs. how often it was given a positive vote.   This is an important way to find ideas that might be superior ideas even though they didn’t get the most votes. (there are lots of reasons why the best idea might not get the most votes, but we’ll hold that for another post).   Analyzing comments to find the frequency of use of certain terms is another piece of implicit data that allows identification of important trends and themes.

    I’ll stop here, but you get the drift.  Include your crowds in a collaborative way for market research, and you’ll likely derive better quality input.

    Up next:  Identifying your best respondents

  2. When Crowdsourcing Goes Wrong

    There are lots of success stories about crowdsourcing out there, but unfortunately, there are also a fair number about crowdsourcing failures.   Recently I’ve read about or heard about some perceived failures, in particular the Pepsi Refresh Challenge and the Mad Men Casting Call, which have motivated me to write about why some crowdsourcing goes wrong.  Naturally, we don’t like to talk about the failures; we’d much rather say all crowdsourcing projects provide benefit…

    There are, of course, many reasons why crowdsourcing can go wrong, but most of the highly visible failures have two things in common:

    1)  “Misguided” purpose:  I’ve found that too frequently the purpose behind crowdsourcing that flops is really about driving marketing and awareness for the sponsoring organization,  rather than on quality outcome. Instead of identifying a need that can be applied for good, the program is just another marketing program for the company.  As a result, less care is taken and more mistakes are made in planning and implementing, and people (participants) get frustrated and angry.

    Do not take the attitude of  “whatever the crowd gives the most votes to will be good enough”.  Care about the result, and don’t allow your site to turn into a simple popularity contest that can be gamed,  such as the Mad Men Casting Call where they belatedly added a recaptcha to stop the use of proxy voting that has drastically skewed the voting.

    2)      Inexperienced Implementation: Failed crowdsourcing sites are often implemented by companies that don’t have much experience with crowdsourcing.  One of the unfortunate things about crowdsourcing is it looks easy to do.  Hey, all you have to do is have people submit ideas and get others to comment and vote right?  How hard can it be?  Well, let me tell you,  some firms are learning the hard way.

    Crowdsourcing done well is a LOT more than ideation, comment and voting.  You must consider crowd management, eliminating bias, ensuring fairness, reporting, providing at least some equality in idea visibility and much, much more.  You must plan for a great user experience so you can avoid problems such as those on the Pepsi Refresh site.  Experience is not an option, unless you want to risk a failure.

    Any web-design firm that hasn’t done crowdsourcing in the past that looks at a crowdsourcing site and tells you “yeah we can build that” is something to be afraid of.  What’s behind the UI is  more important than what you can see.     At a minimum, you should get a consultancy firm that understands things like crowd recruitment and moderation, incentives, reporting and analytics to work with your web-design firm to make sure the implementation is done well, or quite frankly hire a firm like Chaordix, where we do crowdsourcing for a living. Sorry for the blatant pitch, but we’re good at what we do and it pains us to see crowdsourcing failures that could have been breakthroughs.

    So, how to best understand good crowdsourcing practice?   I’d suggest starting off by reading blogs by people like Stefan Lindegaard, Robert Brands, Jason Spector and Andrea Meyer.   Or, join us for a free webinar. And be sure to read our “Eight Principles of Successful Crowdsourcing”  white paper and take a look at the Crowdsourcing Scorecards in our Resources section.

    The good news is with the right planning, purpose and experienced implementation, your crowdsourcing initiative can be one of the success stories.

  3. Why every business MUST care about social media!

    Originally posted at rediff.com on August 16th, 2010.

    Social media is no more a buzzword today. Given the rate at which it is growing and the impact it is making in our everyday lives, we will soon see the answer to this question: ‘Why should I care about social media?’

    Over the last three years, I have closely observed and actively used this medium, as one of the most powerful tools to solve and address multiple business challenges, ranging from hiring an employee to acquiring a new customer.

    In the following pages are my learnings and points of view on why should a business pay attention to and invest in social media.

    Read More…

  4. Moderation – Mandatory for Crowdsourcing Success

    Chaordix at Grow  2010

    Out at the GROW2010 conference in Vancouver (not to be confused with grow events of the horticulture variety), we got to hear from Lane Becker, Co-founder and VP Strategy of Get Satisfaction talked about “well that didn’t work – startup lessons learned.”

    He talked about Adaptive Path, MeasureMap (acquired by Google … Inspired GoogleAnalytics), and Get Satisfaction all with cheery cynicism.

    Get Satisfaction is a peer to us – as Lane described they offer “Customer service communities online – getting customers to engage with and support each other.” Chaordix has a different focus on innovation and insight communities. Our members through crowdsourcing are collaborating with each other, but also with the company personally and via our moderation team. We generate innovation and insight for companies, where Get Satisfaction offloads work from companies, reducing customer support costs.

    Early on we looked at Get Satisfaction and thought “great idea but that won’t work.” Why? Because it’s not a one way input world anymore and people contributing online expect more sometimes useful help from a non-invested stranger. They want connection, appreciation, and a near real-time response from the company on the feedback shared. Participation is the new brand loyalty.

    Ta da! Turns out Get Satisfaction came to same conclusion. Now they bundle in moderation to their service.

    There’s a lot we’ll all discover as online communities mature. At Chaordix we’re working hard to create the human and online expeience to trigger product co-creation, technology or research breakthroughs, open up new markets and predict future opportiny for world-leading brands we work with.

    What do you think human behaviour tells us so far about how people participate and invent online, and what companies find most valuable about customer and other crowd input?

  5. Crowdsourcing for Research Part 1: Getting unbiased results

    With the plethora of market research techniques out there, some people might question the application of crowdsourcing to get information from the market.   What with surveys, panels, focus groups, Neilsen, Ipsos, MyPoints, suggestion boxes, etc. we should be able to get all the input we need, right?  After all, if over 50% of Fortune 500 firms only used focus groups, they’ve gotta be good right?*

    Well, yes and no.  The issue isn’t getting input, it’s getting reliable, accurate, unbiased input that’s most important.  Getting market input isn’t all that hard.   Ensuring that it’s accurate feedback that represents what the market truly wants and being able to assess all of that information to pull out only the most salient information is very hard to do well. And that’s where crowdsourcing differs significantly from traditional research.

    Well designed crowdsourcing campaigns are a far less biased way of gathering and assessing marketing information.

    Let’s take a look at three types of bias often found in traditional research that  crowdsourcing can reduce or eliminate:

    Survey Bias: Generally, unless you go into your research with a pre-determined agenda of building support for an existing hunch, you generally try to avoid bias.  However, like it or not, when you build a set of survey questions, you are building in bias just by what questions you ask and by the way you phrase the question.  With crowdsourcing, you often ask just one question, which is chosen to be open enough to invite creativity yet focused enough to ensure the crowd wisdom is directed to deliver the information sought.  The crowd largely directs the conversation from there, without the survey bias of a list of pre-determined questions.

    Sample Bias: Anyone who has run focus groups understands that when you get eight or ten people in a room around a table, leaders soon emerge that immediate start to bias the discussion and comments.   While you might try hard to pull from different profiles to round out your group, you will ultimate see your groups affected by this factor.  With crowdsourcing, you will typically see hundreds if not thousands more people participating in providing similarly rich data that you might get from focus groups, but without that “in room/cool kids” bias.  (and for the same cost or lower, but that’s another blog post)    Sure, you can reach the same number of people with traditional surveys, but then you get less rich data back, and you expose yourself to survey bias (see above)

    Interpretation Bias: OK, so you do your traditional research – survey, focus group, whatever – and now you have your data. What do you do?  You bring all that data back in and start to sift through it, try to decide what’s most important, and what you should act on.  In doing that, you are subjecting yourself to your own interpretation bias.  Who says what you think is most important is what the market will think is most important?  With a well-planned crowdsourcing approach, you can enlist the crowd to help you vet, assimilate and rank the data.  Get the market to do the work for you and end up with a result that is likely to be more reflective of the market than if you did it yourself.

    So next time you need to do market research, take a look a using crowdsourcing as a way to improve the market-driven accuracy and reliability of your results.

    Next Week: Crowdsourcing for Research Part 2:  Getting better input

    *Adams & Dougherty, Journal of Product Innovation #15