How Understanding Customer Jobs turns Crowdsourcing into Smartsourcing


Share on LinkedIn

Peter Drucker the gurus’ guru famously said, “Because the purpose of business is to create a customer, the business enterprise has two—and only two—basic functions: marketing and innovation. Marketing and innovation produce results; all the rest are costs.” Marketing is hard enough to get right, but innovation is a whole lot harder still. Depending on the industry, about 80% of new products fail on introduction in the market. And up to 60% fail in reintroduction.

To overcome this disastrous failure rate, companies have started to recruit customers to generate ideas for new products. A process Jeff Howe called Crowdsourcing in a 2006 article in Wired magazine. There are a number of great examples of successful crowdsourcing programmes, however, two examples illustrate what happens when companies start crowdsourcing programmes without really thinking them through properly.

The first of these is Dell with its Ideastorm programme. Anyone can come up with a computer-related idea, post it on the Ideastorm website, vote for the best ideas, comment about them and hopefully, see them implemented. Sounds great. Why not harness ideas from customers? And why not get customers to vote for them to cut programme staff costs. Unfortunately, crowdsourcing has a number of serious problems. The first problem is that customers, even large numbers of them, typically produce average, unremarkable, incremental innovations, rather than the step-change innovations that companies hope for. Although 12,483 ideas have been posted on the website since Ideastorm started in February 2007, only 366 have been implemented to-date, a miserly 2.9% of the total. And most of the implemented ideas provide only incremental improvements to Dell’s business. To its credit, Dell says that Ideastorm is intended as an extension of its relationship with its customers, rather than just as a source of product ideas. Just as well, as Ideastorm is a failure as a source of winning new innovations.

The second example is Starbucks with its My Starbucks Idea. Similar to Ideastorm, My Starbucks Idea allows any registered customer to post an idea, vote for the best ideas, comment on them and see them implemented. Or not as the case may be. My Starbucks Idea, despite receiving over 75,653 ideas, has only implemented 315 ideas to-date, an even more miserly 0.4% of the total. You wouldn’t think that having ideas to improve a coffee-house chain would be all that difficult to implement. But the low rate of implementation illustrates the second problem with crowdsourcing; that customers have no idea of how the business works, what business capabilities it has and thus, no idea whether even the simplest of ideas can realistically be implemented, (let alone whether they will turn a profit). In stark contrast to Starbucks, Toyota implements over 1,000,000 employee ideas every year, 95% of them within 10 days of being submitted. But unlike Starbucks’s customers, Toyota’s employees know exactly where the best innovation opportunities lie, what can realistically be implemented and the profit-impact of doing so. Coming up with innovations like this is part of the Toyota Way. It’s what makes Toyota such a unique company.

And there is another big problem too. Customers expend a lot of creative goodwill generating ideas for Dell and Starbucks, only to see the vast majority of them shot down by their peers or ignored by the companies. All that talk by Dell of building a relationship with customers quickly comes to nothing when its customers’ hard work creating ideas is neither recognised nor rewarded. As Dell and Starbucks attempts at crowdsourcing illustrate only too clearly, crowdsourcing’s supposed advantage of harnessing customers as sources of innovation is in fact its biggest weakness. The fact is that most customers simply don’t have any good ideas, those that do are often not implementable and the miniscule implementation rate burns a lot of customer goodwill in the process.

So what should companies who still want to harness customers to generate winning innovations do? How can they turn wasteful crowdsourcing into productive smartsourcing.

One company that got it right is Cisco with its I-Prize competition. Starting in late 2007, Cisco asked innovators to come up with ideas that could it turn into the next billion dollar business. Cisco collected over 1,200 ideas from 2,500 innovators in 104 countries across the globe The ideas were initially filtered to see if they tackled Cisco’s pain paints, if they could be delivered using Cisco’s capabilities and if Cisco could make money from doing so. The filtering was done by a full-time, six-man team, drawn from across Cisco’s business. The best 40 ideas were then assigned a mentor to help the innovators turn their idea into a workable business plan. The final 10 ideas were then selected, and taken though an interview and further filtering process to find the eventual winner. A single idea—for a smart electricity grid—by a German/Russian team was selected to collect the $250,000 prize.

Cisco’s success at smartsourcing shows the importance of focussing ideas on particular pain points or opportunities. Rather than just let innovators come up with ideas, it is much better to focus their creativity on just those opportunities where they can produce a breakthrough. As innovation gurus Tony Ulwick of Strategyn and Prof. Clayton Christensen have shown, in today’s customer-centric business environment this means understanding the jobs customers are trying to do and the outcomes they want from doing them. And not only functional ‘doing’ jobs, but also emotional jobs that describe how the customer feels about what they are doing and social jobs that describe how the customer relates to their peers too. Once you understand customer jobs and outcomes, they should be prioritised to find the areas where the importance to customers is highest but satisfaction with current solutions is lowest. This is the innovation sweet spot. Experience suggests that focussing on the innovation sweet spot can produce an 80% success rate for new products; a whole lot better than the 80% failure rate we commonly see.

Only when you know where the innovation sweet spot is should you seek to harness the creativity of customers. And not just any old customers either. As Dell and Starbucks’ experience with crowdsourcing shows, harvesting ideas from the mass of customers produces a very small number of good ideas and a large volume of poor ones. It also generates a lot of wasteful costs if ideas are to be assessed properly. Recent research on emergent customers suggests that by screening potential customer innovators for their ability to imagine how innovations can be developed that will be successful in the market, the quality of ideas generated is significantly increased. Emergent customers produce much better ideas than the mass of customers, better ideas even than the lead-customers who are alredy pushing products beyond wheree they were designed to go. Just think what being able to identify the best innovators from the broad customer base could mean for companies. Companies could harvest a smaller number of high quality ideas. That would free up resources to help develop the best ideas together with customers. And more ideas would make it successfully to market. It would enable companies to turn inefficient, wasteful crowdsourcing into much more productive smartsourcing.

As the failure of Dell and Starbucks’ crowdsourcing programmes, and the success Cisco’s smartsourcing one shows, understanding customer jobs provides a solid foundation for targeted open innovation, whilst identifying emergent customers provides the best way to harness just those customers whose ideas will help produce winning innovations: new products that help customers get important jobs done well and create profit for the company too.

What do you think? Have you been disappointed by crowdsourcing initiativies? Have you seen great examples of smartsourcing in action? Are you an emergent customer?

Graham Hill
Customer-centric Innovator
Follow me on Twitter

Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

Further Reading:

Jeff Howe, The Rise of Crowdsourcing

Dell, Ideastorm

Starbucks, My Starbucks Idea

Matthew May, Elegant Solutions: Breakthrough Thinking the Toyota Way

Cisco, I-Prize

Harvard Business Review, Inside Cisco’s Search for the Next Big Idea

Tony Ulwick, What is Outcome-driven Innovation?

Hoffman et al, Identifying and Using Emergent Consumers in Developing Radical Innovations

Graham Hill (Dr G)
Business Troubleshooter | Questioning | Thoughtful | Industrious | Opinions my own | Connect with me on LinkedIn


  1. Hi Graham:

    Great article that asks some interesting questions.

    It is amazing to me just how many companies forget about the untapped insights of internal community members.

    An effective “social” strategy really needs to include people both internal and external to an organization.

    Not all companies have the type of solution that can garner vibrant participation by consumers, but all companies leave plenty of untapped staff knowledge on the table.

    Get your community involved in your knowledge base content!

  2. Excellent Graham. You just gave me the last piece in our own puzzle.
    We get a lot of ideas on our new product XeeSM. And I’m very proud that we do. However some of the ideas are based on just random thoughts or sometimes just lack of attention. But it takes time to go through those. Now we are working on the next iteration of the product and here is what we will do:
    1) Allow users who fit a certain profile – basically power users – direct access to our feature request system.
    2) All other users still submit feedback but we will look at those in a different cycle.
    3) Our point system, where you get points for feedback and a lot more points for requests we realized or bugs we fixed, will be adjusted accordingly. Innovators (by points history) may get a higher priority than other (not sure about that yet).

    Anyway I do thank you for your thoughts.


  3. Graham,

    I agree with your statements, but I think that the conclusion could have been drawn a little bit different. I think that the problem with crowdsourcing (and the likely solution) is not to engage everyone to think of anything – rather (and you do say this very well) to focus on those people that know what / how to do it to focus on specifics.

    I think that the Netflix challenge, the X Prize challenge and similar crowdsourcing experiments worked pretty well because they had two things that Dell, Starbucks, and others did not: specificity as to what was needed, and an incentive to determine the value of the work. I am going to go out on a limb and say that most people who contribute to Starbucks are not MBAs or Food Industry experts that are looking to make a change – they are everyday people that don’t know (as you say) what it would take for an idea to succeed. They are doing it because it is fun, they can say they did it, and if their idea is chosen they can gloat about it with their friends. They have on incentive to produce a workable idea, nor a specific focus area (thus, sweeping the retail floor every 20 minutes is right next to the one about serving breakfast food or better food). Same for Dell.

    I think that crowdsourcing is not a great solution, but if you want to make it work you have to use it for a very specific purpose – and with a clear incentive (which, btw, is why Toyota works in addition to being about people who know what / how to improve).

    Still, cannot think of many instances where this could actually produce good results.

    Thanks for the very nice writeup. Great work!

  4. Axel:

    Nice! Very practical tactic for optimizing crowdsourcing.

    The only problem is that some of the best ideas are initiated by people who you may least expect. The ideas may require lots of refinement by “power users”, but they brought in the lump of coal.


  5. Hi Chuck

    Thanks for your comment.

    I think you are right when you talk about the insights that internal staff have, as well as those that external customers have. The two sets of insights complement each other. Companies need to gather both if they are to succeed in innovation, irrespective of whether that is small-i incremental innovation or BIG-I radical innovation. The challenge that companies have is as much knowing which internal staff to involve, as it is knowing which external customers to involve. Fortunately an understanding of social networks allows companies to both identify internal staff that need to be involved and also to identify emergent customers to persuade to get involved.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  6. Hi Axel

    Thanks for your comment.

    You have developed an interesting system of rewarding innovation. As you suggest, not all innovators are equal; some are more equal than others. You seem to have combined two separate strands of open innovation in your programme: Lead-user innovation to allow power-users pushing your product beyond its normal limits to suggest improvements, and a simple suggestions system to allow rank and file users to provide feedback.

    One challenge that you have already identified is ranking the improvement suggestions quickly so that the best ones can be implemented. It will be interesting to see whether offering points for innovations triggers an improvement in the quality of suggestions, or just a wasteful increase in the quantity.

    It might pay to have a look at the evolution of Eli Lilly’s Innocentive programme that pays rewards to innovators for solving the most difficult problems. This might provide some guidance on how to target points at the areas of improvement where you need the most help from outside.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  7. Hi Esteban

    Thanks for your comment. It is always great to hear from you. As is often the case, I agree entirely with your suggestion. But with a subtle twist.

    The Netflix challenge differed from many crude crowdsourcing programmes because it posted a tough job that it couldn’t solve itself. In Netflix’s case, predicting customer behaviour. The evolving roster of winners (one of whom I know from my days at PwC Consulting, when he – Gavin Potter – was the head of Business Dynamics) all used different approaches to solve the given problem. And solve it for Netflix they have.

    This highlights the need to harness external innovators to solve particularly tough problems that you can’t solve yourself internally. Even P&G with tens of thousands of researchers reckons it only has 1 in 200 of the people it requires to solve all its problems. That why it has develop a whole range of programmes to harness different external innovators to help solve different problems.

    Understanding the customer jobs & outcomes that we are having difficulty delivering satisfactorily is critical if we are to focus external innovators where they can help the most. Everyone potentially has a role to play in solving these difficult problems. But we need to understand what the problems are first. Before asking emergent customers to help identify maket-winning solutions.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  8. Graham: thanks for setting the record straight that customer-driven innovation doesn’t always make financial sense in the context of a business model and delivery infrastructure.

    The innovation sweet spot you mention should address two considerations beyond the importance to the customer and the satisfaction with the existing solution. One is the ease of adoption and the other is availability of product. There have been some fabulously wonderful innovations that have failed in the market simply because they were prohibitively expensive or too complicated to adopt (hence no consumer behavioral change), or they simply weren’t available at the right place at the right time.

    Second, I’m curious about your thoughts on customer-driven innovation from existing users versus “emergent” customers versus adjacent market customers. The “power-user” input that Axel describes runs the risk of insulated thinking. (As an example, if GM solicited user-generated innovation for SUV’s, they might have created a vehicle with wider seats, larger cup holders, bigger gas tanks, and more powerful engines–financial troubles notwithstanding. Greater social, environmental, and financial forces aren’t always reflected.)

    Conversely, I’ve read frequently that consumers who aren’t customers are the ones that deserve the greatest attention. For many industries, the greatest growth opportunities come from markets that are tangentially related to the “core” business of an enterprise. From what I read about Emergent Customers, this is a different group.

  9. Hi Andy

    Thanks for your comment. And for your searching questions.

    I agree with you about taking innovations to market. The article focussed on improving the fuzzy front-end of innovation by understanding customer customer jobs & outcomes and by harnessing emergent customers to identify better solutions. The innovation sweet spot represents the business opportunity. Obviously the next stage is to take innovations to market successfully. This requires good product design so that the product is easy to use, good service design so that supporting services are effectively delivered and today, good CEx design so that these all work together as a coherent whole that is greater than the sum of the individual parts. The glue that holds all these things together is the understanding of customer jobs & outcomes. Not only do they drive the identification of the innovation sweet spot, they also provide the foundation for the design process that drives product, service, experience, capability and increasingly, business model design.

    I also agree with you about involving more than just customers in the fuzzy front-end of innovation. The most effective way to do this is to widen the research base to include customers, non-customers and ex-customers when trying to understand customer jobs & outcomes. The subsequent quantitative confirmation and prioritisation will identify whether there are distinct job & outcome-based segments that need different innovative solutions. Different innovation sweet spots. Once you know the innovation sweetspots you should look to emergent customers to help you identify those solutions most likely to be successful in the market. Emergent customers, like lead-users before them, have been shown to produce better solutions than average customers because they are already pushing products to the edge of what is currently possible to solve their own problems. It is likely that they will produce better solutions than average non-customers for the same reason too. All innovators are clearly not created equal.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  10. Hi Graham,

    Thanks for bringing some facts to this discussion. I think there are a lot of excited people out there that want their “ideas” for crowdsourcing to work. The problem is that the online tool really is no different than a shoebox with a slot, or worse, asking a crowd to shout out their ideas all at once. Well, I guess with the online version you can get better feedback that your idea stunk 🙂

    I can’t think of any scenario where an unfocused “mass” of people are going to generate a lot of value. It would really be a lucky stroke for someone to give up a billion dollar idea…or an unlucky one because they wouldn’t be rewarded for it.

    This just gets back to the fact that we already pretty much understand people (in aggregate). Engagement tools like these really just expose that for objective observers.

    Great post.

    Mike Boysen
    Effective CRM Consulting

  11. Hi Mike

    Thanks for your comment.

    The problem with sourcing ideas from crowds is that ideas are unevenly distributed. There is some evidence that ideas are distributed in a long-tailed power-law distribution. That means for every great idea there will be hundreds of merely good ones. And for every good idea there will be hundreds of simply dull ideas. Unless companies use purposeful strategies to find the few customers with great ideas, like looking for emergent customers, most of the ideas produced will be dull. This is the mediocrity of crowds.

    There are scenarios where unfocussed masses provide advantages of selected customers – where you are prevented by regulations from asking customers directly, for example – but they usually require the use of search approaches like ‘random walks’ or ‘simulated annealing’ to search the entire idea fitness landscape for the best ideas.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  12. Nice article, really nice. I wont realize the importance of this issue if last year I was not collecting an ideas of our workers due to program of working process optimization(I was making a social survey as an office manager). You know, when I was talking to every person on their working place during all the day it was 2 or 3 interesting ideas, but none of them were supported by managing director. But then, I noticed, that people do exchange their experience in kitchen. Yup, in kitchen, when they were chewing their dinner!!!For week or more I get more then 20 interesting optimizing ideas, most of them were supported by director. And only couple weeks ago I was looking through some site and read a research paper by Japanese Kaizen(experience exchange) Manager, this practice is wide spread since 60-th in Japan….I was shocked. They got even Kaizen Managers for this, amazing nation!!!

  13. Hi Graham,

    I like your thinking here and agree with the thrust of the article. Your post suggests to me that crowdsourcing is not wrong per se, but that to effectively collect valuable ideas from customers it is necessary to provide them the guidelines as to what constitutes a successful idea. Rather than offering customers a blank slate on which to write, providing them with the same kind of insight that an internal employee has could yield more focused ideas from customers. Of course that’s not always possible but it’s a direction to explore.

    The other question to my mind is whether the implementation rate has anything to do with the quality of the ideas. I haven’t delved into a detailed analysis of either Dell’s suggestions or Starbucks but perhaps you have. It’s clear that simply coming up with ideas isn’t the hard part of implementing them. Having a process to take the ideas, analyze them, filter them, test them and scale them would seem to be far more critical than the sourcing of the idea in the first place (this was the thrust of my post on social innovation networks).

    One aspect that companies should perhaps consider is providing better feedback to customers on the criteria they use for idea selection – by helping their engaged customers understand what it takes to get an idea into production they can refine the thought process through a feedback loop. Providing financial rewards reinforces this process.

    Thanks for a thought-provoking post.


  14. I think a distinction needs to be drawn based on what we are asking customers to do / the role we are asking them to play and the outcome we’re expecting.

    If a company asks the mass of lay customers for the next source of true business innovation or the next “billion dollar idea”, then they deserve the results they’re likely to get. It amazes me that we’re surprised or dissappointed that a large number of ideas from self-interested customers uninvested in the business fail to generate the firm’s next big business model.

    Why not try asking everyday customers to perform the task they’re naturally oriented towards anyway – that being suggesting small, incremental improvements in the customer experience, the service processes, or in ways to close the expectations / performance gaps. (As an aside, I’m guessing that a vast majority of the 1M innovations submitted by Toyota employees fall into that category – and they’re at least doubly incented to do so)

    We make a huge deal about finding the next source of “innovation” while paying almost no attention to finding ways to “do what we do every day – only better”. Which one is more likely to pay the bills this month, this year?

    I’m not saying that we should abandon the notion that real innovation can come from the crowd. But rather than getting dissappointed on the rate of innovation return by the masses, why not use the intelligence we’ve hopefully developed in other areas to segment “the crowd” for experiential lead users capabile of generating (or at least evaluating – not sure we should ask customers to do our development work for us) innovation, separately from those masses capable of generating extremely valuable business process improvements?

    Crowdsourced innovation may or may not work, but until we see some differentiation amongst the approaches, or at least some thought behind matching what we want from the crowd to what their talent & motivation might be, I don’t know we’re prepared to call in the jury with the verdict that the work of the crowd is bound to be mediocre.

  15. Hi Jane Ann

    Apologies for being so slow to get back to you.

    I cannot say enough to support your suggestion about using Kaizen to improve everyday operations. In a previous job I was interim Head of CRM for Toyota Financial Services. There I was taught Kaizen by the masters. And I used it everyday in my team’s work. It produced amazing results. One marketing campaign started out with a 10% response rate. Not bad, but we knew how to make it better using Kaizen. One year and 50 incremental changes to the campaign later, it had a 35% response rate. Customers were happy enough to buy a new vehicle with Toyota financing. Dealers received great leads and support in following them up. And the team at Toyota was really buzzing with confidence.

    Kaizen really works. I am glad you discovered that for yourself.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  16. Hi Nigel

    Thanks for your comment.

    As you point out, crowdsourcing is by no means wrong. It is but one tool with which to fulfil the fuzzy front-end of innovation. The problem is that it is an extraordinarily ineffiecient and ineffective tool most of the time. Who wants a huge pile of run-of-the-mill ideas from customers; with no guarantee that they provide much value to other customers, with no guarantee that the business can implement them effectively and with no guarantee that they will create value for the company?

    As you also point out, even if customers do provide great ideas, many companies don’t have a proper innovation process (or if they do it is protected by armies of strange looking men in white coats with clipboards in the R&D department). Without a proper innovation process, many companies struggle to develop, test, pilot and implement innovations.

    The recipe for success with crowdsourcing is not all that difficult: Firstly, companies need to recognise in which circumstances they should consult the crowd and in which circumstances other open or closed innovation approaches work best. If they want to use crowdsourcing, secondly, they should provide guidance to customers about the problems they need help in solving. Thirdly, they need an efficient process to sort the few good ideas from the many poor ones. And finally crowdsourcing needs to be plugged-in to the company’s open innovation process.

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  17. Hi Chris

    Thanks for your comment.

    You suggestion to harness crowds for continuous improvement is potentially a great idea. But like many great ideas the devil is in the detail. Having been heavily involved in Kaizen (in a previous job at Toyota Financial Services), I can vouch for its effectiveness at driving continuous improvement. We more than tripled the response to some marketing campaigns by using Kaizen. But that was WITHIN Toyota.

    There is still considerable work to be done to work out how we can make a company’s walls more transparent to customers; so that they can see where the company needs help. And more porous to customers; so that they can actually help. Von Hippel’s early work on Lead-User Innovation provided a starting point which Chesbrough’s more recent work on Open Innovation extended. Howe’s work on Crowdsourcing has done something similar in parallel. But this isn’t the same as having a robust framework to guide us where, when and how to involve customers. As partners in the co-creation of value.

    Any suggestions?

    Graham Hill
    Customer-centric Innovator
    Follow me on Twitter

    Interested in Customer Driven Innovation? Join the Customer Driven Innovation groups on LinkedIn or Facebook to learn more.

  18. There is still considerable work to be done to work out how we can make a company’s walls more transparent to customers; so that they can see where the company needs help. And more porous to customers; so that they can actually help. Von Hippel’s early work on Lead-User Innovation provided a starting point which Chesbrough’s more recent work on Open Innovation extended. Howe’s work on Crowdsourcing has done something similar in parallel. But this isn’t the same as having a robust framework to guide us where, when and how to involve customers. As partners in the co-creation of value.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here