Expanding Crowdsourcing without Getting Overwhelmed

By Scott Kirsner |  June 22, 2015

We’re introducing a new Q&A feature as a way to let members ask questions anonymously and get useful answers from other members. If you’ve got a question, just drop us an e-mail, and let us know if you’d like an answer from a company of a particular size or industry (or even from a specific person or company.) The question below was answered by Imaginatik senior executive Chris Townsend. We invite your comments at the bottom…


We currently hold an annual crowdsourcing event for the technical services division of our company.  Teams work on their projects for 24 hours, then post a 2-3 minute Brainshark video that gets posted in Sharepoint Idea Stream so people throughout our division can vote for the various ideas.  Last year, we received 70 entries and it was really hard to find the time to watch all of those videos (not to mention a bit overwhelming to choose a favorite).  This year, we want to expand the event outside of tech services, but I’m not sure how people could watch even more ideas to vote on them.  How have others addressed this challenge?

— From a VP/Innovation at a $40+ billion insurance firm


That’s a great question. In fact, almost every organization struggles at first with how to scale their ability to crowdsource new ideas. Inevitably, the number of decision-makers (and/or project managers) is limited, and the sheer volume of submissions (whether they’re ideas, insights, concepts, proposals, etc.) quickly overwhelms those few people.

Overcoming this natural volume ceiling constitutes a fundamental step forward in your firm’s innovation capabilities. To cross the capability chasm successfully, you need to make ideation both programmatic and systematic. The good news: it’s eminently achievable. After building some of the capabilities described below, we’ve seen companies comfortably able to collect and manage many thousands of ideas each year.

Let’s start on the “programs” side. If the ideas and/or proposals you’re collecting are always flowing through one set of people for adjudication and decision-making, that’s a clear bottleneck. The best practice, proven many times over, is to build out an innovation program where idea vetting and subsequent decision-making (resource allocation, project management, etc.) can be distributed more widely across the organization. This usually means both a formal structure that aligns with established hierarchical leadership structures in the organization (budgets, managers, execs, etc.) as well as a “practitioner” network of innovation champions and facilitators who can assist leaders in successfully running crowdsourcing initiatives, and who can take on enough legwork to make such initiatives both feasible and repeatable.

There are many well-established best practices for making your innovation efforts increasingly programmatic. Here are a few perennial issues that will become progressively more important as you proceed:

  • Business Objectives. The more clearly you can articulate the purpose and expected outcomes of crowdsourcing initiatives, the better results you will get, both in terms of quality input and ease of prioritizing winners and follow-on actions. For example, you’ll get far better (and more comparable) submissions by asking for “codebase hacks that improve customer wait times by 2x or more” rather than saying “please send us proposals to make our team more efficient.” It seems counterintuitive at first, because you’d think the extra specificity will stifle creativity. But just the opposite happens – the extra constraints actually allow people to be more constructive and creative at the same time. Furthermore, setting clear business objectives forces key execs and stakeholders to get directly involved in helping you drive the effort forward – which is also a pre-requisite for ensuring that proper follow-through will lead to material wins for the business.
  • Champions Network. As you begin to target crowdsourcing initiatives toward various business objectives around the company, the project management and facilitation resources needed to keep those efforts running smoothly is stretched, to the point where you will once again bump against a ceiling on what you can achieve. To solve this problem, you don’t need to hire a large staff, but rather build a strong network of “innovation champions.” This type of approach has been in active use for at least a decade, to great effect across a variety of companies. These networks are usually focused on part-time volunteers who are embedded as process experts in different parts of the business. For more on how Pfizer built a 500-person champion network across the globe in only two years, check out this webinar.
  • Governance structure. As you develop new crowdsourcing efforts toward a range of business objectives, and as you leverage a growing champions network, there will be a concomitant increase in the complexity of decision-making and execution follow-through – until eventually it becomes unmanageable for a small innovation team. When you reach this threshhold, it’s time to create an explicit innovation governance structure (usually at the corporate level) – providing the clarity and functional wherewithal to adjudicate your program’s growing size and depth. This usually involves clear dotted-line connections to executive leadership, corporate strategy, and global operations. For more on how several well-respected corporations have approached innovation governance, watch this webinar about Exelon.Now let’s look at the “systems” dimension. The more ideas you collect, the harder it will be to manually complete the vetting, sorting, clustering, and decision-making required to move swiftly into execution mode. Thus at some point, to achieve scale you will need to implement a purpose-built enterprise software system designed to facilitate the seamless surfacing of ideas and their smooth flow into value-creating innovations.

Imaginatik offers one such system, as do other vendors, and your choice will depend on your organization’s level of ambition and desired footprint. Regardless of what system you choose, there are core considerations that matter to any innovation program looking to achieve scale:

  • Process design. Not all crowdsourcing efforts are the same. They can range from idea challenges to hackathons to communities of practice, and everything in between (for more on Innovation’s 8 Programs of Engagement, see this webinar). These processes vary widely in terms of the type of input (ideas, concepts, proposals, etc.) you’ll receive, and therefore the appropriate mechanisms for figuring out “what to do” with all of that innovation fodder. Different software systems will have varying feature sets for the right reviewing, vetting, and decision-making tools corresponding to each program archetype. You’ll need to know what types of ideation process best fit your organization’s needs, so you can find a software platform that accommodates those programs well.
  • Crowd vetting and voting. One way of easing the decision-making burden is to hand some responsibility over to the crowd. A strong software platform will include multiple forms of crowd-based collaboration and vetting tools to help you sift and sort through the database ideas intelligently. This extends well beyond simple “up/down” voting, and into various types of crowd prioritization, scoring and scorecarding, conjoint-style comparison ratings, and certain types of gamification. This can be a big leap forward, not only in allowing ideas to naturally “rise to the top,” but also in growing enthusiasm and buy-in from the team for the implementation / execution challenges that lie ahead.
  • Algorithm-based analytics. By the time you’re receiving thousands of ideas, even the most well-organized innovation programs struggle to keep up with the volume. Many ideas will be ignored as a result, and your ability to find the gems becomes increasingly haphazard. At this point, advanced analytics becomes critical to the further expansion of your program. Tools based on semantic analysis, text mining, and semi-automated ontologies can help you keep track of emerging and latent value across your entire innovation program, while community-tracking and social-graph-mapping tools ensure you’re aware of who the best innovators are, and how they work best together. If used properly, analytics allow you to make the innovation program operate significantly more effectively.If this feels like a lot to digest (much less implement / launch), that’s because it is. Which is a great testament to the speed with which crowdsourcing best practices have grown over the last 10 to 15 years, and how rich a discipline innovation management has already become.

Nonetheless, moving forward doesn’t need to feel daunting or costly. The above points of advice are fairly broad and comprehensive, and in no way must they be implemented immediately, all at once. Rather, think of them as “development areas” over time. Each plays a role in helping you properly scale a corporate-grade innovation program, and it’s up to you to figure out how fast you can build these capabilities, given resource constraints and level of institutional will. This can be scaled back if budget or headcount issues present significant roadblocks in the short term.

I hope this all proves helpful for you!