The (Kinda) CEO Series: Interview with Julie Wittes Schlack of Communispace
If you have not had the pleasure of getting to know the folks at Communispace then you are missing out. Everyone I have met there is just so exceptional, but none more so than the co-founders; Diane Hessan and Julie Wittes Schlack. I’ve had the privilege of getting to know both of them gradually over the last year or so and every time we chat I come away enriched. That is why I was thrilled when Julie agreed to be interviewed; having the opportunity to expose our readers to one of the true innovators in the market research industry is pretty exciting to me!
I added the “Kinda” into the the title for this entry into the CEO series of interviews because Julie is NOT the CEO (that would be Diane), but she is a co-founder and heads up all of the innovation efforts at Communispace, and is certainly one of the most senior people in the company. I think Diane and Julie are two sides of the same coin, so in my book this interview fits right into the series.
I conducted this interview via email over the course of about 2 weeks. I think you’re really going to like it.
LFM: Hi Julie! It was great getting to spend a bit of time together at the recent Net Gain 6 conference, and thanks for agreeing to chat with me a bit more. At the conference you gave a really interesting presentation on the “Community ecosystem”, meaning the differences between various “types” of community models and how Communispace was doing some things differently from the rest. Since many folks were not lucky enough to be there, can you give a summary of your presentation for everyone?
JWS: Thanks, Lenny. I appreciate the opportunity to talk about this, because there’s a tendency out there to refer interchangeably about social networks, public communities, private MROCs, and even panels. This isn’t just an issue of semantics (though I am a bit of word dork). Rather, panels, networks, and communities differ considerably in their organizing principle (or “center of gravity”) – in how the participants get there and what they want and expect from the experience — and that in turn has implications for how companies can best use each of these approaches.
A panel typically comprises hundreds of thousands of people about whom some basic demographic and psychographic data is known. Panelists are typically thought of as “respondents” because the communication with them is usually one way, and they’re isn’t free to pose counter-questions or interact with other panelists. So with panels, the project – shampoo today, cars tomorrow – is the center of gravity. A variant on this model is the “custom panel” or “communi-panel,” in which a sub-segment of anywhere from 300 to 2,000 panelists are sliced off and given the opportunity to talk to the client company via a basic online discussion board. While panelists are able to interact with one another in the context of these discussions, their opportunities to develop relationships with one another and/or to generate their own discussions are limited by technology and/or by the typically short-term, discontinuous nature of what is still a project-driven nature of the research. This approach makes panels a good forum for feedback from generally less engaged consumers, but they afford very limited opportunities for discovery.
In a social network, the individual is the origin for a web of tight and loose connections. Unlike panels, one-to-one and one-to-many communication is possible in an ongoing way in social networks. Indeed, that’s the whole point of them. But rarely is anyone guiding or moderating the conversation, and the individual – the profile, with its content and privacy settings that determine who can see what – is the center of gravity. Interactions are diffusely dispersed across the site, and while clusters of “friends” may link to each other, conversations do not occur across or within the social network as a whole, and are visible only to accepted friends or followers. They’re also a bear to analyze unless you’ve got a great set of web scraping and text analytics tool, which makes social networks a good venue for monitoring trends and doing some general discovery, but your ability to do deep dives is constrained.
In a community, shared passion and purpose is the organizing principle. Public communities tend to be large, form organically, comprise the most passionate people, and have a minimum of guidance or moderation. In private communities like those we run (also known by the clunky acronym of “MROCs”), the community is relatively small (usually 300-500 people), it’s private and very secure (nobody can self-enroll, and members agree to keep community content proprietary), and participants are recruited based on each client’s specifications, so we can ensure that we’re talking to a very targeted group. In small, private communities like ours, members build relationships with the sponsoring brand and with each other. Content is generally visible to all (though we can make it private and visible just to a selected subgroup), conversation is ongoing, concentrated and iterative, and we practice very high-touch facilitation, interacting with members daily. All of these factors combine to generate very high levels of engagement, which in turn makes the community an ideal forum for both speedy feedback and long-term exploration and discovery.
LFM: That is a great taxonomy of the space; thanks! From a client standpoint looking at those options can be pretty confusing amidst multiple business priorities and objectives. How can we help clients identify not just the right tools for the right job, but also to look at where these pieces fit into a holistic framework for driving impact and ROI?
JWS: Okay, I’m going to be dangerously blunt and improvisational, but here goes. In general, I think research serves three broad purposes – insight (which helps you understand your consumers and their unmet needs, and informs the generation and refinement of concepts, products, and messaging), validation and projection (which hopefully helps you refine concepts and pick winners and losers), and monitoring/tracking (which helps you measure the efficacy of your campaigns and the general health of your brand).
MROCs are primarily insight-generating instruments, and, not surprisingly, I think they have the most potential to have a profound, long-term impact on a business. They enable you to dig deep, iterate, and find the golden tone amidst all the noise. So in grand strokes, the biggest ROI of a community is in getting that breakthrough insight that helps you create an entirely new category, form a relationship with an entirely new market, etc. Shorter term, though, it’s also useful to think about the ROI of MROCs is in terms of cost savings. Because Communispace (and most MROC providers) operate on a subscription model (as opposed to a price-per-project model), the community is a continuous asset that supports a multitude of research needs. That’s means that a company doesn’t have to incur multiple one-off participant recruitment costs, and because the community is able and eager to address multiple questions at once, it quickly pays for itself in the volume of research conducted. Another way to measure the ROI and impact of a community is in terms of speed. Again, it’s a 24 x 7 asset, so if you post a burning question on Sunday night, you’ll have a pretty robust answer by lunchtime on Monday. When you’re trying to assess the impact of a PR disaster, decide whether to recall a product, or simply have a sudden and hugely important top-to-top meeting with a key customer, that kind of speed is invaluable.
Interacting with and polling fans and followers, and monitoring social media are relatively low-cost approaches to monitoring and tracking that yield a good ROI for those concerned with surfacing product problems, managing brand reputation, and keeping an eye on the competition. And of course, as a means of building awareness and driving promotions, social networks have great impact as well.
Conjoint analysis and other very specific techniques for refining and prioritizing product features can certainly have an impact on a business as well. But I think that the main purpose and value of most quantitative, panel-driven research is not to create impact, but to provide a (sometimes illusory) sense of safety, confidence, and predictability in relation to new products and campaigns. So, to the extent that it’s accurate, the ROI is largely in averting costly mistakes, and in optimizing marketing expenditures, and that’s certainly not insignificant.
LFM: Of all of the emerging techniques and technologies that are being discussed within the insight community, which one do you find most interesting and the easiest complement to communities?
JWS: Oh, that’s a tough one, as there’s a wealth to choose from, and I hate choosing. (I used to meltdown at Baskin-Robbins when I was a kid …) So let me pick three. We’ve found mobile ethnography to be incredibly powerful in giving us in-the-moment, in-context, media rich content. People are sharing videos of their kids at play, themselves trying on jeans in the store dressing room, their hotel rooms and how they make them homey (or at least not creepy) – all kinds of stuff that we’d never get through recall or conventional journaling exercises. Emotion analysis – principally through free association and metaphor – has also proven to be really engaging for members, and a source of very actionable insights for us. And, after years of painful experimentation, I’m starting to feel more optimistic that text analytics and semi-automated content coding tools are finally sophisticated enough, not to replace human beings interpreting the voluminous content that our communities generate, but to free them up to attend to the nuance and texture, and find the needles in the haystack that can drive real breakthroughs.
But all of these methodologies depend on highly engaged participants who are willing to share in a very personal, often intimate way. Perfunctory, let-me-get-this-over-with input will yield equally perfunctory output. In contrast, when people feel safe and heard, they’ll go to extraordinarily lengths. And when they’re connected to the brand and each other over time, they have the opportunity to reflect, to share what they thought about in the shower the next day. That’s the unique power of communities, so the tools I’m most excited about are those that require a high level of trust and engagement.
LFM: LOL, 51 methods, huh? I think the three you picked are really interesting, especially how they fit into the community model. Can you share some specific examples of how you’ve used these three extension within the context of a community?
JWS: Sure. We use mobile surveys and ethnographies to get in-the-moment responses that we simply wouldn’t be able to elicit through recall-based surveys, and, perhaps more exciting, to more closely see and hear the lives of our community members. Sometimes that takes the form of a very specific assignment, like having members complete a brief survey every time they use a financial services app, or shop for a specific type of product and show us what they’re seeing on the shelf and how they feel about it. In other cases the work is much more exploratory and open-ended. So, for example, for an agency client in a multinational youth community, we captured incredible videos and audio notes about the brand encounters and cultural experiences that evoked strong feelings over the course of a weekend. For another client, members are simply recording the sounds in their daily lives that make them happy. But in all of these cases, the fact that these projects are an outgrowth of their ongoing community experience makes the participation more robust and the content more personal, and lets us follow up and probe and iterate over time.
As for emotional exploration, we’ve got a proprietary technique called Emotion Centric, that we’ve used for retailers and clients in food, pharma, financial services, and beauty companies to elicit unmet emotional needs in specific categories, to surface sharply differentiated feelings and perceptions about competing brands, and to test concepts and messaging. We’ve also done some work with webcam-enabled facial coding and eye-tracking tools to measure people’s unconscious or pre-conscious emotional responses to video, as well as self-reporting techniques. And again, the power of doing this in a community is that we get explore the interplay between emotional and rational, not just do one-off, highly specialized projects, and that’s important, because in the real world, your limbic system – the emotional center of your brain – is generally narrowing or expanding your decision set, then your frontal cortex – the reasoning core of your brain – is kicking in to make the call about what you’re going to do. There’s a constant interaction going on between these two centers of the brain, and it’s great to be able to tap into both.
Given the large volume of both formal research activities and member-generated content in our communities, it can sometimes be challenging to see both the forest and the trees, to see not just the answers to the questions we’ve asked, but the themes or trends that may just be quietly percolating in the background over the course of weeks or months. Finally, after years of painful struggle, I’m finally starting to believe that text analytics tools are getting both nuanced and speedy enough to provide some real value to us in doing both. We’ve worked with multiple platforms in the past, and are currently experimenting with Decooda – a company whose board I know you serve on. Though it’s early days, what excites me so far is the fact that the system mimics how a human being approaches the task of surfacing both broad themes and individual nuggets from large volume of open-ended text. We start by looking at what’s bubbling up organically from the content, and only after we’ve done that, apply our own classification scheme onto the content for purposes of coding and quantifying, whether it’s an individual survey or everything that’s been said over the course of weeks or months. In that way, we can continually refine our coding schemes, and as importantly, start with what’s actually there in the words, not with what we’re used to looking for.
LFM: OK, since you mentioned it I’m going to be a little self serving here and want to drill into what you’re doing with text analytics and Decooda (full disclosure I am on the Board of Advisers of the company). Taking a step back from the coding uses, how do you see text analytics and/or social media analysis outside of Communispace communities fitting into the insight delivery process? What other ways are you looking at merging data from different sources to deliver a more holistic view to clients?
JWS: We generate an enormous amount of free, unstructured text in our communities, and while we’re perpetually reading, responding, and analyzing this content for our clients, that work still tends to happen on an activity-by-activity basis. So text analytics tools like Decooda have terrific internal application. They help us see the forest as well as the trees, to see what’s organically bubbling up from all of that content — especially in purely member-generated conversations — and point us to themes that we may want to drill into in more depth. I would never want us to rely purely on text analytics tools, but instead, see their value in taking care of a lot of the obvious classifying and counting of verbatim so that we human beings can be freed up to really find the nuance, the unexpected idea, and notice what’s changing over time.
I think that there’s also tremendous value in applying text analytics to call centers and support line databases, as a way to surface problems or unmet needs that can then be further explored in the context of a more free-form, iterative conversation in a community. In the same way, social media analysis can surface emerging product problems, competitive threats, or even ecstatic delight with products and brands – all issues that can be explored in a more directed, reciprocal fashion in a community. And by the same token, when an issue surfaces in a community, a company can use “listening platforms” – whether applied to their own databases or to social media – to see if it’s specific to the community or has broader resonance and should be followed.
We’ve experimented with a variety of web scraping and text analytics tools in the past, with the thought of augmenting what we do in order to provide a more holistic view and address what’s happening at the micro and macro levels. However, we’ve found that most clients already have existing relationships with CRM or”listening platform” providers that they’re quite happy with, and that often that work is funded by a different function within our client organizations. So where we’ve been able to provide most value is to partner with those companies to coordinate our efforts, and more importantly to serve as a synthesizer. Many of our clients are just plain overwhelmed with the volume of data coming in, and what they really need is someone to help them make sense of it all.
LFM: So you’re a year+ into the Omnicom relationship; how has that changed your business and/or your offerings?
JWS: You might get a different answer from our CFO, but from my perspective, our Omnicom relationship had done little to change our business, but is doing increasingly more to inform and support it. In the first year, we went through the usual pains and occasional pleasures of getting our corporate infrastructures more aligned in terms of payroll systems, equipment providers, and so on. It also felt like a speed dating marathon, as we met with many other companies and agencies in the Omnicom family. But now things have quieted down, and we’ve been able to focus on partnering with other Omnicom companies to meet new potential clients, present joint offerings that may be more robust, and grow internationally. Given the size of the company, there’s also just a terrific brain trust, and Omnicom is good at creating opportunities for company leaders to learn from one another. Most important, though, they’ve delivered on their promise to us, which was to grant us the autonomy and freedom that has helped us grow to where we are today.
LFM: The data from the most recent GRIT study seems to indicate that MROCs are now mainstream and will continue to grow. You helped pioneer that model and I know it was heavy going in the early days. What changed? Why do you think many of the adoption barriers have come down?
JWS: Great question! in the early days, approaching prospects with the idea of an ongoing, online private community of consumers or customers who would be not just a respondent pool, but an asset to the company, was a missionary sale. After all, in some respects communities as an insight-generation tool violate some of the old paradigms. People would say that they’re too big to be conventional qualitative research, but too small to be valid for quantitative research. They’d worry that community members would say negative stuff about the sponsoring brand in front of everyone else; they’d also worry that feedback would be positively biased because community members would develop a closer relationship with the brand over time. They’d push us to recruit community members who were typical of their target audience, but fret over the fact that because these people were online, they couldn’t possibly be typical. They’d push back on the price of a community, which was higher than the price of a one-off project, and worry about what they’d do with all those consumers during downtimes when they had no burning business questions to ask.
Then a few visionary early adopters came along, and proved the ROI of communities in a big way. Hallmark quickly recognized that by listening to member-to-member conversation, and not just to the answers to their own questions, they could get a deep and ongoing understanding of their consumers feelings, aspirations, and unmet needs. Colgate analyzed what a community cost them vs the cost of conducting all of that research via traditional means, and realized that it would pay for itself within the first three or four months. Kraft realized that they could engage consumers much earlier in the new product development process and throughout it, engaging in discovery and not just soliciting feedback on already-developed concepts, and through constant iteration with their community, ended up creating the 100 Calorie pack and a whole new product category along with it.
So those early successes helped make the case for the ROI of communities. Then along came MySpace and Facebook and Twitter and TripAdvisor, and helped us to completely reframe the research conversation. What does a “pure” or “unbiased” sample really mean in this environment? What can we learn just by listening to the organic spread of ideas and opinions among consumers? What risks and opportunities are inherent in social media? All of these questions and opportunities have made some of the initial objections and concerns almost irrelevant, and that’s also helped to fuel the adoption of MROCs.
LFM: So what’s next for Communispace? What is the next “big idea” that you’re working on for the next year or two?
JWS: Oh, now you’re making me choose again! Let me narrow it down to two. First, our clients are looking to us to help them connect with consumers in markets all around the world to help fuel their growth. Every month we learn from our new communities in new languages and cultures. How do we move even more quickly and nimbly in bringing our services to emerging markets, via new platforms and business models that deliver the results our clients need? Another thing we’re focusing on as we generate insight using ever more methodologies is how to develop a truly platform-agnostic way to help our clients co-create with their customers by tapping into their whole brains — the emotional and the rational, the unconscious and the conscious — and to make that partnership between companies and customers as seamless as possible.
LFM: Last question Julie. I understand that you are also a writer; what do you write and where can I find some of your work?
JWS: Oh, you’re too kind. Besides writing for the Communispace blog (http://blog.communispace.com), I write book reviews for The Boston Globe (which you can find by searching on my name at www.Boston.com), and short stories and personal essays. Most of the latter have appeared in print-based literary journals, so the best way to find them is by looking on my mother’s coffee table. But if you search at the websites of publications like The South Carolina Review, Ninth Letter, The Dos Passos Review, The Tampa Review, and The Monarch Review (to name a few …. but almost all of them end in “Review”), you can probably find some of my stories and articles that have absolutely nothing to do with social media or market research. Those are my husband’s favorites.
LFM: This has been great Julie; thanks so much for the time and great conversation!
JWS: Thank you Lenny, this has been fun. Take care.