NetBase: Insight Delivered

Web Marketers who are responsible for running online ad campaigns, with Google AdWords, for example, could benefit from a new way to find highly differentiated, non-obvious keywords for their campaigns.

Semantic Keyword Discovery—A New Category of Tools

I’m creating a new category of tools by calling ConsumerBase a Semantic Keyword Discovery tool to differentiate it from the existing category of Keyword Discovery tools. Keyword Discovery can tell you the search phrases people use to find products and services, as well as the search terms that drive traffic to your competitors. One example is SpyFu. A search analytics company, SpyFu shows the keywords that websites buy on Google AdWords, as well as the keywords that websites are showing up for within search results. The main value proposition is to see or to “spy on” the keywords that competitors use and improve SEM and SEO strategies based on those keywords.

Semantic Keyword Discovery is different because it finds semantically related keywords that we believe will resonate with shoppers in much more interesting ways than some of the Keyword Discovery approaches. Semantic Keyword Discovery isn’t intended to replace standard Keyword Discovery—it can serve as an adjunct, albeit a very valuable one. Standard Keyword Discovery gives you bread and butter keywords while Semantic Keyword Discovery gives you gravy. You need both.

An Example: Unexpected, Relevant Keywords for the Wii

Suppose you were a Web Marketer developing Google AdWords campaigns for the Nintendo Wii. You’d start with regular Keyword Discovery tools from Google, SpyFu, and others to get your bread and butter must-have keywords—the ones you want to advertise on to prevent your competition from doing so. But to get an extra edge, you need to find other related keywords that highly resonate with your consumers—semantically related keywords.

Here’s how you’d do that with ConsumerBase: In a ConsumerBase search on the Wii, you’d discover that Injury is listed in the Dislikes word cloud as something people dislike about the Wii (repetitive stress injury) but, counter-intuitively, you’d also discover it’s in the Likes word cloud as something they like, under the headings of stroke recovery and Wiihab. You’d find that people are using the Wii to improve coordination, for stroke recovery, and to treat Parkinson’s Disease. Drilling further into those results, and filtering the search on the word Injury, you’d learn that Kansas State researchers use the Wii to help restore soldiers’ balance after traumatic brain injuries. (For a quick overview of ConsumerBase and this process, you can watch this YouTube video.) A traditional Keyword Discovery tool wouldn’t uncover this because the relationship between Wii and Parkinson’s Disease can only be found by semantic analysis of consumer chatter in social media.

You could then use those diseases and conditions as keywords in your marketing—an insight your competitors probably don’t have and a means to direct people searching on those keywords to your site.

Comments (0)

While meditating on the train one day recently, I overheard two guys talking. It was a happy disruption because one guy was expressing the pain that I founded my company to address. He was a business development guy working at a Silicon Valley hardware company. He was saying he liked his job, and got to work on cool new technology. But what he didn’t like was that it was one step removed from the customer, which made it difficult for him to know what customers really wanted. In Silicon Valley, finding that out is the job of product managers/market researchers.

Well, that same frustration is exactly what led me to found NetBase! When I was an young engineer at Ariba, I spent a year working on an innovative new product. But the project was cancelled and all the work that my colleagues and I had done was thrown out. It was devastating—all my professional work up to that point was just discarded. I didn’t want to see that happen again because I thought it was bad for the economy and I didn’t like it personally.

So I began researching the process of innovation with the goal of understanding it better and, ideally, finding ways to improve it. I started by examining the Ariba project and came to the conclusion that it had been cancelled because the work wasn’t related to a customer need; it was just an interesting idea. That’s when it occurred to me that the same explanation applies to many failed innovation projects within corporations: They fail when people developing products or services deep within a company aren’t connected to their customers and don’t know what their needs are.

That experience and that conclusion led me to study innovation at MIT and to found NetBase.

So now, using the ConsumerBase tool we’ve developed, business development guys can quickly and easily get a much better understanding of their customers—they don’t need to be in the dark anymore. They can source technologies and strategic partnerships with justification based on real quotes from consumers about what they want.

Going one step further, it’s worth making the point that people who can benefit from using ConsumerBase to understand consumers aren’t limited to market researchers or the biz dev people I mention here. ConsumerBase is for anyone in a company who could do a better job if they could just get a clearer understanding of what customers really want.

Comments (0)

We’re putting together a best practices forum on netnography in collaboration with Rob Kozinets. If you’d like to be notified when it’s up later this summer, leave your email address at http://www.netnography.com. Be sure to tell us what topics you’re interested in discussing!

Comments (0)

I had an hour to kill so I continued my research on the market for photo-sharing sites. My previous netnography of Shutterfly users One Hour Market Research pointed to a lot of complaints users have about the photo print and photo book quality. I didn’t think Flickr even had a printing option, so I expected to see a very different set of issues with Flickr.

It turns out Flickr does have a printing option, but I didn’t find any complaints about it from the user complaints I sampled out of ConsumerBase. A far different picture emerged about Flickr users—the biggest category of complaints was around functionality, which accounted for roughly half the negative comments. Since functionality is a pretty broad category, I’ve broken out the functionality complaints into sub-themes which include such notable areas as compatibility, navigation, and ratings.

When it comes to compatibility, Flickr users had a lot to say about using Flickr with Apple’s iPhone. My research went back a year, so at that point the comments were around Flickr needing to put out an app for iPhone. It looks like Flickr responded—great job Flickr! But then I started seeing some comments about not liking the iPhone app. For instance, one blogger reviewed a bunch of iPhone apps and said “I hate the Flickr app for iPhone. I usually check Flickr using Safari for iPhone” (source). Next up ought to be the iPad I guess.”

The next area Flickr users talked a lot about was ideas and issues they’re having with navigating pictures and organizing them into Flickr’s version of groups or collections. Here’s a good example of what I mean: “I really wish Flickr had more layout customization options. For example, I’d like to go through medium-sized pictures throughout my entire photostream instead of just the first page.” (source) A couple of other users on that site—which happens to be Flickr’s user forum—agreed that this would be a good feature.

I can’t say I understand what it is, but it shows how ConsumerBase allows you to research areas that aren’t familiar to you. The semantic lenses pick up on what people like, dislike, or in the case of the person quoted above, what people wish for. Our lenses literally read and parse what people are talking about and extract enough meaning that their comments can be summarized and searched easily by a market researcher.

Probably the most interesting thing I picked up on, from my perspective anyway, is a string of tweets where people were talking about wanting better ways to rate photos. Apparently Flickr only lets you rate a photo as a favorite so one person on Twitter said “I wish Flickr had a ‘like’ feature or ‘+1’ concept. ‘Favorite’ is too much of a commitment.” (source) Given my long-standing interest in group photo-sharing, I agreed.

I hope you enjoyed this brief netnography on Flickr. Flickr’s a great site, and it has some of the group sharing features I was looking for. If you’d be interested in seeing more of the data from my Flickr study, email me and I’d be happy to share it.

Comments (0)

Our internal documentation expert, Matthew Lindgren, recently produced an in-depth PowerPoint to summarize the results of a netnography on consumers’ perception of diaper shopping at Walmart. (We did this as an example of a netnography on a well-know brand; it wasn’t done at Wal-Mart’s request.) I wanted to post a couple of slides from the presentation to show how effective they are in making it easy for anyone to quickly grasp key results.

The presentation summarizes findings, explains our research approach, defines our social media sources for consumer insights, describes the overall sentiment regarding Walmart diaper shopping, analyzes main opportunities and threats, and offers ideas for potential actions.

Here’s a sample slide from the presentation. The pie chart on the left shows all the negative themes associated with diaper shopping at Walmart, while the chart on the right drills down into the biggest negative, quality, and provides specific information on what consumers don’t like about Walmart-brand diapers.

The benefit of documentation like this is that researchers and marketers aren’t simply left with a general conclusion—shoppers are unhappy about the quality of Walmart-brand diapers. Instead, they can look at highly specific reasons why consumers are unhappy—and can analyze those reasons and develop action plans to remedy the problem.

The slide below provides further insight and information on the quality issue. Note the negative “sound bites” quoted at right—they’re the source material that generates the pie charts, they’re easy to scan, and they exemplify the kind of authentic feedback ConsumerBase gathers from social media.

By the way, a great way to find visuals that reinforce the insights in a netnography presentation is to use Google image search http://images.google.com/. We entered “leaky diaper” and found just what we needed to make our netnography presentation more engaging.

The slide below focuses on the insight that consumers have highly ambivalent feelings toward Walmart as a retailer. In both this slide and the one above, note that after presenting the insight from the netnography, we present Potential Actions the company could take to help remedy the problem.

Thanks again to Matt Lindgren for doing the legwork on this example.

This gives you a glimpse into our style for netnography presentations. What’s yours? Do you incorporate images? How do you present your information visually to make it clear and give it impact?

Comments (0)

I’ve long been interested in the problem of digital photo sharing. Back when I was an MBA student at MIT, another business plan I worked on besides NetBase was for a photo-sharing site enabling groups to share photos after events. See, in b-school there are lots of parties, which couldn’t have been good for our grades—that’s why they call it “B” school.

What would inevitably happen after events is students would clog the school’s email system sending each other pix. Some might try to post the photos to a photo-sharing site, but it was a mess. I imagined others would have this problem. For example, what about weddings? Don’t the bride and groom want all their guests to put their photos in one place? I thought there would be some cool things we could do to preserve photos and memories with a group photo-sharing site. Alas, I had to choose between NetBase and that other idea and thankfully I chose NetBase because I think we’ve created some pretty useful technology.

What brought this issue to mind is that I just returned from a family vacation and was struck again by how hard it is for our family to share the photos. One person says to use Facebook. But the kids in the family aren’t allowed to go on Facebook. Someone else wants to use Flickr; someone else wants to use Shutterfly. Some of us just didn’t even want to be bothered about uploading the photos. A mess. Again. Six years later. I was incredulous.

A Quick Netnography on Shutterfly

So I thought I ought to research the market for photo-sharing sites again. Why hasn’t someone implemented the idea I had in mind? I know I’d buy it. Maybe the biggest issue isn’t in fact the inability for groups to easily share photos. Maybe that’s just a niche problem. Time for some proper market research. Oh, but I’ve got a day job, I don’t have time for market research. Well it just so happens that netnography is the perfect way to do market research on a shoe-string budget, when you need answers fast and don’t have a lot of capacity. Faster, better, cheaper.

So I decided to take a look at Shutterfly, which I really like for making photo albums. I expected to see people complaining about sharing photos after events. Well, turns out I’m wrong. The difficulty of sharing photos after events did not emerge as the top issue for Shutterfly. See below for what did emerge as the top issues. There was pretty much a tie between quality, usability, and functionality. I’ve broken out the quality issue into its various sub-issues, but they mainly had to do with the quality of the prints (flimsy paper, color issues, etc.) and of the albums (bindings fall apart, bad alignment of images, etc.).

Shutterfly—Negative Themes

In my hour of research—remember all this research only took me an hour to do!—I did find one discussion of the difficulty of sharing photos after events. The event type happened to be weddings, just as I had predicted. I decided to join the forum, www.makeupalley.com, to ask the group some questions. Following on Rob Kozinets’ ethical guidelines for participatory netnography, I identified myself as someone doing research and posed some questions to the group about their difficulty sharing photos. The impression I got from posters was that they had gotten by with just creating a Shutterfly account (or the like) and giving the password out to their guests to upload photos. One person complained that it was difficult to get the guests to actually upload the pictures.

But it wasn’t a major problem. Perhaps that explains why nobody else has gone ahead with the idea. Glad I chose to build out NetBase instead To tie this back to the benefits of netnography, sometimes people say that it helps you “fail faster, cheaper.” Indeed, innovation is a game of chance so if you can lower the cost of failure, you can spend more effort on the good leads than the bad ones.

Comments (1)

Interested in business challenges, technologies, and solutions in Content Analytics, where Content Management and Publishing, Search, and Analytics intersect? Then the Smart Content Conference is the place for you.

Seth Grimes, contributing editor for TechWeb’s Intelligent Enterprise and analytics strategist for Alta Plana Corp., just let me know he is organizing Smart Content: The Content Analytics Conference. It’s slated for October 19, 2010 in New York.

The conference is about digital transformation and enhancing the business value of information, both enterprise content and social media. You can learn more at smartcontentconference.com.

Seth is also looking for speakers and is accepting speaking proposals. You’ll find information on the conference website under the Call for Speakers tab, but here’s most of it:

“Presentation proposals from end users, analysts, researchers, and consultants in one of the areas that follow are especially welcome, but we will happily consider all proposals. Topic examples:

• Content management, enterprise search, and findability
• Content tagging and enrichment
• Content targeting including contextual advertising
• Analytics-driven social-media and publishing strategies
• Brand and content tracking, measurement, and optimization
• Competitive and market intelligence
• Knowledge bases and data services
• Sentiment and social-media analysis
• Rich media
• Semantics, Web 3.0, and the Semantic Web
• Taxonomies, ontologies, and knowledge management
• Content analytics technologies: theme and topic extraction, metadata, classification, automated summarization, machine translation, plagiarism detection

Please submit your proposal by Monday, July 26, 2010. We will aim to notify you by July 30 whether your proposal is accepted. Selected speakers— one per presentation—will receive a free conference pass.

If you’re from a solution provider—a start-up or established vendor—please consider submitting a proposal for a lightning talk: a 5-8 minute presentation/demo of your content analytics technology, solution, or service (exact length to be determined based on the number of accepted submissions). Just use the form to tell us who you are and what you propose to demonstrate.”

I think content analytics is a great logical next step for text analytics.  It brings it up a level from technology to business value.  Should be a great conference!  Thank you Seth for all the great events you do!

Comments (0)

The Wall Street Journal recently ran an article entitled “Are the Yankees Truly the Most-Despised Ballclub?” It states that “Contrary to popular belief, the Yankees are only the fifth-most despised team in the majors, according to an Internet algorithm built by Nielsen Co. that analyzes how people feel about certain things.” Here’s the article: http://online.wsj.com/article/SB10001424052748704471204575210384180269378.html

Interesting topic, but I had this reaction:

• How transparent is that method?
• Can you run the analysis yourself using Nielsen’s algorithm?
• How willing am I to trust results when I can’t see the underlying data myself, can’t evaluate the algorithm, and can’t do the analysis?

With our PreferenceSphere tool, you can drill down from the high-level 2×2 map showing Net Preference and Influence to the underlying Directional Graph, and drill down from there to see every one of the actual sound bites from consumers that generated the graphs. It’s an affordable, self-service approach that’s transparent and lets you examine source data and understand how it’s presented in graphical form. We believe such an approach gives you a great deal of confidence in the tool’s findings.

To illustrate the difference, here’s The Hatred Index from the WSJ article, created by Nielsen. It does make me wonder about how the algorithm generated these numbers and what the underlying source data had to say.

And here are the 2×2 graph and Directional Graph that shows visually how the numbers were calculated.  When we put this graphing capability into our product it will also show the underlying data, or “sound bites” as we call them.  For more on how to interpret these graphs see part 1 and part 2 of my postings on visually representing brand preference.

Whether you’re from Missouri or not, we think you’ll want to run these analyses yourself and have complete access to the method and data for reaching conclusions about brand preference. Let us know if that’s true. We welcome any other feedback on this evolving tool.

Comments (0)