Nielsen BuzzMetrics Tries to Measure Buzz in Social Media

Last year was a watershed for social media, with millions of people creating and sharing their own media on sites such as MySpace, YouTube, and Flickr and turning away from traditional one-way media such as TV, radio and newspapers. But for the proprietors of these new media sites, there’s one very big problem: How do you make money off that popularity?

The conundrum for social media is that if you try to commercialize that user-generated space, it loses its allure for the user who wants to be in control. But for research companies, there’s no conundrum — just a good niche to exploit for juicy data. One company stepping into that niche is Nielsen BuzzMetrics, an amalgamation of three other companies — BuzzMetrics (word-of-mouth marketing), Trendum (linguistic analysis) and Intelliseek (enterprise search)— which merged last year under the Nielsen/VNU umbrella. The idea for Nielsen BuzzMetrics is to measure buzz in social media such as blogs, forums and usenet groups and then package that data for corporate clients.

Marketers at companies such as Toyota, Sony and Coca-Cola have tapped BuzzMetrics to tell them what people on social media sites are saying about their brands and products. If they can gauge and nurture this buzz, the marketers hope to eventually turn positive buzz into sales. But building buzz online has its pitfalls, as various companies from Wal-Mart to Sony have found with the recent fake-blog scandals in which marketing blogs posed as fan-generated content and were outed by the blogosphere.

I recently spoke to Jonathan Carson, CEO of Nielsen BuzzMetrics, who noted that much of the data that his company finds online is freely searchable on its public blog tool, BlogPulse. Using that site, I checked out the way the conversation about Saddam Hussein’s execution changed over the past weekend, as the cell phone video leaked out to various blogs. You can also compare trends like the number of times bloggers mention “mother” vs. “father” (mom always seems to double up on dad, with a spike in both mentions around Christmas).

ADVERTISEMENT

Jonathan Carson

Fun facts aside, Carson believes the ground is shifting mightily under the feet of corporate marketers, who need to follow changes in online opinion even before a product comes to market. One recent pilot project for BuzzMetrics combined grocery sales of certain foods with the online buzz around trendy diets. Not too surprisingly, people were talking about low-carb diets for months before sales of such foods actually went up at the cash registers.

The following is an edited version of my conversation with Carson — who is, naturally, a blogger too. He talked about the fake-blogs controversy and defended his company’s decision to keep their own user-generated media conference, CGM Summit, off the record for attendees who might want to blog about it.

Tell me a little about the history of Nielsen BuzzMetrics and what you’re trying to do.

Jonathan Carson: We’ve been at this for a good bit of time. I started with BuzzMetrics in ’99, and the folks at Intelliseek started with the early versions of it in ’95, and they were certainly in this market by ’97 and ’98, and the folks at Trendum started around 2000. We were all pretty early, and all three companies were focused on tapping into data for consumer insight purposes. It took us a few years to appreciate how important it all was in the changing media landscape.

Our chief marketing officer, Pete Blackshaw, coined the term “consumer generated media” in 2002. That was the time where we realized this was going to have a big impact, the concept of citizen journalism or user-generated content or whatever you want to call it. Not only would it be an efficient and economic way to collect consumer feedback but it was also going to have a big impact on the media landscape.

We look at what’s going on [with consumer-generated content] as a pretty big revolution, and figure that with a whole new set of dynamics in the media landscape you’ll need a whole new set of measurement tools to understand size and account for media. To be totally honest, we don’t totally know what that means yet. We know some of the basics, and some of the core technologies to do it, but it’s still very early in figuring out what it’s going to look like.

How would you describe what you do to someone who isn’t technically savvy, like perhaps, your mom?

Carson: The most simple description is that we take all of the content that individuals are creating on the Internet and we perform market research on it. We understand what people are talking about and understand how they feel about various brands and products. We’re still walking the line between traditional consumer insight research, where you’re trying to understand just what individuals who may buy your product think, and what motivates them. And the other side of traditional market research is media measurement, understanding the media that people consume and their reactions to advertising.

We’re kind of a fine line between those worlds. We think that makes it very powerful, because they’re wrapped up together, but it’s also messy, because those are two sides of market research that haven’t really been mixed much in the past.

In web measurement in general, there’s been so many difficulties. Everyone talked about the web as being the most trackable, most measurable medium — that we’d know every detail when it comes to traffic. Yet there’s still arguments over unique visitors and page views, and numbers are thrown around and no one knows what to believe. What’s your take on that?

Carson: The problem is that we’re swimming in way too much data. In past lifetimes in research, there was such a limit of data and data was so difficult to pull together that one single very simple construct would be created. But it would be created in such a way that it was fairly rigorous, fairly defensable, and very repeatable, and then the industry stuck with it.

You look at something like the TV ratings or Consumer Price Index or even the Dow Jones and that’s the way it worked. With the web you have this massive, massive almost unlimited pool of data. From an intelligence standpoint we have more to work with than anyone could have imagined 20 years ago. But from the defensable, repeatable market research standpoint, you have a totally new ballgame, because those accepted norms haven’t come to bear. It makes for a very different environment where for any particular question or piece of information, you have a huge number of ways to find an answer, but you don’t have a single, verifiable metric to pin an industry around.

In some ways, it’s just the immaturity of the space, but in some ways the research and data industries will change and produce intelligence but not market research as it was classically defined. I think the new research companies that have come up in the past few years are going to give media companies the data to sort through as opposed to coming up with standardized metrics.

Do you think there will eventually be accepted standards for web measurement or do you think there will always be this mountain of data?

Carson: Standards will emerge, and I think that to some degree that’s important and it helps the marketplace progress. Within Intenet advertising, that space was a mess until the IAB [Internet Advertising Bureau] came up with standard ad sizes and banner sizes. Even simple things like that helped the market to explode because it made the process of buying and selling ads that much more efficient. Something like that can be credited with helping to drive the resurgence of Web 2.0 because it helped bring more funds back into the market.

Some degree of standards will definitely emerge, and that’s a good thing. The downside of standards is that they can chop off a lot of the nuance of data because they will be more generalized. The genie is out of the bottle on that, and there is more data and intelligence that can come out of that data. The market research world is headed for a real revival where whole new types of data and intelligence will be accessible to support the new media landscape and new marketing paradigm.

How does what you do help out the average person or media creator?

Carson: One obvious way to answer that is that we run a free and public tool called BlogPulse. It’s a way to provide some of our tools to the public for free, and expose our data and information to folks. But the macro answer to that question is that the things that are happening in the broader marketing world are good for individuals. The fact is that the world is becoming more focused around the individual, and individuals have much more control over what they purchase and consume. These are fundamentally good things, and what we’re trying to do is change the world of media measurement that responds to that new reality.

If companies are able to participate in that new reality, then that will speed up the process of people getting more control and autonomy.

What are some of the roadblocks for companies that want to start understanding this new world? Is it an old mindset vs. a new one?

Carson: We tend to think about companies as these monolithic faceless structures, but at the root companies are a whole bunch of people who work for the company. If you’re an employee that works in marketing and you have a budget to spend, you can show results with the traditional methods that your boss will know, understand and appreciate. The actual quantifiable results might be shallow or inaccurate but if they are the numbers your boss is used to, then you will probably be safe. If you are going to do something radically different, then there’s a risk that it won’t work; and even if it does work, you can’t prove it because you don’t have the numbers to prove it. So people’s natural tendency is to be safe and stick with the traditional.

The companies that we see changing the most have either always sat outside the traditional and never bought into the broadcast model, or they are companies that are very product driven, rather than brand marketing or image. These companies have a much easier time adapting to this world, because it’s becoming a world where bad products fail and good products do well because of their inherent goodness. Also, companies [do well] that have some kind of very strong top-down push, a mandate from the chief marketing officer to dramatically reduce television advertising.

BuzzMetrics put together a list of the Top 100 Most Popular Blog Posts for 2006. I’m curious what you learned from that list and the trends you found.

Carson: I thought it was interesting that there was a mix of the big popular A-list bloggers who had a whole bunch of posts on the list mixed with Long Tail [niche] bloggers with just one post on the list. It is the hypothesis of the Long Tail, that every once in a while, one of the Long Tailers can create a hit that makes it up to the head, and we saw that in action. We’re still in a stage of looking for examples of how the Long Tail will play out in the blogosphere, so it was interesting to see that in a quantified way showing up in the top 100 list.

Even among the A-listers on the list, there was a handful of them that had between 5 and 15 posts on the list, and there were also A-listers that we hear about every day who weren’t on the list at all. Some of them are driven by hits — by one post that gets a lot of attention. Crooks and Liars was a good example of that, but Daily Kos had only one post on the list. Daily Kos is an example where it’s an individual outlet that has a consistency of message and can get a daily readership [without “hit” posts].

We now have three or four years of good solid blog data now, and we can see these patterns emerging and see how the blogosphere will mature as a medium.

There’s been a lot of talk about fake blogs or “flogs” by companies such as Wal-Mart and Sony, and there’s also been controversy over the PayPerPost idea of marketers paying for blog posts. Do these things undermine the credibility of blogs or of the companies’ credibility?

Carson: Obviously, companies have to be extremely careful about the way that they approach [blogs]. Any time you approach a new medium, you’ve got to have a good understanding of how the medium works and how your brand would work within that medium. Certainly this year there were some major lapses of judgment by some of these brands.

But I also think it’s a pretty positive story, the fact that the reactions from the blogosphere were so swift and effective is pretty encouraging for any of us who are concerned about the future health of blogs as a medium. Companies are not really able to get away with a lot. Compare it to email marketing, where it took years and years for companies to conform to rules and standards that we all could live with. And most of us would still just hit delete rather than file a complaint with a company that was sending too many emails. In the blogosphere there’s a very low tolerance and a very swift reaction for any companies that do things the public doesn’t like. That’s a very good sign.

You recently had a conference, the CGM Summit, where you asked people not to blog — even though it was a conference on user-generated content. You explained on your blog that this conferece was a private event for clients and not for the public, yet there were a few negative comments about your policy from other bloggers. Do you think you could have handled that better?

Carson: I’m at peace with how the situation played out in the end. I think we learned a great deal from the experience, and it was a healthy debate. We took our punches in the process, but we thought about it in advance and took a very deliberate point of view for a reason. It was a justified situation. I hope that people, at the end of the day, even if they didn’t agree with us, at least thought it was an understandable decision. The reaction from a lot of individuals was extreme and they didn’t take a whole lot of time to look at the situation.

To put out the idea that every possible corporate meeting should be on the record is a little bit crazy. You and I should be able to have an off-the-record conversation. Just because we’re both bloggers doesn’t mean that everything should be fair game. There’s a reason for off-the-record conversations, whether that’s because of confidential information or just for background or whatever. I’m comfortable with our decision. We talked about having portions off the record and some on the record with our own people blogging it. But because it was the first conference and we didn’t have time to get permission from our clients to have some portions on the record, we just made the whole thing off the record.

What about the terms “consumer-generated media” or “user-generated media”? A lot of people don’t like those terms, including the term “consumer” because people are taking a more active role in media creation and not just consuming it.

Carson: I do completely understand it. A proper broad umbrella term is social media for a lot of this stuff. Consumer-generated media is a marketing-focused term. As an individual, I don’t want to be looked at only as a consumer but I don’t have a problem with companies looking at me that way because that is what I am to a company. Consumption does define the role of the individual from a company marketer standpoint. That’s our market: Nielsen BuzzMetrics sells to marketers at large corporations, so the term “consumer-generated media” is useful for us and highly useful for our clients, because it does describe the units of measurement they’re dealing with.

Some people would look at what you do and say, ‘I’m creating this blog or these photos on Flickr or I’m talking on an online forum. I don’t really want someone to come in and measure what I’m doing or follow what I’m doing.’ It’s in the public sphere online and people know that, but is there a reaction against measuring this type of media?

Carson: I think the marketplace is going to do a good job of taking care of that. Already, a lot of the big social media tools have or are developing tools to put permissions around the content. Within websites now there are technical protocols to alert [search engine] spiders that you don’t want them spidering your content. Our job is to just follow those protocols.

That level of permissions and control is significantly higher than what we all have in public life. When we walk down the street, we’re being counted for the billboards that we walk past. We’re being observed when we go to grocery stores and shopping malls by research companies. It’s pretty hard-wired into society now that there is some level of measurement of public activity. Our argument would be that the activity that takes place on the public Internet that’s being archived by various search engines and spiders is part of that dynamic.

Do you have any innovative approaches you’re working on within Nielsen?

Carson: We’ve started to do analysis on how online buzz correlates with sales or with audience size for television programs or things like that. One of our first pilot projects was with AC Nielsen to compare buzz around nutrition trends to sales of products that are associated with those trends. We’ve seen buzz actually leading sales data by up to a year.

It matches with what your gut might tell you. People talk about something before they actually do it. In the case of the food industry, where trends like low carbs or antioxidants seem to just fly onto the map really quickly without warning, the idea of studying buzz ahead of time to predict what grocery sales will be later is very powerful. That’s a big area of focus for us. AC Nielsen tracks all the sales within grocery stores. They have a product called ScanTrack which pulls in sales data from all the scanners in grocery stores. They have one particular view into that data which organizes the products according to nutritional trends. We did correlation analysis month by month to see how many months out the strongest correlation was.

If you look at the level of buzz — and low carbs was the case study we did — if you look at the buzz around low carbs and look at the sales of low carb products, and look by month, there’s almost no correlation. We ran this out over a couple years. But when you give it a lag, the correlation shows up four to six months later. People were talking about low carb dieting before they were really going on low carb diets [and buying the food].

*****

What do you think about Nielsen BuzzMetrics and the idea of a market research company quantifying what’s said on social media sites? Is everything fair game online, or should there be better standards or ways to opt out of measurement? Share your thoughts in the comments below.