Big Data: A Long Way from Plug-and-Play

One of the key markets for all the new big data analytics providers is marketers themselves, a group that should be a natural for turning deep customer insight into increased revenue. But are they ready? Well, according to a study by Columbia Business School and the New York American Marketing Association, although nearly all (91 percent) of marketers value and want to make data driven decisions, 29 percent report that their marketing departments have "too little or no customer/consumer data." Thirty nine percent of the marketers surveyed said their data is collected too infrequently and "not real-time enough." Two in five marketers admit that they cannot turn their data into actionable insight and about an equal number (36%) report that they have "lots of customer data," but "don't know what to do with it."

Researchers found that despite widespread adoption of digital marketing tools like mobile ads and social media, they are less likely to be measured for ROI. Eighty five percent of marketers are using brand accounts across Facebook, Twitter, Google+, and Foursquare but 14 percent of the social networking users are tying them to financial metrics. Fifty-one percent of marketers said they use mobile ads (in-app, or SMS) but only 17 percent of those using mobile ads are tying them to financial metrics. Forty one percent of email marketers measure their results with financial metrics. Overall, 60 percent of companies report that comparing the effectiveness of marketing across their different digital media is "a major challenge."

Forty two percent of marketers report that they are not able to link data at the level of an individual customer and 45 percent are not using data to personalize their marketing communications. Twenty-eight percent said they do not know which high-value customers to focus their marketing on.

Marketing success in the advent of Big Data means mapping marketing metrics to objectives aligned with corporate strategy, collecting and sharing data at the individual customer level throughout the organization, targeting and personalizing marketing efforts and measuring ROI across touch points. But marketing success in Big Data also depends on having the data in the first place. We're not all there yet, which is why the excitement around Big Data needs to be grounded in the mundane reality of where most marketers are today.

-- Nancy Ciliberti

Got Klout?

Imagine a business based on a mash-up of social media, analytics and ratings. And that's exactly where a company called Klout plays.

Klout exists to assess your social media importance. Using advanced algorithms, it looks at how active you are in social media, how big your audience is, how influential are the people in your audience, and the impact of your social media activity. All this gets rolled up in a Klout score - a number from 1 to 100.

If this sounds like nothing more than an interesting academic research exercise, you might be surprised. Klout reportedly has over 5,000 large companies tapping into its database to determine who really matters online. Uses are varied and fascinating. PR companies use Klout to assess whether or not to personally engage with someone who has made a negative online comment about a client. Marketers are creating customized pitches to those with the highest Klout scores in the hopes of engaging with them and getting them to talk to their audiences about their products. And this is just the tip of the iceberg in terms of potential applications. Consider, for example, that Klout has already built a connector to Salesforce.com.

In terms of potential applications, some are cutting edge, but not all are necessarily positive. There are numerous reports floating around of people applying for jobs and being rejected due to low Klout scores. Some hotels reportedly will look up your Klout score at check-in, and provide free upgrades to those with high scores, presumably in the hopes of favorable online mentions. Similarly, Cathay Pacific airlines will make its San Francisco frequent flier lounge available to anyone with a high Klout score - regardless of what airline they are flying. The objective again is favorable mentions.

Implications? What we may be seeing is a devolution in advertising where marketers move to a bottoms-up approach to distributing their messages, with the hope that they can achieve powerful and cost-effective reach by having a small group of individuals amplify their brands and their messages for them. This could have serious impact on those that make money today by aggregating fixed audiences.

Of course, as the rewards for having social influence grow, so too will the number of people gaming the system to improve their scores to reap all these upgrades, free samples and attention. As these activities accelerate, social media measurement could end up getting so polluted and undependable that it becomes too difficult to isolate true influencers, likely a fatal blow to this innovative new marketing approach. Alternatively, Klout, like Google, could try to keep the game going by regularly tweaking its algorithms to maintain its value. But as we add the wisdom of algorithms to the wisdom of crowds, are we really getting any smarter?

Big Data, Small Price

If you're not measuring your data in petabytes and exabytes, you're not truly in the world of Big Data, at least according to the purists. But increasingly, big data has come to mean the powerful analysis of much more reasonably-sized datasets, and that's where big data becomes a tool for all publishers to consider.

Consider just one of big data's outstanding virtues: its ability to drive innovation by providing
insights from data correlating customer behaviors, patterns and specific requirements -- including data that have traditionally been cast off as odds, ends and by-products of "analyze-able" data.

Big data enables the collection and analysis of offline and online data across touchpoints. Customers are generating more data about their buying behavior, likes and dislikes in more places than ever before. In addition to data stored in widely used CRM applications, untapped and tremendously useful customer satisfaction data exists in tweets, blog posts and other social media
content.

Using big data collection and analysis to capture behavior including preferences, product
selection and spending patterns across thousands of customer interactions enhances the ability to measure and impact customer satisfaction and loyalty in ways never before possible. Operational, financial and customer data across a business can now be integrated and processed efficiently to aid the identification and attribution of revenue drivers, yield actionable insights into product development and to provide deeper levels of customer engagement. Customer level revenue attribution, channel optimization, triggered marketing and marketing can occur more efficiently and reliably than ever before.

Big data is quickly becoming mainstream, and as a consequence, both the tools and associated pricing are becoming accessible to almost everyone. Earlier this week Google launched BigQuery which puts a powerful, straightforward and relatively low cost cloud-based data analysis in the hands of a broader category of companies.

Powered by Google's might, BigQuery's best application is interactive analysis of very large datasets. BigQuery is a SaaS program service that runs on Google infrastructure. It accepts CSV files uploaded from customers via an API. The API uses concurrent compressed streams which allow customers to upload several hundred gigabytes in minutes with analysis that Google estimates is "about 10 times faster than the speed of many corporate data systems." Uploaded data is secure, replicated across multiple data centers and can easily be exported. Data is accessed via group- and user-based permissions using Google accounts.

One advantage of Big Query is its straightforward nature and its ability to keep costs low. Cost is $0.12 GB per month for storage with a 2TB limit and queries are $0.035 per GB with a limit of 1,000 queries per day. Prices are negotiated beyond those limits.

As capital "B" Big Data increasingly becomes small "b" big data, we're going to see even more tools and services that will allow publishers to easily and cheaply find important new insights about their customers and their markets, and that's what makes big data a big opportunity and a big deal.

-- Nancy Ciliberti

To Find Gold, Dig Deep

The travails of the traditional yellow pages industry are serious, with no end in sight. There are some interesting lessons and insights that can be drawn from the remarkable and relatively rapid meltdown of this seemingly bulletproof and impossibly profitable segment of the data publishing industry.

What made yellow pages arguably the most profitable form of publishing ever was the combination of monopoly-like characteristics, coupled with a product that worked extremely well for advertisers. But with every community sporting a yellow pages product, those monopoly-like characteristics became a two-edged sword because, while margins were wonderful, there was little room for geographic growth. For many years, yellow pages publishers contented themselves with eye-popping annual price increases in their home markets. After the AT&T divestiture, there was a huge effort to poach territory from other publishers, resulting in many people finding anywhere from two to ten different yellow pages on their doorsteps. That’s why when the Internet came along, what the yellow pages publishers saw was a way to vastly expand their territories. Forget adding a new market or even a new region. In one fell swoop, they could become NATIONAL. Nirvana!

In reality, this move to become national online publishers was horribly ill-advised. Yellow pages publishers know nothing about content. All those pesky names and address listings they publish? They simply buy them, and view them as a necessary evil, useful only for separating the display advertisements. And this national expansion overlooked the core dynamic of all yellow pages and buying guides: the advertising is the content. By extension, no advertising, no content. Yet with regional sales forces, these publishers had no capability to sell advertising on a national basis, so instead they offered a thin gruel of content: company name, address, phone and general category. Not exactly compelling even in the early dates of the web, and almost pathetic today. The Internet goldmine turned out to be a black hole.

That’s lesson number one. If you are going to produce an information product, that product needs to offer … information. And the competition is keen in the online world where information begets more information that a user can get elsewhere. That’s why the most successful information producers tend to keep a tight focus. Better, deeper, fresher. And the best data publishers are now wrapping software around their content to make it more useful as well. If none of this sounds like yellow pages to you, you’ve gotten my point.

The other useful lesson revolves around momentum. Until very recently, people pointed to the yellow pages as one print medium that while not thriving, was at least holding its own. Yellow pages, some believed, were different and immune to the forces of the web. It’s amazing how quickly that story changed.

That’s lesson number two. Momentum is a powerful thing. Couple it with a strong brand, years of measurable results, a largely unsophisticated advertising base (potentially as many as half of all local business do not yet have websites) and a saturation distribution strategy, and you can appear to defy gravity for a long time. It’s comfortable sticking with what you know, especially when the old way is more profitable than the new way. But when you’re defying gravity, reality will intrude unexpectedly and you’ll hit the ground hard. The smarter path is to engineer a soft landing, not to convince yourself that your market is somehow special and different.

To sum it all up: if you stay focused, dig deep and keep your feet on the ground (making stumbles is a lot less painful than crashing), you’ve already mastered some of the fundamentals to successful information publishing.

To Market, To Market

I have long been interested in the fine line that often divides marketplaces and buying guides, a topic that I am sure keeps all of us up at night at least every so often. A string of recent new website announcements has me back thinking about this again.

As I see the world, a buying guide exists to help you find a source of supply (e.g. rubber gaskets) and a marketplace helps you find a specific item for sale. Buying guides are designed to introduce buyers and sellers. Marketplaces are transactional, designed to get things sold.

Buying guides, which have their roots in traditional yellow pages, tend to be advertising-based. Marketplaces prefer to get paid a percentage of each sale they effectuate. Generally, getting a piece of the action is seen as a much more attractive revenue model than advertising (think eBay). That’s why even old-line yellow pages publishers such as Yell are trying to pivot to become marketplaces. As a general rule though, marketplaces designed for existing markets tend to be highly disruptive, making it hard for them to get traction. Create your own market, however (think Airbnb), and getting traction becomes a lot easier.

And whether buying guide or marketplace, traction determines whether you succeed or fail. All marketplaces and all buying guides need to bring together enough buyers and sellers to be viable. Let’s look at two recent launches:

Archability is a new hybrid buying guide for architectural services. I say hybrid because in addition to directory listings of architects, it also allows buyers to post projects to which architects will respond. This is an efficient hybrid because sellers can get a stream of qualified prospects and buyers can post their needs without a lot of work to identify the architect who meets their exact needs. Critical to any marketplace, Archability stays in the middle of the transaction all the way through (don’t expect to get commissions on transactions you don’t know about), in part through a clever escrow service it offers. And while this is an existing market, Archability isn’t displacing any current industry players, just functioning as a lead generation service, so market resistance should be limited. If Archability can get enough buyers and sellers in the room, this looks like a strong concept to me.

Now consider another start-up, PlaneFinder.com. Got an unused leg on your private jet? Post it on PlaneFinder and sell it to someone. Conversely (just like Archability), buyers can post that they’d like to hop a jet between Point A and Point B on a specific date and wait for a response from an interested jet owner. It’s a useful concept, but I see two concerns. First, PlaneFinder is a marketplace that acts like a buying guide. By that I mean, it makes the introduction and steps out of the transaction. Convenient for all parties to be sure, but not a great way to assure payment. Second, I worry about traction. Aircraft flights are perishable. That makes it even harder to have enough interested buyers and sellers in the room at the same time. And you need a lot of buyers to assure you can move that Boise to Tampa flight leaving on Thursday. Which also means you need a lot of flights, all the time. When buyers and sellers go away unsatisfied, they tend to go away for good. While I like the idea, you not only need a lot of buyers (a given), you also need an extremely large inventory of flights at all times (an unusually steep hurdle).

Lesson learned: for buying guides thinking of adding a transaction side, remember that transactional businesses are different and hard. While the term “marketplace” is widely used and badly abused, you’ve got to have buyers and things for sale immediately; there’s no option of slow growth, results (or lack thereof) are easily discerned, and there is no ability to finesse low participation levels. If you’re up for the challenge, the prize is worth it. If not, maybe a hybrid model is worth exploring.