Big Data Doesn’t Know A Thing, But It Sure Can Ask the Right Questions

When I was 12, I remember telling my mother that if a human being could know all the events that factor into the moment just before an automobile accident, it would be possible to avoid the accident. Little did I know that, a) I was talking about “big data;” b) I was wrong.

Fortunately, a couple of guys who actually know what they’re talking about have written the book that explains it all: Predicting the Future With ‘Big Data’.

Kenneth Cukier is data editor for The Economist and Viktor Mayer-Schönberger is Professor of internet governance and regulation at Oxford University. The co-authors appeared on the Kojo Nnamdi show March 7.

If I understand what Cukier and Mayer-Schönberger are saying, big data leads us not to more facts, but to more data, within which will be more questions and, therefore, greater opportunities for insights. And yet, the whys of “the universe” will remain a mystery.

More Data Is Just More Data
The paradox, apparently, is this: More data negates — or at least counter-balances — the need for perfect data. Mayer-Schönberger puts it this way: “… as we have more data, we can also be accepting some inexactitude in how we collect the data and how perfectly curated the data is because we just have so much of it … [And, any move away from] an elusive quest to find causality, targets something much more pragmatic and much simpler called correlations. It means that we are not looking for the why. We are looking for the what and that’s good enough.”

The Power of the Spread Is the Greatest Power Yet
Mayer-Schönberger thinks both the invention of the printing press and the invention of the Internet will be dwarfed by another major leap forward: not the mere spreading of information, but the spreading of the power of information. [Crowd-sourcing, anyone? Or, maybe, targeted sharing?]

Say what? Cukier explains. “You can imagine that [big data is] going to actually change the way that businesses run. They’ll find their most precious asset might not be what they’re actually building [or selling or offering], but the data that goes into it — because they can learn from [that data] and cross-apply it to other things.”

Got data?
You’re in the cat bird’s seat. And you know it. As a marketer, you know better than to despair at data results. You just test again … and again. You just get more data.

Cukier gives us a good example: “Imagine an algebra teacher who would find out that 60 percent of her students got the same question wrong with the exact same answer. She would, therefore, learn that, in fact, maybe she taught the algebra wrong, that maybe she wasn’t clear enough.” Insight. New testing. More data.

Data Is Fake
Should we worry about the issue of big data and our privacy? Sure, because we always need to be vigilant as a society not to misuse or abuse big data. But keep calm.

Cuckier explains. “Data is only a simulacrum of reality. It is not the real thing, firstly. Also, we’ll never have all the data. That’s not actually possible … So this sort of hypothetical thought experiment of what will happen when we know everything about everyone, that day is just not going to happen. So on a practical level, to get wound up in knots about this doesn’t seem useful..”

If only I’d known then what I know now …

scrubbed by MarketingBrillo

"They let their data be their spin."

The headline above echoes Chuck Todd, Chief White House Correspondent for NBC News, talking election strategy -- and success -- today on Morning Joe.

Todd -- an admitted enthusiastic for the power of data -- attributed President Obama's re-election victory in large part to the data-directed dedication of Obama's top political advisers Jim Messina, campaign manager; David Axelrod, political adviser; and political strategist David Plouffe.

As Scarborough's panel noted, post-election results prove that this group of strategists and analysts knew they would win and they knew why. How could they have been so sure?

Marketing Lesson #1: Get the Data
In Obama's reelection effort, Big Smart Data did the trick -- big data that already knew where to find Democratic leaning constituents, including where they shop, what they buy, where their spouses/partners work, what magazines they read, what TV programs they watch, what moves them -- all of it.

Richard Stengel, managing editor of Time Magazine, also touted the value and impact of sophisticated data mining described in the magazine's November 7 story by Time White House correspondent Michael Scherer. A similar Poynter article quoted a senior Obama campaign official saying, "We ran the election 66,000 times every night," said a senior official describing the computer simulations the campaign ran to figure out Obama's odds of winning each swing state. "And every morning we got the spit-out -- there are your chances of winning these states. And that is how we allocated resources."

And so it went ... "All into one gigantic database," Stengel noted.

Marketing Lesson #2: Put Effort Where It Counts Most
Marketers call it the 80/20 rule. In the 2012 election, Big Data was applied to instructing Obama's reelection efforts -- particularly in the nine swing states -- exactly where and how to boost voter registration. In other words, the point wasn't to convert non-believers. The point was to grow and nurture the believer base.

The Obama campaign also took Big Data to an art form in areas where the Democrats were traditionally less strong, but where great potential lay. "[In Ohio, among labor] the effort was particularly helpful in targeting some of the more difficult demographic groups - white men, for example."

Marketing Lesson #3: Follow-up with CRM
From the gigantic database, relevant information was converted into "boots on the ground." In all nine swing states, for years before the election, committed volunteers worked to "get out the vote." They registered voters, of course, but it didn't end with registration. Many volunteers actually became friends with those they had recruited: they had coffee together, they stayed in touch, they practiced the proven tactics of customer relationship management (CRM). And, on election day, these volunteers made as certain as possible that their constituents actually would vote: they phoned, they visited, they offered rides, the followed up, close and personal.

Marketing Lesson #4: Don't Talk Down to Your Audience
To marketers, of course, Big Data is no mystery. "Companies like Proctor & Gamble, are accustomed to calibrating data against message," Stengel pointed out. The "secret," of course, is to spend money only on messaging that's targeted and effective.

Making this point, Stengel noted that some of the opposition's repetitive TV ads actually helped Obama. In particular, a Romney campaign ad playing in industrial Ohio that many viewers believed misrepresented Romney's position on the auto bailout, angered the well-informed viewers in Ohio's heavily unionized areas. "Watchers do become experts on ads," Stengel noted.

Having Said All This, If It Don't Work, It Don't Work
The Republicans had data, too, of course -- a huge machine they nicknamed Orca. Somehow, though, Orca went awry. An article by political columnist Paul Glastris that appeared in Washington Monthly sought to address the question, The Mystery of Why Republicans Were So Sure They’d Win.

Glastris wrote, "Orca, which was headquartered in a giant war room spread across the floor of the Boston Garden, turned out to be problematic at best. Early in the evening, one aide said that, as of 4 p.m., Orca still projected a Romney victory of somewhere between 290 and 300 electoral votes. Obviously that didn’t happen. Later, another aide said Orca had pretty much crashed in the heat of the action. 'Somebody said Orca is lying on the beach with a harpoon in it,' said the aide."

Just for the Fun of It, Remember: Anecdotal Evidence Also Counts
Rising Democratic star San Antonio (TX) Mayor Julian Castro explained how voters in his city were persuaded to approve a modest one-eighth of one percent tax increase dedicated to underwriting high-quality pre-kindergarten for thousands of children. Castro noted that -- contrary to the "no new taxes ever" mantra, people will accept taxation when they know its purpose. Voters understand the need for education for young children, understanding that "brain power is the currency of success," he said.

Incidentally, Castro predicts the effect of the Hispanic vote will take Texas to the Democrats in six to eight years.

-- scrubbed by MarketingBrillo

Why Associations Should be Afraid, Very Afraid

A few weeks ago, I blogged about how free flowing content exchange on the Internet -- both video and written --  is striking a blow at the heart of what trade and non-profit associations have offered members for so long: industry-specific information, straight from experts. 

Today, I find out that the Internet also is attacking another major association strength: networking with peers.

When I signed on to LinkedIn this morning, I had an invitation from Citi® to join "Connect" their "premiere network for professional women." In the words of the "invitation," Citi® and LinkedIn are teaming up to bring you a premier group experience designed specifically for professional women that provides career advice and networking opportunities.

How much does that sound like a typical "Join our association" pitch? Except for one thing: it's free.

So, really, why wouldn't I accept this invitation? It could have some benefits. And it costs me nothing to find out.

I hadn't envisioned Citi® partnering with LinkedIn, which goes to show that the CMO at Citi is a lot smarter than I am and tells me that big players (like banks and financial institutions) will continue to refine social media marketing -- using it, in this case, to bypass traditional "associations" and involve themselves directly in the networking and social activities of prospective clients (women).

So how might this work? One dimension of the Citi/LinkedIn effort assists the formation of geographically aligned "small groups," to wit: Network in Your Neighborhood. Does anybody care to network in their neighborhood? Apparently, yes, says their pitch. "A number of you have expressed interest in meeting up with women from Connect who live near you ... If you want to start a group or are looking for one, use this discussion to let us know where you are!"

LinkedIn is a brilliant partner in this effort, of course, because it possesses a superior database of exactly the individuals Citi® wants to reach. Databases, of course, lie at the heart of all social media: Facebook, Twitter, LinkedIn, Pinterest, Flickr, StumbleUpon, ad nauseum.

Talk about effective direct marketing! Really, it doesn't get much better than this.

Expect more of the same ... much more.

-- scrubbed by MarketingBrillo

No Registration Required, So I, Too, Must Be a Segment

A few weeks ago, HubSpot offered me a report. They had me at “Hi, Nancy,” saying:

The best-performing email campaigns don't always have great subject lines. Or email copy. Or even calls-to-action … The experts at MarketingSherpa have discovered that two specific components greatly impact the success of your email campaigns: integration and segmentation.

I loved that copy, but the REAL treat was THIS Line: Get the Report Now (no registration required).

I’ve downloaded dozens of HubSpot’s great content over the past 18 months, but this was the first time I haven’t had to register all over again.

Perhaps I’m now in a “regular customer” segment. Whatever … it was sweet. And that was just the dessert. The meat followed: The report itself had great advice about how to segment your list, including these tips:

1. Be prepared to make your case and be patient. Email researcher Michael Wexler advises, "There's a high cost to entry for classic segmentation testing; it costs more than a simple test. However, this investment helps a lot. It stems lowered results, reporting of spam, unsubscribes, and it results in higher lifetime value per name."

2.  Gather your assets, namely these three: your email database, your process for testing, and great content.

3. Collect Data. MarketingSherpa identifies four types of data : endemic data from the subscriber; transactional data, behavioral data, and computed data developed from calculations performed on one or more variables.

4. Base your segments on long-term behaviors, for example:

    a) Were your customers brand-sensitive or price-sensitive?

    b) Do customers self-segmented based on the particular product purchased?

5. Identify segmentation types, for example: geographic, product type, lifecycle, personal data (self-reported, appended or behavioral).

6. Consider a variety of segmentation approaches. Even simple segmentation by customer profile or email activity can reap considerable rewards.

7. Start with a single segment (for example, subscribers). This tactic enables marketers to manage unforeseen challenges, adjust strategy, and prove the value of segmentation before getting too complicated.

8. Treat inactive subscribers as a segment, too.

9. Keep it under control; excessive segmentation can be wasteful.

10. Leverage and repurpose content to keep up with the demand of segments.

Download the full Special Report, “How to Segment & Integrate your Emails for Better Results,” from HubSpot. 

Quickie Revelations for Better Blog Results

I downloaded HubSpot's "6 Deadly Marketing Myths Busted." Author Dan Zarrella is HubSpot’s award winning social media scientist and author of Zarrella’s Hierarchy of Contagiousness.

This free HubSpot download turned out to be a great analytical gem and I suggest you get your hands on it. In the meantime, here are seven revelations that popped at me while reading:

1. Zarella's research has found no significant correlations between the number of comments a blogpost received and the amount of traffic that blog post got. In the words, conversation doesn't drive traffic. Attracting visitors takes interesting content and "broadcasting" [letting everyone know about the content].

2. Zarella concludes that, on Twitter, “engaging in the conversation” doesn’t work to increase reach. That requires broadcasting more interesting content. Highly-followed accounts tend to share more content.

3. When it comes to blogging, more posts equal more traffic.

4. The words "please retweet" do work. Zarella concludes, "Make sure that your tweets, and all social media updates, contain verbs that prompt your readers to take an action and learn more about your organization."

5. Weekend emails have a much higher click-through rate (CTR).

6. Which self-descriptors (that is, which titles) garner the most Twitter followers? [In descending order]: officials, founder, speaker, expert, guru, author.

7. Remember The Seven Year Itch? Your new subscribers love you best. The longer a subscriber has been with you, the lower the response rate.

A Database Will Become Something Else Altogether. But What?

Multics released the first commercial relational database management system in 1976. Soon thereafter, in the 1980s, direct marketers moved to computerize and warehouse names and addresses. Zoom forward to today. Databases have become the fundament of direct marketing, customer relationship management, and a lot of other core business initiatives. The question is: Where is this evolution going?

In particular, I'm thinking of Facebook, which is becoming an Internet of its own [or even the Internet itself, some argue], with vast quantities of personal information databased in one location. With the addition of Facebook business pages where folks actually shop from within Facebook, this "FB universe" takes on a whole new dimension of control [and, reportedly, profitability].

Should FB begin to thoroughly monetize it's database by selling its data warehouse [and how can it possibly resist the temptation?] consisting of trillions of terabytes of information (names, email addresses, photos, educational background, connections, shopping and Internet habits, photos, videos, birth dates and records of children's growth and activities, from over 500 million subscribers internationally, as of this hour) how will external databases compete? Despite a firestorm of warnings about what not to do on Facebook and the fact that we sign away our rights to everything we post there, the user base keeps growing. So ...What will happen to databases as we currently know and love them?

It's a huge question. And only imaginings of the Brave New World variety can intuit the possibilities. As I work on the April "Lists and Databases" issue of DMAW's Marketing AdVents, I wonder how many years into the future this vital element that has driven direct marketing for three decades will remain as we have come to know and love it.

Thoughts anyone?

-- scrubbed by Marketing Brillo