Using Social Input for Marketing Decisions

Should likes result in price cuts?

The impact of social media on commerce may be in consumer feedback rather than direct commerce.   A recent article in the Chicago Tribune outlined a number of retailers - large and small - using the communal media as a way to inform their decisions.  From what colors to stock to what prices to charge social networks provide a means for either conducting market research or letting consumers decide.

The use of natural communities for research is by no means new - but the quick ask and answer paradigm of social media does raise some questions to consider.   If we give consumers such a visible role in the process, then as a business...

• Do we care what the decision is?   
• Colors, assortments, and other merchandising decisions are good places to test the waters.
• Can we deliver? 
• In essence, asking for opinions is on the path to building trust and that outcome should not be put at risk.
• Can we afford to be wrong?
• Certain topics, e.g. pricing and new product development, may result in unintended consequences that impact other parts of the organization.  

Strictly speaking, market research is a tool used to confirm or reject a hypothesis so there are some limitations to consider.

• Selection and sample bias -- no one ever said those who participate and respond are 'representative consumers'.  Can the results be applied broadly or into the future?
• Classification and segmentation - verification of the interests and demographics used to analyze the results is tough when at a click of a button I can change, include or exclude things.  There is truth to adage - 'on the Internet, no one knows you're a dog.'

That said, there are clearly innovative market research, data collection and analysis opportunities to explore.  This area is important for marketing since it is the data of self-expression that holds the keys to context and intent and thus purchases. A few examples..

• ChatThreads combines a bit of market research methodology, technology and social behavior to track where and how a recruited panel of consumers interacts with a brand.
• ThinkVine creates a virtual world of consumers by building personae from all types of data in order to understand how the online and offline mix of marketing spend impacts sales.
• itracks focuses on mobile focus groups so they can interact in the real world instead of the conference room with one-way mirrors.

So, since there is no reason to believe these trends are simply fads it is time to rethink how we go about figuring how to learn what we need to know.  

There needs to be some amount of experimentation in the generation of insights.  Suggestion for the new year: set a budget limit to which you simply say "Yes, let's try it."

Mismatching Brand and Expectations

What if your stucco repairman drove a BMW?

We recently had some work done on our house to fix some cracks and chinks in stucco.  The crew did a good job, communicated well and earned a strong review on Angie's List (in exchange for a discount).   What was interesting was the question that the contractor asked at the end:  "Did you mind me driving to the job in a BMW?"

To be sure, I did have pangs of being overcharged.  But upon reflection, it was "just" a 3-series and I've seen enough pick-up trucks to know they can cost just as much - if not more.   So, it wasn't the price that caused the twinge but something else.   Maybe it was a case of branding at work.

BMW is (was) an aspirational brand; the 'ultimate driving machine' is something to strive for, a reward for success, a pinnacle of perfection.   It is very personal.  A truck on the other hand is there to get the tough jobs done - just consider the Chevy Silverado winning Super Bowl ad that reinforced it's message as the longest lasting.  Even Viagra uses solving challenges as part of delivering their brand promise.

In Simon Sinek's TEDx talk he makes the point:  "people don't buy what you do but why you do it".  So, my twinge came from a conflict in the why side of the equation.   The problem a BMW solves has to do with personal needs; the problem an F250 solves has to with the job. 

So, yes I would have preferred that he drove up in a truck.

The Rise of Native Advertising

Will something replace display ads?

As click thru rates for display ads continue to slide into the abyss - now estimated at less than 0.2% - advertisers and publishers are looking for alternative ways to take advantage of consumers' use of the web as a source of information.    

A recent (and probably welcome) trend is toward 'native advertising' something that Solve Media has defined as "a specific mode of monetization that aims to augment user experience by providing value through relevant content delivered in-stream."   In general, they're talking about custom content associated with the context and intent of the consumer - things like sponsored stories, web videos and interactive graphics.  

Given that these assets are delivered as part of the experience, as opposed to simply occupying the 65,520 pixels of a leader board, a lot more care and feeding in their creation and publication is required.  Bidding for eyeballs, whether in real-time or thru fixed line-item bids, typically focuses on impression attributes such as the characteristics and behaviors of the visitor along with page/section information and often has a short-term payback based on conversion. Since the highest-bidder wins, there is no guaranteed link between the subject of the ad and the consumer context.  (I still see ads for things that make no sense anymore - like a pressure cooker (got one) and an online MBA (don't need another one.))

So, while it is an interesting trend particularly in light of mobile and tablet usage, figuring out native advertising requires answering 3 questions:

1. What content are different people or segments interested in?
2. How do various content types impact sales?
3. How do we optimize the creative and content mix across the journey?

The answers to those require a different approach to generating insights than applied historically.  To allocate budgets, we'll have to develop some set of 'content attributes' in order to answer them.  

In short, we need to figure out how content works.

3 Levels of Identification: Consumer, Segment, Context

How do we move along the touch points of a consumer's journey?

Media fragmentation and interactive touch points results in media plans that probably look more like a 3D chess board than Candy Land.  So, how do we plan our next move?

Recognition is the key to the personalization and customization of consumer experiences.  But, there are clear issues revolving around privacy, governance, and permitted use that may prevent us from doing as much as we would like or even can.  Zappos has had to find the right balance between privacy and targeted ads because too much of a good thing is a bad thing.

There are several ways in which we can approach the need to help people choose.

• Consumer: the ideal scenario in most cases is the the explicit permission to market with an identified individual.  This typically relies on a preexisting relationship where opt-in has already occurred thus naturally focuses on retention.  Cookies, because of their ephemeral nature, are a substitute (sometimes good, sometimes poor) of actually knowing a person and bridge to acquisition programs.
• Segment: in lieu of identifying the individual, it is common to identify and track behaviors and classify them into segments.   Sessions of ecommerce sites might have 'researchers', 'tire kickers', and 'category buyers' for which different combinations of content and offers are presented.
• Context: there are cases where even anonymous tracking should be avoided, e.g. medical, pharmaceutical and health information.  In these scenarios it is the behavior of the content that drives segmentation, and thus the business rules. 

Given the need to cover a wide spectrum of the consumer journey, the challenge and thus opportunity becomes the design of scheme that simultaneously leverages all three levels of recognition.  To do that it seems we need three things:

1. Identification Linkages - a means to link individuals, identified or not, as deep into the journey as possible.  For example using a match service to link email and cookies, or even cookies to Prizm clusters.
2. Portable Segments - a structure to behavioral segmentation that can be applied across touch points in order to deploy common business rules.
3. Content Archetypes - a classification scheme (meta tags) that allows us to look at how similar content works across multiple scenarios.

  And to become a grand master we'll have to experiment and test a lot. 

Playing Poker with Big Data

Is finding insights a skill or just luck?

In Forbes' "Big Data News of the Week" several points were raised about the insights side of the equation.  

• Machine learning requires a human touch.  We can't just implement a silver bullet because the problems require either business domain or technical knowledge in how to address the problem in the first place.
• More doesn't mean more. There are marginal returns to more data.  In fact, the proportion of value derived probably decreases rapidly the closer we move to the realm of self-expression. Drawing the appropriate line in the sand requires judgment.
• Entrepreneurs love scarcity.   The lack of 'data scientists' is leading to the funding of companies that will do the job analyzing the data for you.  Even the reporting platforms are embedding analytics into their offerings.  Clearly a disruptive innovation in the making.

All this suggests knowing when to hold 'em and when to fold 'em is the critical skill.   In an article in the Economist, skilled poker players had an ROI of 30% while the unskilled were down 15%.  Victory was based on the ebb-and-flow of in game decisions, just like in the world of big data.

Applying Big Data Philosophy To Silos

Can we use big data architecture to solve big company problems?

One of the most significant challenges to overcome in any organization is silos - those technological, political and budget-driven islands of activity.  Email is here; ecommerce is there; social is in another department; and display is managed outside.  Each produces data that might be useful in achieving objectives, but they are locked up.

We know, or at least believe, that we need to deliver a seamless experience based on a 360-degree view of the consumer.  What we do not know a priori is what tactical campaigns or even strategies might be appropriate.  Since insights change how we market, rather than just tweak a plan, we need to explore and test first.  And that's the rub - I can't define a complete data architecture because I don't yet understand how I'll use it.

The typical, big company approach is to kick of a project lead by IT to design an integrated solution based on structured data models and ETL (extract, transform and load) processes.  The vision is a common company database that can be shared and leveraged appropriately through-out the company.  I've worked on those projects throughout my career.  But is that the right approach in light of the realities of today's needs?

When data management sources were scarce, ie expensive, then it paid dividends to design the solution first. In the early data warehouse days, we'd estimate building a marketing solution at nine months and $1,000 per gigabyte.  Today, I'm not so sure we have that much time due to the volume, velocity and variety of data. 

Because we're dealing with a data fire hose, we can't design first because the sources will change before the ink dries on the umpteenth version of the Visio diagram.  In fact the whole notion of applying structure to unstructured data is somewhat of an oxymoron.  We need to get the data together first, start working with it and finally apply some structure.  Since Big Data architectures put the data design at the end of the process, maybe they're appropriate for more pressing tasks - like looking at the integration of silos.

The design stage needs to shift to focusing on the environment rather than the contents.  The key marketing questions actually become technology ones: How will we stream data into a useful place?  How can we reduce the time to get answers to questions?  How will we marry on and offline data? How do we deploy insights? 

So, I think it is time to jump in and see what I can learn from email and display data.

2013 Trend Season Has Begun

What's on the horizon?

'Tis the season to start publishing trends for next year.

The first set is from trendwatching, a global trend spotting organization.  Three of note:

• Data Myning: the reversal of data ownership from businesses to the individual consumer.  The challenge for marketers will be to find the right balance between anticipating needs and watching. 
• Full Frontal: changing the mind set from saying 'we have nothing to hide' to proving it.  The discussion needs to move toward permitted use and governance and away from privacy.
• MobileMoments: using digital snacks to maximize experiences.  Apps like 'snapchat' allow images to be shared for seconds before self destructing.  The half-life of interest just got shorter.

And those macro trends seem to be wrapped up in content trends from D Custom, an agency out of Dallas.  Summarizing, content will be...

• Tailored to the Consumer - I want what I want, not what you want to send all of us.
• Placed above the Channel - I am a cyborg, my life is more social and rich because of seamless mobile interactions.
• About the Experience, not the Word - I can dedicate all my senses for a brief moment of time, so please stimulate them

Other sources reflect the same theme of experimentation to get it right; some with examples of technology rather than exploring the trend itself, others focused on key marketing disciplines.

• EBrkiks Infotech's slideshare on digital trends.
• From a brand marketing perspective, Brand Strategy offers this view. 
• It's back to basics; less tricks as control continues to shift toward consumer.