Digital Analytics: Big Data; Big Mess

Digital analytics, seen by some as a panacea for optimization leading to enhanced performance, is full of problems that are often ignored. This is part of the inherent bias toward numbers and the belief that therein lies the truth. Ushered in by the expansion of big data analytics tools, many managers simply allow the numbers to guide their actions without thought or understanding that those numbers might be lying to them.

In today’s post, we’ll explore how to use data effectively, considering that this data might not reflect the truth. We’ll discuss ways to question your data and derive more nuanced insights from it. So, let’s dig into this very weighty and uncomfortable topic. Buckle up!

social media analytics

Bad press for big data

I’m not alone in questioning the rationale of putting absolute faith in numbers. Recently, a rash of articles blames such a reliance on numbers for everything that’s bad with business.

In some ways, the backlash against overreliance on data is natural—people resist anything new. But some of the bigger names in marketing perpetuate this type of thinking, so we must consider their arguments.

Here are just a few of the top blogs bashing data with respect to what it’s done for marketing:

  • squeezes all the creativity out of marketing [source]
  • leads to inaccurate marketing strategies [source]
  • doesn’t offer trustworthy insights for marketers [source]
  • data and predictive analytics are getting in the way of doing basic marketing [source]
  • lied about election results, so why should we trust it to help with marketing [source]
  • doesn’t belong in marketing’s future [source]

And I could go on and on with the number of people who now question whether marketing should rely so heavily on data, regardless of its size.

Is data killing marketing?

I think several issues underpin objections to using data in marketing — big or otherwise. Here they are:

  • Expecting data to provide answers, not insights
  • Using bad data and expecting good answers
  • No connection between data and marketing concepts
  • Don’t understand how to use data to drive actions

Let’s deal with these one at a time.

Expecting data to provide answers

In some areas, data provides answers.

If you’re an accountant, data generates an income statement and figures your taxes.

A financial planner uses data to determine how much money you need to invest to retire comfortably.

If you’re a doctor, data helps you calculate the proper dosage of a drug to prescribe or even whether your patient’s data justifies the prescription.

If you’re in production management, data determines how much raw material you must order to fulfill demand.

In each of these cases, practitioners have years of education and experience allowing them to USE the data to answer their operational questions.

Do marketers have the same training in data analytics? After teaching at universities for over 30 years, I can tell you the answer is a resounding NO. Marketing curriculum at most universities in both undergraduate and MBA programs do nothing to prepare marketers to analyze the vast amount of data generated in the digital world, much less how to use that data to build insights.

big dataA bigger problem is that some folks, like the ones in the cartoon above, try to use marketing data the same way you might predict raw material needs—if X, then Y. They assume marketing data is determinant. Marketing just doesn’t work like that because consumer decision-making is a function of a variety of factors, and we can only calculate probabilities that, on average, consumers will do Y if exposed to X. Does that mean marketing is broken or that using data won’t help with marketing?

Absolutely not.

You just have to use marketing data in a different way than you do accounting data or medical data.

First, take a look at where marketing data comes from (according to IBM—and they should know). Some of it is concrete data, including transactional data, such as how consumers responded to changes in marketing strategy, and data from existing marketing efforts, like email and digital marketing campaigns. That’s pretty accurate stuff. Still, you can’t say with certainty that a price change that generated X increase in revenue will produce the same change in revenue if attempted again. In fact, using the strategy too frequently trains consumers to wait for a discount, so they put off shopping when prices are not discounted.

Another big batch of data comes through social media. That stuff is a little wonky. First, it doesn’t consist of numbers — unless you include vanity metrics like shares and likes (which most data analysts don’t). Most of your data is unstructured, which is IBM’s way of saying it consists of messy words that are hard to interpret — so many companies just don’t. That means you’re losing 80% of your data (IBM estimates that 80% of data is unstructured).

So, a big part of the marketing problem with data is that you’re ignoring 80% of it.

You also need to recognize that consumers aren’t robots. Sometimes, they say one thing and do another, or they do something today and don’t do it tomorrow. They say variety is the spice of life, and consumers try to prove that every day.

Data can only provide insights into consumer behavior; it can’t provide answers. We’ll discuss this more in a later section.

Here are some of the problems I uncovered with an overreliance on data across all types of digital analytics:

Problem #1. Response bias

Sure, we use that term more when we discuss survey responses, but I think we can use the same term to represent the problem in digital analytics.

Response bias (or bias of any type) refers to the problem you encounter when your data is systematically different from your population of interest. Using data from your website or email list results in response bias because data comes from a select group that doesn’t encompass your entire target audience only those who visited your website or who subscribed to your email list.

In a client project, I built an algorithm to score email list subscribers and discovered that a very significant number of subscribers weren’t even IN the company’s target audience but were competitors, students, and others who weren’t valuable to the organization. Without this algorithm, the company wasted resources trying to convert folks who weren’t even prospects and likely made content and other types of decisions based on feedback from readers who didn’t matter to the company’s bottom line.

There’s also a fundamental bias when it comes to posts on social media, where most of the conversation comes from a small number of users who don’t represent the consumer population at large. That problem happens with your website, where visits from different devices by the same individual further obscure the true customer journey by fragmenting it across various devices.

Problem #2: Digital analytics are plain wrong

That’s right. You heard me.

In addition to being biased, your data might be wrong — full of errors, duplicate counting, and just plain wrong!

In a recent PPC campaign for a client, I discovered just how inaccurate data is — even data coming from Google, which makes its living by providing accurate data. As a data-driven agency, we monitor analytics on a daily basis — sometimes on an hourly basis. I noticed something funny (or not so funny) in my data — the PPC campaign charged my client for a higher number of clicks than were recorded by Google Analytics (combined with Webmaster Tools).

Which number was accurate?

I never got a satisfactory response from the Google Ads Team.

I noticed a similar problem with Sprout Social. I routinely unfollow accounts that remain silent—why follow someone who never Tweets? I noticed a friend of mine appeared among my silent accounts, but I thought that didn’t seem right. So, I switched over to Twitter and found Tweets as recently as one hour ago. Now, I’m afraid to delete supposedly silent accounts.

Which leaves me pondering:

  • What other data is inaccurate?
  • How big is the difference between my digital analytics and reality?
  • Are my decisions based on real data or just some fiction?
  • Is there some way to fix data problems or even understand the extent of these problems?

Problem #3: Digital analytics don’t speak

Your digital data doesn’t speak—you have to construct queries to answer questions. Construct the wrong query or misinterpret what results mean, and you’ll make bad decisions.

Here’s what Scott Liewhehr told TechCrunch:

Everybody can use data to tell whatever story you want to tell and it’s a big challenge for marketers. If they don’t know how to run studies, they can make a lot of bad decisions.

Ask the wrong question, get the wrong answer, and make the wrong decision.

This means data scientists need an understanding of the firm — its business model, customers, strategies, etc. Which means pairing up data scientists with marketers within the firm or, better yet, training marketers to be data scientists.

The same goes for tools. Tools don’t provide answers; they provide a means to ask questions. Buying another tool isn’t going to magically solve your digital analytics problems.

Problem #4: Correlation isn’t causation

By extension, it bears repeating that correlation isn’t causation — no matter how big the coefficient of correlation is. But, big data offers the tantalizing option of seeing correlations and using them to inform decisions.

As an example, let’s say you pick up a series of click about my search for a new car. You pick up similar signals about me using cookies or other data sources that indicates my lifestyle and predicts income. You then start sharing information encouraging me to buy your high-end car.

But, whoops. You guessed wrong.

I don’t have the kind of income necessary and I’m not a prospect for your expensive car.

Now, you’ve wasted resources and maybe even damaged a potential relationship with me in the future. This happened to a friend of mine just last week.

Problem #5: Behavior isn’t understanding

Watching what people do, even in digital and mobile space, isn’t the same is knowing them or why the customer journey evolved the way it did.

  • I may let someone use my mobile device to find information.
  • I may research a product for someone else or as a gift — something I have no interest in.
  • Maybe I’m researching a term paper, job prospect, or blog post.

Using behavior to infer WHY I chose a particular path along my customer journey is dangerous, but, when combined with big data about other behaviors, the danger is compounded.

Even when we ask consumers why they do something, we might get inaccurate data, but certainly inferring attitudes based on behaviors is wildly inaccurate.

For instance, my daughter is getting married this summer. I’ve been bombarded with emails for wedding albums, honeymoon trips, and other related wedding paraphernalia because I attended a wedding expo with her.

But, I’m not getting married — been there, done that, have the bruises.

I’m not making the decisions. In fact, I have little time or expertise in such matters so she’s got other folks helping her put the wedding together. All a company does by sending me this deluge of unsolicited email is incur my wrath — especially when I ask multiple times to unsubscribe from their list.

Problem #6: Reaching the top

Even when digital analtyics are working well, it’s tough getting top management to make decisions based on insights.

You make your report. Make recommendations. You move on to the next puzzle.

Management hears your findings. Ohs and ahs over your colorful infographics and visualizations. Nods heads appropriately.

Then.

Nothing.

Maybe it’s inertia or maybe fear of the unknown, but using customer insights to guide future plans is challenging for even the most data-driven organizations.

Many advocate for a C-level information officer, such as a Chief Analytics Officer or Chief Data Officer as a champion for digital anlaytics in the C-suite and as an advocate for using information as a tool in strategic decision-making.

Need marketing help to support business growth?

We welcome the opportunity to show you how we can make your marketing SIZZLE with our data-driven, results-oriented marketing strategies.  Sign up for our FREE newsletter, get our FREE guide to creating an awesome website, or contact us for more information on hiring us.

Hausman and Associates, the publisher of MKT Maven, is a full-service marketing agency operating at the intersection of marketing and digital media. Check out our full range of services.


Using bad data and expecting good answers

big data causes big problemsSome data is just dirty and needs cleaning. Maybe some missing values shifted all your columns off — just like when you were in school and messed up the responses on your scantron. Sometimes weird values got inserted. Maybe your key was off when you merged two databases. The bigger the data, the more like you have some dirt in there.

Whatever the reason, you need to clean your data periodically and there are good tools out there to help with cleaning.

A bigger, more subtle problem might exist in your data — it’s not representative. Yet, you make broad assumptions based on skewed data. For instance, Facebook comments over-represent young, affluent, outgoing folks, which don’t represent the total population and might not represent your target audience. The discrepancy between predictions and results from the US presidential election are a good example of how your predictions are wrong when you assume a biased sample represents the whole.

No connection between data and marketing concepts

managing digital dataFor me, this is the most serious problem I see with big data. The people running the analysis don’t understand marketing concepts (at all). Often, these are engineers or data scientists who are great with numbers and stats, they just don’t know what they’re looking for.

I’m working with a mentee now who’s trying to bring more analytics to her marketing role. I think she expected some magic bullet — like go learn R or SQL (yes, I recognize these are challenging software programs, but they’re concrete). Instead, I told her to go back to the organization and determine what marketing goals should drive decision-making. Without this information, I have no clue what data is important.

The same is true for marketing concepts. For instance, I need to understand adoption/diffusion to understand which data help me speed the process. I need to know, for example, that observability speeds adoption, so I know to look for images containing someone using my product to see how well I’m doing visually.

Don’t understand how to use big data

Just as there are too many data analysts who don’t understand marketing, there are too many marketers who don’t understand how to analyze data.

In businesses, firms dealt with increased needs for data by either hiring people who could analyze data (but weren’t marketers) or hiring analysts to provide nice, neat visualizations so marketers could make decisions without having to deal with messy numbers.

As a marketing professor for the past 15 years, I understand how this happened. Students chose marketing simply because it didn’t involve numbers. They liked the touchy-feely nature of marketing that made it a practical application of psychology (which is a very popular major). As marketing became more data-oriented, we failed to incorporate data into our classes.

Now, it’s time to bite the bullet, for marketers to retool with better analytics skills and for marketing programs to become more analytical.

A variety of online school offer classes in data analytics, such as EdX, where instructors from top schools offer classes similar to those offered on campus for a fraction of the cost. So, there’s no excuse to stay ignorant of analytics.