From mysterious buzzword to realized potential, the concept of “big data” is now well-known to marketers. A growing and maturing landscape of Big Data vendors now make it easier-than-ever for marketers to apply “big data” to their strategies, campaigns, and customer experiences.

The concepts within and around Big Data still drive confusion for many marketers.

Three Big Questions

I’ve been asked these questions by startups during the Google-sponsored “marketing hackathons” my friends and I run. I heard them in my past life leading digital at MIT. I am still asked these questions by our Fortune 50 clients here at our brand experience agency, Cramer.

These questions are still very common and still very much asked by marketers within many different types of business:

1. What is Big Data?
2. What are the challenges around Big Data?
3. How can Big Data become meaningful in our marketing?

While each question can — and has — been addressed ad nauseam around the web. For those who like streamlined lists and simplified frameworks, here are my referenceable answers to each question.

What is Big Data?

The Five V’s of Big Data

The “Vs” of data have been around for a few years now. Some people will tell you there are just three Vs, while others, like IBM, champion four Vs.

For me, the conceptual framework of Big Data’s key concepts requires five Vs. The five Vs are:

1. Volume: The quantity of data that is generated and stored; this is the V most directly associated with the term “big data”
2. Velocity: The speed of data generation and processing
3. Variety: The structured and unstructured forms of the data
4. Variability: The consistency, and often lack thereof, of the data
5. Veracity: The accuracy, fidelity, and quality of the data

When explored individually and holistically, these Vs set the framework for a marketers’ strategies and tactics for applying Big Data to their businesses.

What are the challenges around Big Data?

The Four Big Challenges of Big Data

Once marketers grasp the concepts behind Big Data; they find that their strategies and tactics must address Big Data’s four core challenges: aggregation, access, attribution, and analysis.

Aggregation and access and are straightforward concepts: these are the challenges around capturing, storing, scaling, and then, once aggregated, get hands-on access to the data.

Attribution — tying a business outcome to a particular set of historical actions — is a challenge faced by all businesses, especially those focused on direct and performance marketing. For well-trained data scientists, nights-and-weekends data hackers, and the marketers in between, there are now powerful platforms that open advanced data theory and statistics models for their use. Even free tools like Google Analytics, now include robust attribution modeling. While still far from perfect, not fully understood, and underutilized by marketers, it is now easier than ever to address attribution challenges with Big Data.

Analysis is the fourth, hardest, and most critical challenge within Big Data. Analysis unlocks the insights that inform real-time actions and future-looking strategic decisions. From enterprise-grade software to an ever expanding portfolio of analytics-focused startups, marketers now tap into powerful platforms that support and accelerate the analysis of data. Plus, the current advancements in technologies such as artificial intelligence and machine learning will soon usher in the next evolution of analysis within marketers’ technology stacks.

How can Big Data become meaningful in our marketing?

Two Recommendations for Analyzing Big Data

Of the four challenges marketers face within Big Data, the challenges around analysis are typically the most discussed and the most questioned. These challenges had existed since long before “big data” was a thing. With this extensive history and a rapidly expanding and improving toolset, all marketers are now compelled to do some data analysis.

Many businesses employ data scientists, marketing analysis, or agency partners whose sole purpose is to help marketers go deeper into their data analysis and unpack the insights and actionable recommendations needed by their businesses.

What if your business does not have access to these unicorn-like resources? How can you go beyond the canned reports? Where should you be going deeper into your data? How can you find the stories and the insights your business needs?

There are two places to start.

Data Needs Context

Without context, numbers lack meaning.

Is 100 good? Is 10,000,000 bad?

Without context, there is no story for your data. Moreover, without a story, data remains passive with its actionable value still untapped.

The first place to look for context is through peer comparisons. By sourcing public and syndicated data from your competition and industry, your comparisons and resulting stories become straightforward and meaningful to your entire organization. However, this type of data is rarely available.

Most marketers are left looking for the context within their pool of data. This means context is best driven by historical analysis.

The types of context-creating, historical-based comparisons that marketers should explore:

• Time-over-Time. This provides a straightforward look at changes and the impact of time-related events (e.g., Past 30 days compared to the previous 30 days).
• Benchmarks (or averages). At these most basic level, these provide a much-needed baseline for analysis (e.g., “Q1 was 15% above our quarterly average for visitors from paid search”). For simple-but-a-bit-more-advanced analysis, by looking at your Bell Curves, the distribution of your data. The straight-forward plus/minus one standard deviation from mean provides you with a benchmark (or average) that is a bit more forgiving to time- and event-based business fluctuations.
• Indices. Any marketer who has worked within the paid media space is familiar with the power of indices to help tell a data story. I find that marketers often do not think to apply the power of Indices to metrics outside of paid media. This is a lost opportunity. By building your indices so you can see clusters — particularly when things are over/under-indexing — and unlock compelling new ways to frame (and reframe) perceptions around your data.
• Trend Line Forecasting. Once you have a good set of historical data, you can do simple forecasting and trend analysis. For example, within Excel, the simple R‑square-oriented trend lines can be extended into the future. This gives marketers decent forecasting data to showcase how their campaigns and customer experiences will continue to grow.
• Growth of the Growth Curve. This is my favorite and still surprisingly rarely report. It was born out of past experiences where my teams and I were part of many fast growing, highly successful businesses and marketing programs. After a stretch of “we are growing at 15–25% per month” that metric begins to lose meaning. So we started to look deeper into it to find new data stories; what we found is the compelling Growth of the Growth Curve concept. This would tell us not that we had growth, but rather the health of the growth. As a marketer, say your social following is growing steadily each week, take a look at how your growth rate is changing over time. You are sure to find insights and actions within that analysis.

Structure Reports to Business Challenges, Not Data Buckets

Most of the software marketers use for data analysis is now sophisticated and easy to use. The software does a strong job breaking out its reports into pre-canned buckets. However, these buckets are often based around generic types of data (e.g., product sales, page views, clicks); these reports are rarely something a marketer can directly use to answer business questions.

As soon as you start using a new piece of analysis software, my recommendation is not to dig into its pre-canned reports, but rather start building your own. Be “business challenge first” and create your reports from there.

For example, most marketers could build reports that look to answer these types of questions immediately:

1. Acquisition: Are you attracting new users? If applicable, new sign ups? How? From where?
2. Retention: Are people coming back? How often? How soon (time between visits)? How do returning visitors differ from new? How do heavy-return-visitors (individuals who often visit, e.g., weekly), compare to people who come less often (like daily)?
3. Engagement: What are they doing? How many pages/screens are they seeing? Where are the clusters?

In the end, it is fairly obvious to all marketers: That faster we see data in a view that aligns with our business needs, the sooner we can make business decisions based on that data. Therefore, as soon as you start using a new-to-you piece of analysis software, don’t expect it to be able to have pre-canned answers to your key questions — be ready to build your own.

Going Forward: Evolving this Framework

From understanding the concepts behind Big Data, to addressing the challenges around Big Data’, to going behind canned reports and unpacking meaningful analysis from within Big Data, this story pulled together frameworks my teams and I use when discussing and using Big Data-oriented projects with our clients.

If you have other frameworks, approaches, and recommendations, then please send them my way — I would love to see them.

One More Thing.

Complete this Analogy.

The Internet is to the World Wide Web, as the Internet of Things (IoT) is to…?

When the World Wide Web emerged it became a marketing channel on top of the Internet. What will be the equivalent marketing channel on top of the Internet of Things?

It is called The Web of Things.

It has arrived, and it is changing the marketing landscape.

I wrote an in-depth piece about the Web of Things’s arrival. Want to give it a read? The booklet is packaged with two other brilliant pieces written by our crew at Cramer.