Of course, we are all talking about big data. And we should be: It promises precision, real-time targeting, improved engagement and acquisition, competitive advantage—and faster growth and greater revenue.
To realize the full promise, however, we must admit that big data is simply the raw material input to the true sworn value: intelligence.
Intelligence derived from big data supports better decisions, informs strategies and corresponding tactics, and powers iterative improvement to more efficiently reach our goals.
Transforming those raw materials into intelligence requires specific expertise.
Big data does not conveniently spew intelligence. A big dataset may be varied and vast, so it requires experts to select and extract meaningful intelligence—and apply those insights to strategies, tactics, and decisions.
Commanders of big datasets have four particular, well-honed areas of expertise: foresight, selectivity, discrimination, and interpretation.
1. Foresight: Identifying the intelligence that will support strategic definition.
Success begins at the beginning: with strategic objectives. Declaration of intention(s) is critical and it must be clear. Setting the point on the horizon will ensure that all tactics, including analysis, support laser focus on the goal. It is too easy to become sidetracked—or overwhelmed—otherwise.
I have worked with clients who are awash in data. Every analytic-driven professional dreams of these rich datasets. These same clients, however, can be crushed by the metrics they can generate.
Without the defined goal, they can easily fall into a common but unproductive practice: They generate haphazard numbers furiously. Everything is everything. Random numbers are set as KPIs (key performance indicators) but no one asks, “Key indicators of what?”
The result is “fun numbers”: good for an infographic or a momentary hallway conversation—but nothing actionable results, and no reliable business decision can be made.
Conversely, if a strategic objective is defined (a doubling of revenue, for example) then big fat data can be applied to project the various sources and scenarios by which the intended increase could occur.
With the necessary intelligence, options, and confidence indicators, informed strategies may be developed.
2. Selectivity: Discriminating so as to introduce purposeful analyses, metrics, and KPIs that support the larger objective.
I worked with a client who wanted to evolve customers from a modestly priced product to a premium product line. The objective was ambitious, as the price increase was over 500%. The client compounded the challenge by setting the objective without the foundational intelligence described above.
The client conducted a series of disassociated, random analyses. Through this arbitrary progression, they identified “best” customers.
It was assumed that those best customers would adopt irrespective of interest, need, or disposable income. There were a few further, equally unsuccessful, efforts to stumble into intelligence. The company marketed the new line at great expense—and with little resulting adoption.
While a few of the analytic results had some application, the use of big data wasn’t adequately precise. It required higher selectivity and greater specificity:
Happily, the client had a clean, robust, and relevant dataset.
Exploratory analysis revealed that adopters of the premium product line routinely adopted a specific mid-priced line first. We had found the gateway product!
The new intelligence opened worlds: We could target mid-priced adopters to boost them to the premium product. And we could analyze modest-priced customers’ behavior to recognize behavioral signals and evolve them to the gateway line.
Suddenly, a progressive adoption cycle was clear, and cross-channel marketing encouraged clients efficiently through the phases of adoption from one line to the next. It was because we knew what to look for and when.
With a single (arguably simple) piece of intelligence derived with correct and applicable data, analyzed with intention, and keenly interpreted results, objectives could be confidently turned to productive—and now valuable—action.
3. Discrimination: Ensuring that what is selected is indeed relevant, complete, and correctly provides the necessary guidance.
Understanding the available data, how it was sourced and stewarded, and how to apply it is a necessary expertise to achieve big data’s great promise.
I worked with another client, also awash in data, who was trying to meet their quarterly sales number. They had analysts projecting top-line numbers—average sales, top performers’ averages, etc.—but could not act on the resulting insights.
They could not act because the insight was not complete. They needed to add new data and generate additional intelligence.
Specifically, the client needed to understand the sales cycle: how prospects moved through the decision phases and, to achieve their quarterly goal, how to do this quickly.
Although there was no data present in the primary database to describe the required timing, there were elements available in a peripheral system. With that source data, and some “frog DNA,” we derived the missing attributes.
Analysis promptly revealed the type of company and size of deal the client could pursue to hit the quarterly goal. The sales teams were focused on those opportunities and it worked well, achieving their objective.
4. Interpretation: Extracting meaning and intelligence from analytic results.
This is the rarest of the necessary skills for big data wranglers. This is where true genius emerges.
Analysis is not circling the biggest number or recommending the next most obvious action. The easy-does-it approach may support tactical definition, and perhaps limited success, but it simply does not exploit true opportunity presented within analytic results.
This is where the promise and the power is realized: creating specific context from findings and results, recognizing their significance, and applying them for optimal benefit.
Analytic results are highly specific to each organization. They are personal. Every company has to work within parameters: resources, budgets, customers, systems, abilities, time, and data. We have to know how to apply them to the specific business.
I worked with a SaaS client who believed that certain industries adopted their software largely to support traffic volumes. However, an analysis of crawler data revealed that an entirely different set of industries adopted their software. The initial reaction was to target those newly identified industries with the existing message focused on user volumes.
But by looking more closely, by getting intimate with the data, a new insight was revealed: These newly recognized industries adopted for a different purpose. Value propositions and messaging had to be created to address their need, capture their attention, and their business.
Quite happily, it worked brilliantly.
Conclusion: Leveraging the power and promise of big data.
I have listed theses discrete skills that masters of big data must develop in order to achieve the great promise woven within vast datasets.
Most importantly, this expertise must exist in combination. I will evoke the analogy of the stool missing one leg. You get it: It topples over if these necessary abilities aren’t collected and cohesive and strong.
In the wrong hands, big data can be harmful. And “harmful” is the correct word.
If we are to effectively base decisions—big and small—on the intelligence derived from big data, we must know how to treat it, how to apply it, and what it means for our business.
Although the right marketing professional able to unearth rich intelligence from big data are rare, they are recognizable: they bring vision, selectivity, discrimination, and critical interpretation.
© 2016 Hampstead Solutions LLC
All Rights Reserved