Which phrase most accurately represents your current marketing strategy?

Data-driven. Data-informed. Data-enabled?

If you’re scratching your head (or rolling your eyes), you’re not alone.

The recent uptick in conversations around data and the resulting divergence has left many a marketer torn about which approach they should be taking with their business.

But instead of taking a side in the matter, I’d argue that it’s less about semantics and more about why these subtle distinctions are being made.

The major complaint against the term “data-driven” is that it insinuates that the data is in charge. The new decision-maker. The end all and be all of modern marketing.

Most data scientists will talk till they’re blue in the face about how insights gleaned from sophisticated algorithms are far superior to those of a human hunch.

Numbers are concrete. They’re accurate. They’re tangible.

But consider for a moment the case of the self-driving car.

The self-driving car is made possible by advanced GPS, navigation and mapping, sensors, lasers, cameras and computers.

In theory, computers and other advanced technologies should be far superior at processing all the factors and information required to safely navigate roadways.

After all, they’re able to take in and process so much more information than the human brain is capable of. They’re also better at recognizing and identifying complicated patterns, in this case traffic patterns.

But where self-driving cars fail is their void of the uniquely human capacity to interpret context, intention and well, humanness.

The same can be said about consumer or market data. There is simply far too much nuance to human behavior for us to ever be completely AI/data-driven.

Now don’t get me wrong.

As someone whose role in demand generation requires a lot of facetime with numbers, I’m not here to vilify or discount data as a necessary or effective marketing tool. I just think it’s important for the success of other organizations, as well as my own, that it remain simply that – a tool.

In order to truly harness the power of data, you have to first recognize and understand its limitations:

Data doesn’t lie, but it can’t tell the whole story

Albert Einstein once said, “Not everything that counts can be counted, and not everything that can be counted counts.”

As complicated and intricate computers and machines are, they mainly offer us very basic and straightforward information – the who/what/when. While knowing that information is extremely helpful, our learnings are limited without the additional context of how and why.

For example, data can tell us how many people liked or shared a post, but it can’t tell us why. In other words it can offer us quantitative results but not qualitative reasoning.

Or think about it this way: If someone analyzed the hours you spent with coworkers, friends and family in a given week, data would suggest that your coworkers were most important to you.

Probably not the case, but a good example of how data can be misleading without the proper context.

Data can also point to a potential relationship between different factors, but it can’t prove it. It’s the familiar adage, “correlation does not imply causation.”

For example, data may show a correlation between a month of high website traffic and high revenue, but that doesn’t necessarily mean the increased revenue was caused by the increased traffic. There could have been a third factor affecting both of these numbers, or another indirect variable.

Only your trained eye and experience will know to approach these numbers and metrics with caution and do further testing. If you discover a correlation in your data, try digging deeper to either replicate your results and isolate the true cause, or segment in various ways to see if different patterns emerge.

It can also be helpful to gather qualitative feedback from methods like site and email surveys as well.

Data is a realist, but it can’t take risks

Several years ago, 29 year-old Morgan Hermand-Waiche set out to buy his girlfriend lingerie for her birthday.

Once he discovered just how expensive most of his options were, he realized there was a serious gap in the market for an affordable lingerie company and immediately began researching a possible venture opportunity.

The problem? Data told him to stay as far away from the lingerie business as possible. There was a clear industry kingpin dominating the marketplace, countless barriers to entry and numerous failed attempts including several big-name brands.

But despite his findings, Hermand-Waiche couldn’t ignore the one thing still pushing him to pursue this venture: His gut. There had to be a market for affordable, quality lingerie – even if the data suggested otherwise.

Hermand-Waiche is now founder and CEO of Adore Me, an e-commerce lingerie company revolutionizing the industry. In just a few years, he turned his gut feeling into Inc. 500’s #2 fastest-growing company in NYC and has raised around $11.5 million from VCs and private investors.

Data can only tell us the current state of things, and at best make informed predictions.

Try utilizing more qualitative methods like posing questions/polls on your social channels, social listening or even good ‘ol focus groups to get more honest, intimate feedback on an idea you may have.

And remember, sometimes revolution just means ignoring the status quo and taking a risk.

Data can inform, but it can’t imagine

Repeat after me: Big data is not the big idea.

Data did not come up with “Just Do It” or tell Apple to “Think Different.”

It’s all too easy to get caught in the weeds of numbers and statistics, but remember that great marketing is about telling a great story – and telling a great story means understanding human behavior, emotion and experiences.

We can learn all sorts of things about our audience’s actions from data. But it can’t tell us about their motivations, their struggles, their desires, etc. We need those uniquely human insights to tell great stories and be creative.

But it’s not data’s fault.

Creativity is an art. By its very definition “art” is the expression or application of human creative skill and imagination, producing works to be appreciated primarily for their beauty or emotional power. Keywords here being “human” and “emotional.”

Case in point: In 2016, the University of Toronto Computer Science Department tried to teach a computer how to write a song.

Researchers fed the machine over 100 hours of music while a sophisticated algorithm “learned” patterns in the beats, chords and lyrics. And while all that sounds impressively high tech, the resulting “song” was somewhat of a disaster – with strange, nonsensical lyrics and an uninspiring, robotic melody.

Turns out, data is a pretty crappy composer.

The good news is, there is a way data can provide the type of human, emotional insights that inspire great creative. But instead of listening to the numbers, you’ve got to actually listen to people.

The most recent advancements in social listening tools allow brands to discover things about their audiences that might otherwise take months of qualitative interviews. Topic affinity is a great example of a listening capability that is much more impactful than most people realize.

Imagine the doors that can open when you find out what else your audience is talking about on social. Do they love a certain kind of music? Or sport? These insights can lead to new sponsorship opportunities, product integrations or even a brand new audience segment.

Another good example of how social listening can inspire great creative is through sentiment analysis. Learning the way your audience feels about something like recent news or relevant topics gives you the opportunity to create content or campaigns that will resonate with them on a deeper, more emotionally resonant level.

Coca-Cola used sentiment analysis to create their Coke Tweet Machine. Using natural language processing and location, the brand was able to identify the least happy city in the country.

In keeping with their brand strategy, “choose happiness,” they brought a Coke vending machine to the city that analyzed the sentiment of each user’s Twitter profile.

The machine then only dispensed a can to users with a more positive, happy presence on the platform.

It’s amazing how brands can learn and create so much simply from analyzing people’s activity on social. Try doing some of your own social listening the next time you’re looking for creative insights.

So whether you consider your organization to be data-informed or data-driven, all that really matters is that you’re leaving plenty of room for humanity in your decision-making. Because data without humans isn’t insight – it’s just numbers.