« See all blog posts

Marketing 2018: When Data Meets Science

By Connie Quach on April 24, 2018


Q: What do all marketing executives need?
A: Data-backed proof of the financial success of their campaigns

Q: When do they need it?
A: Twenty years ago.

One of the most stressful aspects of running digital marketing campaigns using cloud-based martech systems is the feeling many marketing executives have that they’re drowning in data but have next to no actionable insight—while also being asked to show proof of a positive return on marketing campaigns.

It is unfair at best and cruel at worst that today’s marketers are then asked to create meaningful predictions for campaign success out of this onslaught of data without the analytical solutions necessary to do so.

In times past, marketers made decisions based on their experience and resulting intuition. Now, they’re being asked to make data-driven decisions and to prove the profitability of marketing campaigns—but without the proper solutions.

How do marketing professionals do it? Well, in truth, without a data science team in place, the answer is often that it doesn’t get done: 39% say their own company’s data is collected too infrequently or with too much latency to be useful.

At conDati, our passion revolves around providing data science solutions to marketing’s most pressing problem: Seeing the profitable storyline in the weeds of digital data. Shining a light on the profitable path to a positive return on marketing investment (ROMI). Knowing that our past successes have provided us with the knowledge, the solutions and the team to provide the needed science for marketing data.

That, truly, is a wonderful feeling. And it’s even better when market innovation dovetails to create the environment where marketing campaigns can flourish.

20 years later, the time has come: Marketers, you have nothing to lose but your spreadsheets.

From Enigma to the Cloud: Marketing Data Science Gets Practical

According to Forrester, 80% of marketers feel challenged by the amount of data they have to deal with and the difficulty of accessing it. All this data is currently stored in the cloud: easily accessible but not easily managed, because it all lives in different silos.

Managing this martech stack can be daunting, even for the most organized marketing team. It was this need—to give marketing teams the capabilities to manage all their data from one real-time, visual dashboard—that drove the conDati team.

Integrating data and providing real-time visual dashboards is the first step: eliminating the “spreadsheet hell” of manual reporting. The obvious next step is to apply cutting edge machine learning algorithms to generate powerful and actionable insights from the data.

Backing marketing campaigns with data-driven reports showing return on investment, alerting teams to anomalies (both positive and negative) and predictive revenue forecasting transforms the deluge of data into practical, accurate and reliable information that marketers can use to create more successful campaigns that ultimately preserve—or increase—their budgets.

Why data science for marketing? Because we can, because it’s time, and because marketing data needs science to improve campaign results—not just describe them.

Marketing Budgets Need Data Science

By 2012, the warning bells were already sounding that marketing professionals were going to have to prove a positive ROMI. However, at that time, “More than 70 percent of CMOs felt they were underprepared to manage the explosion of data and ‘lacked true insight.”

Data from campaigns continued to pour in, year after year, but few if any tools were created to provide actionable insight that marketing teams could use to demonstrate return on investment.

The first generation of attempted solutions were (and often still are) unwieldy general purpose business intelligence tools that lead marketers down labyrinthine paths, often ending with “insights” that hardly helped to prove positive ROMI. Today, the most widespread solution to this problem is to export data from all the different reporting systems into a single spreadsheet—aka “spreadsheet hell.”

Marketing expert Jay Baer of “Convince and Convert” writes that “only 44% of CMOs say, ‘I can measure ROI,’ but this number has to increase if the CMO wants to increase their marketing budget. The CMO must have the right technologies to measure, test, and analyze their marketing efforts.

Wouldn’t that be helpful? What if marketing teams had the technology that financial analysts have had for decades to measure, test, analyze…and we’d like to add…to predict the future success of their marketing campaigns?

Where Data Meets Science

According to IBM, we create 2.5 quintillion bytes of data every day. It’s overwhelming but also the source of a tremendous opportunity for marketers.

What if marketing teams could collect and blend all their data, from all their silos—from Advertising to Analytics to Content and Email to Sales to Social Media—store it all in one place, and then create stories in the form of visual reports and dashboards to share with colleagues?

What if the information gathered from these reports could be seen, visually, in real-time on dashboards, and real-time notifications of exceptions could be pushed to marketing teams?

This is where data science steps in, to provide the innovation that enables marketing professionals to focus on marketing, rather than on rote data collection and reporting. Instead, machine learning algorithms continuously watch all their marketing campaigns, alerting the teams, in real time, to what needs their attention—rewards and opportunities as well as weaknesses, threats, and risks—so they can make the right decisions, at the right time. This effectively stacks the deck in their favor and increases their ability to obtain that holy grail of twenty-first century marketing—a positive ROMI.

With data science on board, the marketer is finally in control.

The Power of Prediction—Marketing’s Key to Success

The real power comes from the science of predictive analytics. With the right algorithms in place, machine learning can produce compelling predictions.

A family of algorithms based on Bayesian inference helped crack the Enigma code during World War II. They lie at the core of nuclear blast simulations. Since the 1970s Wall Street has used them to predict stock prices. And now they can help predict marketing campaign success.

Using these robust and proven algorithms allows for new levels of reliability and predictive power. The predictions are solid, actionable and easy-to-digest. The math takes care of the dirty work, combing through the data and pulling useful insights out that marketers can use to ensure that campaigns produce positive ROMI.

By integrating data across channels, then aggregating that data into real-time, interactive and highly visual reports, marketing executives are now perfectly positioned to know what’s working, what’s not working and just as importantly, to be notified while there’s still enough time to take action.

Real-time reporting is great, but seeing into the future is even better: Forecasting that shows the entire team what will work, what may need to be changed…and why.

Ditch the Black Box: Bring Clarity (& Profit) to ROMI Reporting

If all this has been available since the 1940s, why have we had to wait until now to use it? Why did marketers have to languish with the “black box” described in a 2013 GenPact report?

“Part of the problem behind marketing ROI is associated with the black box used to calculate it. The host of techniques that estimate ROI, and the key word here is ‘estimate,’ require artistic application of statistical techniques, rather than commonly accepted financial principles.”

Until remarkably recently, the sheer cost of corralling, herding, taming and then preparing data for show was insurmountable. Yes, cost effective for nuclear testing. Yes, cost effective for code breaking to save the day during World War II. Yes, cost effective for Wall Street analysts. But fast forward to the new field of data-driven marketing, and the costs could not be justified.

Marketers resorted to their storytelling superpowers to apply “artistic statistical techniques” to ROI reporting.

Fortunately, opportunity favors the prepared mind…and marketers are nothing if not always prepared. Cloud-based data warehousing and server costs have plummeted, making it affordable to retain and store reams and reams of data.

Once that data is stored, applying well-chosen algorithms is the next logical step. Following that up with easy-to-share visuals just makes sense.

And now there’s an all-in-one solution for savvy marketing professionals. Finally, the days of taking creative credit for good times and shouldering the blame for negative anomalies are gone. Marketers can now present the CEO, investors, and board of directors with data-driven insights, showing the evidence that their campaigns make financial sense, while alerting them to required changes in direction before it’s too late.

Even better, the power of communication now rests firmly with the marketing team—for their own story, as well as the stories they tell every day about the brand they represent. Visual communication of data is powerful, and marketers perhaps know this better than anyone. Sharing a strong, well-told story, backed by data and information-rich visuals, pairs perfectly with how marketers already work. The time has come. Data science for marketers is here.