The secret weapon of a CRM agency: the data scientist

“March, we have a problem.”

This could very well have been the message sent in 1999 when Mars Climate Orbiter caught fire and exploded. What should have been a celebration for NASA engineers has turned into a disaster. All because of the data, in this case the incorrect measurement of units.

Who or what was the culprit? The Jet Propulsion Labs navigation team used the metric system. millimeters and meters. Lockheed Martin Astronautics, who designed and built the spacecraft, provided data in inches, feet, and pounds. As the team sounds like a scene from Iron Man, there was no Jarvis to reconfigure the units to milliseconds. The result was catastrophic for the Orbiter.

Everyone has data. It’s what you do with it that really matters. Data can be lost, obscured, inaccurately analyzed, or ignored. Add many parties to the equation (multiple vendors, agencies, business stakeholders, data sources) and the potential for data crime accelerates. So how do you optimize these most valuable assets?

The answer is as elegant as the discipline itself: data science. Data science is the art of turning existing data into useful information that businesses can use to make decisions. Artificial intelligence is at the heart of the data scientist’s toolkit, which strives to enable machines to mimic human intelligence to perform reasoning. Memorable examples of AI onscreen are 2001‘s HAL 9000, Lieutenant Commander Data from Star trek and Arnold Schwarzenegger in all Terminator movies. Machine learning, a subset of AI, can drive value for your CRM by applying predictive analytics to drive significant growth. This is where the real advantage exists. CRM agencies leverage this advantage to segment, be predictive, personalize, and optimize channels.

Think about the amount of data that exists in the digital age: smartphones, Internet of Things, connected vehicles. Then take social media into account. There are 500 million tweets per day. Five billion videos are watched daily on YouTube; 500 terabytes of data are shared on Facebook daily. Billions of data location pings come from smartphones every month. Information derived from artificial intelligence is not only useful; they are essential for traversing billions of data points to divide sentiment, propensity, advocacy. Old-fashioned research companies and focus groups would investigate sentiment towards a product or campaign. In the new world order, the data scientist is your personal assistant.

But, at the end of the day, data is about understanding people. Here are a few examples of brands using data science to leverage information to their advantage.


In predictive analytics, there is the infamous Target case study. An angry father confronted a Target official with a flyer for baby clothes and a crib. The leaflet was sent to her teenage daughter. The father reportedly asked, “Are you trying to encourage her to get pregnant?” Several days later, the father apologized because his daughter had confirmed the pregnancy. What came into play here was the brilliant ability to consume data, compile behaviors, and predict a life event – pregnancy. Not only could Target predict pregnancy, but its algorithm could also report pregnancy stage based on purchase. Prenatal vitamins at 20 weeks; six months unscented lotion; fragrance-free soaps, giant cotton ball bags and hand sanitizers just before the due date. What could be better than knowing your customer’s needs? Receive coupons for the items you need when you need them. Data science couldn’t have been more precise than if the customer were buying What to expect when you expect.


Amazon has set the standard for big data-driven personalization. Consumers can buy anything from Amazon, but the choices can be staggering. To reduce the churn rate and help consumers get what they really want, Amazon has adopted a collaborative filtering engine. CRMs strive for a 360-degree view of the customer. And Amazon takes this holistic view seriously. The collaborative filter engine analyzes previously purchased items, online shopping cart or wishlist, reviewed and rated products, and frequent searches. Then the personalized recommendation system offers products that people with similar profiles have purchased. This method generates 35 percent of the company’s sales each year.

To enable faster shopping and reduce distractions, Amazon will tell you everything about ordering the item. Is it eligible for free shipping and when? Is it in stock? And it offers quick one-click ordering. Amazon also packs the shopping experience with product images submitted by customers, along with Q&A and reviews from verified buyers. And if a review appears in another language, you can translate it into English. Thank you for suggesting buying a size larger!


Peloton was taking the fitness world by storm long before COVID-19. Several factors make the success of this $ 4 billion start-up. Time: lack of time to go to the gym. Unused fitness equipment: stationary bikes stored in the basement. Competition: remote spinning that ranks you against a virtual exercise community. Social networks: bragging rights on social networks.

Building on the spin craze, Peloton was founded in New York City in 2012. The revenue model consists of two components: the actual stationary bike and the monthly subscription for virtual lessons. The inspiration factor kicks in by using the client’s own data against rivals. Joining the community can be an initiation. Users create their exclusive nickname, under which they are known to the virtual community. Fun, pun-oriented names are encouraged, unlike alphanumeric names. (Who wants to shout, “Congratulations on ranking number ten, MZG1991!”?) The name of Patrick Mahome’s Peloton is 2PM. Spinners provide stats like age, sex, and weight, and while it may seem intimately intrusive, it pays off later as it more accurately measures calorie burn and ultimately leads to the coveted leaderboard of the leaderboard. Bikes track performance and remember preference settings. Instructors shout during class to give props, recognize birthdays and galvanize their students.

Then, the power of competitive data takes effect through gamification. Spinners are invited to compete against each other and follow their progress live. It’s like playing a video game – all the adrenaline, visuals, and challenges, except you’re not sitting on your couch, you’re breaking your ass. You can also send high fives and video chat during workouts. Excuse me, my little cousin burns more calories than me? Talk about primary motivation. And loyalty. The more the spinner takes root in the Peloton community and feels included, the more he uses the service. The retention rate is 95 percent. Word of mouth is one of the most lucrative ways for Peloton to acquire users.

Peloton shows how data innovation can succeed even in extreme crises: confined to your home with your own sanitized bike, while you connect with people like you.


This brings us to data cleanliness. Part of the beauty of understanding the more technical parts of data science is the foundation of data hygiene. Data integrity is always an issue. It makes absolutely no sense to add AI or machine learning if the data itself isn’t clean. And what does that mean? No duplicate records. Unique customer identifier. Always fresh data.

Data cleanliness is the path to data piety! What does it mean?

First, it’s critical that your data successfully identifies a unique and distinct person, regardless of how they enter your data sphere. Is the Alex Garcia who visited your website on Monday the same as Alex Garcia who ordered an item in Jersey City, New Jersey on Friday? Signals in customer data that allow a brand’s customer data store to successfully identify and merge the actions of a separate customer are essential to machine learning and artificial intelligence. Businesses need to invest intelligently in Customer Data Platforms (CDP) and back-end system integration to achieve the goal of a unified view of every customer.

Second, it’s just as essential, if not more, to ensure that your customer data is refreshed and updated on a regular basis. Email lists, according to email validation service FreshAddress, are degrading at rates of up to 25-30% each year, and similar aging rates are seen for phone numbers critical to business. SMS messaging, location analysis and other data points. It cuts you off from the precise fusion of customer data and key channels for effective, personalized offers and branding. It is essential that your data management strategy creates a strong database that captures and recaptures current customer contact data.

How does the data scientist put this together? By understanding the data and using its validity and recency to create a holistic picture of customer behavior, feelings and actions. This information can be generated to create personalized content, offers and calls to action, as well as strategic support for back-end operations, including supply chain management, inventory, l operational efficiency and future projections. Ultimately, the goal is to effectively use customer data as the foundation for all aspects of a brand’s operations.

While data science is science, the instrumental factor is humanization. There’s a reason a CRM agency’s secret weapon is the data scientist. This discipline allows marketers to answer fundamental questions. Who to target? When to target How to predict behavior? How to remember? Which channel? Don’t get lost in translation like the hapless Mars Orbiter. Or worse, leave money on the table by misjudging your data. Use your secret weapon: the data scientist.

Rekha Gibbons is COO at Raare Solutions. She evangelizes the customer journey with B2B2C companies seeking to increase sales and brand loyalty. For Gibbons, the secret sauce for increasing conversion and loyalty is a flawless customer experience. She nurtured this philosophy at companies like AIG,, Jaguar, Land Rover and Lindblad Expeditions. Gibbons is a sought-after speaker on topics such as marketing in times of crisis, pivotal customer journeys and inspirational marketing. She lives in New Jersey with her husband and their three-legged COVID rescue dog, Sonny.

CRM covers

Sean N. Ayres