Have you heard of Facebook’s magical moment that created explosive user growth? Their goal was to help connect a user with seven friends in 10 days. Have you ever wondered where that number came from? Or for Twitter, to get an user to follow its first 30 users?
Wouldn’t it be great if your marketing team knew what articles helped get the most sales? With this data, it’s easy for your marketing team to know exactly where they should put their focus. You could answer these kinds of questions if your team could leverage big data marketing.
Here’s what I’m covering in today’s article:
What exactly is big data, data science, and data mining? And how does it apply to marketing, and your startup? To understand big data marketing, let’s take a brief history lesson to help us define our terms.
Data mining first became popular in an article written by Gregory Piatetsky-Shapiro, founder of KDNuggets, Usama Fayyad, and Padhraic Symth in AI Magazine. Data mining was defined as “the practice of examining large databases to create new information.” (Source)
In simple terms, data mining is the process of discovering useful knowledge from data.
Then in 2001, William S. Cleveland brought data mining to a whole new level. He sought to combine data mining with the power of computer science, or data science. So data science is the use of data mining with computer science. (Source)
Around this same time, Web 2.0 came about where we could interact with websites, and each other. Web 2.0 allowed millions of people to connect and interact with one another. As a result, we saw an explosion of social networks like LinkedIn, Reddit, and Facebook. As a result, there was a ton of data created.
The large amount of data from people and computers interacting with one another became known as big data.
That means big data marketing is how one applies large amounts of data to improve your marketing. So why does that matter to your startup?
Have you ever had answers to questions that did not make sense? You wake up to go to the coffee shop, and get that nagging feeling you should grab your umbrella. As you interview a marketer on Skype, after 60 seconds you get a feeling in your gut, she’s the one.
While common wisdom says you should “always trust your gut,” smart co-founders know it isn’t always that simple. Instead, you need accurate data to know which marketing channels are best. Good data gives you the right insight to make smart decisions.
Max Galka, data wrangler, visualizer, and founder of Metrocosm, explains:
"A marketing campaign is the product of countless strategic decisions. Which marketing channels should I use? Who should I target? What message should I communicate? How should I measure its effectiveness?
Smart marketers may have great intuition for answering questions like these. But unless they use big data to test their assumptions, these assumptions are ultimately only guesses.
The challenge is there’s a ton of data. To use that data in a meaningful way, you need to know how to collect it, analyze it, and display it. A second challenge is the data is spread out in many tools and platforms. This means your team is missing out on untapped opportunities without this complete picture."
James Phoenix, a Python analytics marketer explains:
"Big data allows you to spot trends, patterns, and behaviors your competitors will miss. Tools like Segment help you to find insight hidden in your advertising, analytics, sales, and CRM platforms. When coupled with machine learning, the tool will eventually tell you what to do."
What specifically can big data do to help you grow faster?
Have you ever wished you understood your true customer lifetime value? Or how about your exact customer acquisition costs and how to lower it? Big data touches each part of the marketing funnel to improve sales.
A 2013 study asked 171 Chief Marketing Officers where big data has the largest impact on their marketing programs. 58% said search engine optimization (SEO), email marketing, and mobile marketing. 49% said customer segmentation. And 41% said big data is having the largest impact on their marketing strategy. (Source)
I’d be willing to bet most startups benefit from big data marketing, and don’t realize it. Do you do any A/B tests? Then you use big data. Do you use Google Analytics, Google Search Console, Kissmetrics, or Hubspot? Then you are benefiting from big data.
But because of all the data, some of these tools are not optimal out-of-the-box. For example, if you do not make any changes to Google Analytics (GA), you may think Reddit is a bad promotion channel. GA’s default setting measures time on page by counting the time between engagement hits. (Source)
If a redditor visits your blog, reads an article, and closes their browser, the session duration would read 0:00. This would skew your results, making Reddit look worse than it is.
This is where having a data scientist on your team can come in handy. If you’re not sure how to set up GA and can’t afford a data scientist, ConversionXL has an excellent paid course on how to do so. (Source)
Beyond Google Analytics, how else can a startup use big data marketing to grow a business?
Every marketing campaign has the same goal. You want to send the right message, to the right person, at the right time.
Without big data, content marketers have to settle for delivering the right value, for the right audience, at their right time. But with big data, you can create a near-perfect customer buying experience.
"Content marketing is one the highest-ROI [channels] because it delivers something surprising, interesting, and useful to the consumer. This is especially true if you use machine learning (ML) to match the right content to the right user at the right time.
The blending of multiple data sources is leading to an age of hyper-personalization. For example, one customer was searching on her home PC for a specific kitchen gadget. [Later on,] a pop-up ad appeared on her husband’s computer suggesting that he buy this item.
Hyper-personalized content can really move the needle by driving the customer from awareness, interest, and conversion... to having the customer become your vocal fan in the marketplace."
Adding machine learning takes retargeting ads to a whole new level of personalization. But retargeting requires an interested buyer to visit your website. Or if you swap pixels, a partner website. What if you used big data to spot a potential buyer simply from their Facebook photos?
"Using unstructured data enables marketers to "interact" with potential buyers on a 1-on-1 level. For instance, let’s say a person posts a picture of their skiing vacation on Facebook. An insurance company could offer them a short-term insurance policy to cover any potential sports injuries.
The analysis of massive amounts of unstructured data enables "psychological segmentation." Here, machine learning algorithms can apply to target customers with relevant personality traits. Content and ads can tailor to the customer’s preferences, increasing the conversion rate dramatically.
Many growth marketers know keeping profitable customers longer is key to sustainable growth. As Brian Balfour once said, “Growth is good, but retention is 4+ever.” What if you could use machine learning to predict the odds a customer will churn, and stop them before they do?"
“Instead of relying on expenses to reduce customer churn, telecommunications companies are turning to machine learning.
The following graphic shows how risk models can help determine what actions to take to lower churn, based on their probability and risk. [This] allows marketers to consider the level of intervention could affect the chance of churn based on the customer lifetime value (CLV).”
So using machine learning can help you predict who has the highest chance of churning. But it can also help you find who’s worth investing more time into by finding customers with a high CLV.
If you want product marketing services to help you from idea to scale, then you need data to direct your decisions. Otherwise, how will you know if you're investing in a strategy with no ROI?