One of the major differences between social media and traditional media is DATA. In traditional media, users struggle to find data to support the success (or failure) of their campaigns and develop insights to guide future decisions. In digital marketing, we’re literally DROWNING in data. Knowing what matters and how to generate insights from your data is key to your firm’s success. Recent studies find only about 0.2% of the mountain of data generated each year gets analyzed. A valid question is how to use the other 99.8% of the data, or some meaningful fraction of that amount, to guide businesses toward better decision-making. The opposite also happens, where data is analyzed into graphs and tables that provide meaningless mountains businesses must weed through looking for insights — something I call puking on your data. Today, we’ll discuss developing market information that leads to meaningful insights and long-term success.
Developing market information is more a question of weeding through mountains of data to find valuable insights. Add to this the complication that 80% of data is unstructured (text as opposed to numbers), and you see the problem even the savviest analytics find in understanding how to proceed.
Too much data
Take a look at this infographic, which shows the amount of content ADDED to the internet in just 1 MINUTE. Obviously, NO ONE can keep up the pace necessary to analyze all this information — even for a single minute. Even Google’s spiders struggle to crawl all the new content uploaded.
According to IBM, data possesses 4 qualities that make it difficult to analyze:
Data comes at the world in terms of Terabytes (TB). For instance, the NY Stock Exchange generates 1 TB of data on trading every day. The average car contains more than 100 sensors generating data constantly. And, the average person generates 1.7 megabytes of data every second; that’s nearly 150K megabytes of data per day. And that’s per person!
Analyzing data in real-time is a huge challenge for modern businesses, even with the understanding that an individual firm only captures a fraction of the data generated every day. But, a large firm like Amazon, with its commitment to data, likely processes as much as a TB of data every day when you consider they process sales, record new products offered by sellers on the platform (including video) as well as their own, update customer search for remarketing purposes and to improve recommendations, and manage the billing for subscription products. That doesn’t even include all the data handled by clients on Amazon Web Services (AWS), which is rapidly becoming the biggest data warehouse in the world, outside of governments.
IBM estimates there are 40 Zettabytes of data as of 2020, and, as you can imagine from our discussion on velocity, the pace of data creation means the cumulative volume grows rapidly. In fact, the cumulative amount of data grew 300 times between 2005 and 2020. Even as far back as 2015 (which is eons in digital time), experts estimated the amount of data available would double every 2 years.
Possibly the biggest challenge businesses face is the variety of data they must analyze. Hardest of all, 80-90% of all data is unstructured (text) coming from places like social media posts, content uploads, comments, etc. And, despite the concerted efforts of many analytics companies, our ability to analyze unstructured data is pathetic, with 95% of businesses citing problems evaluating unstructured data in their businesses, according to Forbes. Even determining the sentiment behind utterances is terribly difficult because non-verbal communication missing from written content often changes the meaning entirely, especially for languages like English.
Of course, we’re getting better at handling some unstructured data. For instance, most larger companies use chatbots as a more personal form of answering questions than placing FAQs on their websites. Also, chatbots never take a day off or get sick, so they’re available 24/7; 365 days a year. Powered by artificial intelligence (AI) and machine learning (ML), these chatbots use NLP (natural language processing) and are pretty effective at answering questions while providing the feel of human touch to the interaction.
Check out this graphic from Wordstream depicting the common chatbot application.
Is your data accurate? According to IBM, the question is commonly asked by managers, with 1 in 3 managers questioning the accuracy of the data they receive. In fact, IBM estimates that poor data quality accounts for $3.1 trillion in lost revenue for US businesses alone.
Certainly, automated systems reduce the problem with data quality by transferring data into centralized storage locations, such as AWS, which reduces the introduction of humans into the processes and the errors they make in data entry. Still, businesses experience hiccups in their data collection, and humans analyzing the data make mistakes by using improper techniques.
Obviously, you need tools and processes to deal with all this data if you hope to glean insights that guide decision-making. So, read on to see tactics for developing market information.
How to: Developing market information
Notice, I don’t use the term marketing information system in our discussion today. That’s because what I’m talking about is distinct from a marketing information system, which is often somewhat automated. I’m talking about developing market information to guide decision-making from the mountain of data available to managers. Once developed, managers can automate the process with a marketing information system.
Here’s the process I recommend in developing market information into meaningful insights.
- Set goals
- Create KPIs
- Define metrics to assess KPIs
- Choose marketing information system software
- Transform data into insights with appropriate analysis, such as pivot tables, correlations, etc.
- Use the analysis to make decisions
The first step in developing market information is to set goals and objectives for the organization. What do you hope to achieve as an organization? Greater sales? Better customer service? Improved brand reputation? Likely, your organization has multiple goals, and that’s great, but they should all be SMART goals.
From goals come objectives, which are more concrete than goals. Objectives set out specific actions designed to help the business achieve the overarching goals established.
Next, identify how digital marketing tactics (or a combination of traditional and digital tactics if you plan to combine both) help the business achieve those goals. Build a hierarchy leading to your goals. For instance, if you hope to achieve greater sales, identify the steps necessary to achieve that goal. For instance, to increase sales, you must increase brand awareness, improve brand image, generate a higher CTR (click-through-rate) on social media, better performance on keywords, a consistent content marketing strategy improved performance on email marketing efforts, and smooth progress through the sales funnel created by understanding the customer journey then matching it to efforts to nurture leads.
Determine which metrics effectively assess each action; we call these KPIs for key performance indicators. For instance, assessing smooth transition through the sales funnel involves assessing how visitors move through pages represented by the funnel — evaluating exit and looping that occur during the process. KPIs must both assess how well the organization is doing toward achieving its goals and offer insights to improve performance.
Beware as not all data represent KPIs. For instance, the size of your social network only loosely correlates to achieving your goals. A better metric is engagement (comments, RT, shares) since engagement amplifies your brand messages as well as represents greater intention to purchase than simply following, as depicted in the graphic below.
Marketing information systems
Choose marketing information system software that provides the most information regarding KPIs, with high accuracy (many tools are only about 70% accurate) at a reasonable price. For instance, Salesforce and its modules, such as Pardot, provide metrics for CRM, while Google Analytics is a free tool for monitoring your website performance. Try to use marketing information system software that provides metrics for multiple KPIs in 1 platform. But you’ll still need to combine data from multiple data sources, such as internal data, data from information software, and external data such as economic activity into databases, then create custom queries using analytic tools like Python. Once you create reports, you can automate them then combine visualizations to drive insights using tools like IBM Cognos to create dashboards or Tableau to create visualizations from raw data.
Your software should do a good job of COLLECTING data, but transforming the data into something that makes sense can still be challenging because you’ll have tons of data. Creating custom reports that help you make sense of the data lead to better decisions. For instance, you might use filters on a Google Analytics report to eliminate results with small numbers. Hence, in analyzing traffic sources, instead of looking at hundreds of sources, you might filter to look at only those sources sending more than 50 visits to your site. The same goes for analyzing pages driving visits to your site — filter to view only those pages sending visits that surpass a certain threshold. Using pivot tables generates covariance — or crosstabs — showing how 2 variables in your reports are related. For analyzing conversions, use multi-channel attribution to apportion credit across the multiple channels accounting for visits that generated revenue.
Charts and graphs help visualize data, which often eases interpretation, as mentioned above. Don’t forget that many metrics are more valuable when viewed as trends over time rather than static at a single point in time.
Use the data. If you can’t use the data to guide decision-making that improves your market performance, there’s no use in collecting the data in the first place or analyzing the data.
This means getting data to decision-makers in a form they can understand. Some market information system software allows each user to create custom reports and views to satisfy their particular needs. For instance, a manager might view the performance of certain brands while the CMO might need a report detailing the performance of the entire portfolio of brands.
Some managers are less savvy about business intelligence than others. You might need to provide written reports to certain managers detailing data interpretation to guide their decisions.
Developing market information. Analyzing data collected from multiple sources, generating insights from that data, and making informed decisions greatly impacts your performance. While some aspects of developing market information are more technical, such as using Python, most are less sophisticated. The key is to create a thoroughly considered strategy for marketing information to ensure you have the right information, you analyze it correctly, and form the correct interpretations from the analysis before basing decisions on the results.
Need marketing help to support business growth?
We welcome the opportunity to show you how we can make your marketing SIZZLE with our data-driven, results-oriented marketing strategies. Sign up for our FREE newsletter, get our FREE guide to creating an awesome website, or contact us for more information on hiring us.
Hausman and Associates, the publisher of MKT Maven, is a full-service marketing agency operating at the intersection of marketing and digital media. Check out our full range of services.