Organizations are increasingly adopting big data analytics to understand and then fix business problems. They’re learning how to extract value from multi-sourced information and then relate that information to an issue in their marketing, manufacturing, advertising, or shipping, etc.
For example, T-Mobile (and the other main carriers) consistently use big data analytics to spot the reasons for (and prevent) customer turnover. Customer attrition is a significant expense in this industry, so firms that can best use data to improve retention and improve satisfaction will have a leg up on the competition. Big data is not just a tool for the enterprise level firms. It’s appropriate for businesses of varying sizes that want to better understand customer behaviors and improve their marketing tactics.
Every organization wants to use data to find actionable insights. It’s “cause and effect” on a broader scale, where there could be multiple factors working in concert that are producing a certain result. The difficulty is in generating the right data, keeping it organized, and then having the right analytics tools and staff members who know how to extract correlations. Doing this right to optimize the customer experience and boost sales requires adherence to several best practices:
Use Statistical Modeling
Marketers working on TV campaigns now have at their disposal a number of modeling tools to help them gauge performance. They can use customer demographics, Nielsen-derived viewing data, airing size, and specific data on the actual stations utilized and the airing timeframes. Marketers can use this clean data to gauge current performance and then dynamically adjust future campaigns accordingly. There can be surprises uncovered in this process, as marketers might find for example a previously under-served demographic that is generating impressive sales in response to TV campaigns.
Clean up the Mess
The “mess” in this context refers to data that is not properly structured and is essentially useless when it comes to analysis. Even the best data scientist and marketing wizard can’t pull insights from broken data. Do some work on the front end to ensure all of the data streams coming into the analytics tool are organized and clean. Attempting to fix data after the campaign is launched is a time-intensive process that doesn’t give the marketers a chance to suggest campaign changes in real time.
Follow Customer Behavior and Actions
Companies have at their disposal a powerful but often underused source of rich data. It’s the touch point for most customers – the website. Whether it’s a landing page, mobile site, or the primary corporate website, all of these channels offer a wealth of information. Firms can track this information through pixels that can be placed throughout the sites to measure customer behaviors, from what they visit to how long they hover the cursor over the “add to cart” button.
Understanding customer behaviors provides unbelievable context and the opportunity for segmentation. Marketers should develop “playbooks” on consumers so they can then be grouped together in new and surprising ways. Tracking should also include device information, especially as consumers move to a mobile-centric way of communicating and ordering. Social media tracking provides another layer of data on how customer’s share information about a brand and gives marketers a way to reach social media influencers.
Measure the Retail Responses
Big data analytics is essentially a new way to look at “cause and effect.” Instead of reviewing a direct mail piece’s performance against actual sales, big data analytics can correlate seemingly unrelated company actions and customer actions.
Marketers should focus on the retail responses amongst the various channels to spot these surprising correlations. With many campaigns, there can be a brand boost in sales for a campaign that is meant to only promote a single product. For example, TV spots about the durability of a housewares maker’s blender could drive sales of the firm’s carbonated beverage machine. Marketers will need to use analytics to look deeper at the behaviors and actions behind such results, and then adjust accordingly to boost sales of both products.
The ROI of campaigns can now have multiple layers, so marketers should understand how to analyze campaign impacts on a deeper level. Marketers that take a measured (yet creative) approach to big data analytics will be the ones most likely to uncover surprises and be able to prove campaign ROI. Doing this right requires some patience and hard work on the front end to introduce clean data, build structured analysis models, and create analytics that are built for multi-channel environments.