Rob RistagnoBig data gets a lot of attention. But is it just a shiny object?
On the contrary, in today’s world, big data is table stakes. Companies that do not use data to make better decisions will lag behind their competition. Data makes an undeniably positive impact. According to a study by the McKinsey Global Institute, data-driven organizations are:
- 23 times more likely to acquire customers.
- 5 times more likely to retain customers.
- 19 times more likely to be profitable.
But many companies are overwhelmed by the seemingly large investment needed to take advantage of modern analytical methods. Leaders are worried they’ll need to put years of work and millions of dollars into building out a data warehouse. And then comes the expensive new hires of six-figure-salary data scientists. Plus, the costs associated with new software to make sense of all this data.
The good news is, these fears are unfounded. You can get started no matter what state your data is in, and with little to no budget for the latest in artificial intelligence.
It All Comes Down to Asking the Right Questions
Pablo Picasso is credited with saying, “Computers are useless. They can only give you answers.”
Even with advances in machine learning, if we bank on computers to tell us what to do, we’re going to be sorely disappointed. By asking the right questions, then seeking only the information we need to derive an answer, we simultaneously get more business benefit from our data and reduce the need for significant investments in technology and talent.
As a marketer, what questions should you be asking? We recommend keeping a learning agenda, which is a list of all queries you have about your business. From there, you can create a research and testing plan to get answers, either by crunching the data you already have or setting up a batch of experiments to gather new information. It’s all about testing and learning.
Here’s a starter list of three questions that all marketers should answer.
- Who are your best customers? (Or as we like to call them, your whales.)
- What else are your customers likely to buy from you?
- What are the early warning signs that a customer is going to leave you?
Do you have your answers? How confident in them are you?
Let’s explore how to get more confident using data for business growth.
Segment to Find Your Best Customers
By knowing who your best customers are, you can better target your marketing spend and sales force deployment. You can tailor your messaging, product, pricing, and customer experience accordingly. Identifying and understanding your whales is the bedrock of your marketing strategy.
You can use a methodology called k-means clustering to do this. This statistical methodology has been around since the 1950s, but the ability to handle large data sets quickly and inexpensively is a modern advancement. You don’t need to do the analysis yourself or hire a costly chief data scientist. There are plenty of analytics consulting firms who can help you with this, quickly and inexpensively.
Once you have your segments defined, calculate the acquisition cost and lifetime value of each group. Look at the ratio between the two. As a rule of thumb, the lifetime value should be three times the acquisition cost. Otherwise, it’s an unattractive segment, or something is dramatically off with your marketing mix (price, product, promotion, placement). Usually, it’s not your fault; you’re just barking up the wrong tree. In those cases, shift your focus – and dollars – toward your high-value segments, where the lifetime value far exceeds the acquisition costs.
For a business services company, we found that 50% of their sales and marketing budget was targeting a customer segment with an acquisition cost higher than their lifetime value. Talk about a going-out-of-business plan! How would you like to lose money for each new customer you gain?
By shifting focus to the correct segments, the company was able to drop millions from their marketing budget while simultaneously driving double-digit revenue growth.
Know What Else to Sell
By now, we’re all familiar with the “frequently bought together” strategy deployed by Amazon and other e-commerce retailers. Any business can take advantage of this approach. In our experience, this drives an increase in average order size of 15 to 20%.
Have someone build you a next-product-to-buy model. In doing so, you’ll be able to detect what other products to offer existing customers, either during the shopping or checkout experience or afterward. This application may seem to be for D2C only, but we’ve found great success with B2B organizations as well. For example, if you’re a B2B company, arm your sales team with a data-driven list of recommended upsells they can present to your existing customer base.
Of course, to do this, you need multiple products. If yours is a brand with a small number of products or services, consider investing time in building out your product pyramid for each high-value segment.
At the base of the pyramid should be low-cost products (or even free offerings, like content) to establish your prospect base. Gradually, you should introduce higher-value tiers (at higher price points and margins). At the top of the pyramid, you have your elite offering.
Even though the number of customers who reach the top of your pyramid will be small, they’ll be so profitable that they’ll represent a disproportionate amount of your overall margins.
Use Data as Your Smoke Detector
It’s six to seven times easier to retain a customer than find a new one, but we put so much attention into new client acquisition that we sometimes ignore customer retention. As Ryan Dohrn, CEO of Sales Training World, says, “You need to work as hard to keep your customers as you did to earn their business in the first place. Period.”
This is where predictive models come in. Have your resource build you a churn model. This model will calculate how likely a customer is to abandon you based on their behavior (how they are interacting with your brand, if they had a negative customer experience, how they pay, where they are geographically) and macro-trends like seasonality or weather.
There are two benefits to building this type of model. First, you will get a specific list of customers that are predicted to leave, so you can engage in proactive measures to keep them (a call from the CEO, special offer, or surprise bonus in the mail).
Second, you can understand the drivers of retention, allowing you to make systematic changes to your business model. For one D2C brand, they found credit card auto-renewal customers were so much more profitable than “bill me later” ones, so they stopped giving consumers the option. While in the short term they took a small hit on acquisition, they more than made up for it over time, with a 70% increase in renewal rates.
There is a myriad of ways marketers can use data and analytics to grow their brands without spending a fortune. By seeking answers to the three questions outlined above, you will be well on your way to mastering the art of applying data to make better decisions.
