Sherlock Says: How to Create a Predictive Marketing Strategy
in Digital Marketing Strategy & Trends | by Erika Levy
A few years ago, a job title like “data scientist” sounded more like a wacky Mad Lib insert than an actual career. But now, the science of interpreting complex data is emerging as one of the fastest-growing and most promising careers in marketing. Why? Because these “Einsteins of data” use predictive modeling software to look at micro-patterns in customer data, offering a gold mine of insights for better business decision-making.
Here’s the challenge: predictive software can be a little pricey, and hiring a data scientist can be a tirelessly long and expensive process. But have no fear—you can still harness the power of predictive marketing by taking a deep dive into your existing data; no advanced degree required.
You can find all the juicy secrets to ROI success in our guide, 5 Steps to Marketing ROI Stardom, but we’ve pulled together a few tips here to refine your predictive marketing strategy:
Learn From Historical Data
Closely Review Your Database of Past and Current Customers
If you sell directly to consumers, you should analyze demographics such as your average customers’ age, gender and location. For example, if you specialize in B2B, you’ll want to identify your average customers’ industry, business size and location.
You’ll also want to assess the ways in which they have done business with you. What did they purchase and how much did they spend? Was it a one-time event or have they renewed their contracts?
Remember: aggregated information is great, but it will be more helpful when looked at in segments. If you are using a CRM platform to house your customer data, run a few reports that show:
- Product and Average Spend for Age/Gender (B2C) or Industry/Business Size (B2B)
- Multiple Purchases or Contract Renewal rates for Age/Gender or Industry/Business Size
Get as granular as possible to yield the greatest insight into specific behaviors across groups.
Look at Channels & Campaigns
While reviewing your audience, you’ll also want to look at how they found your business. What channel did they use? Facebook? Email? A phone call? If you aren’t tracking lead sources in your CRM, get on it. You’ll be thankful you have this data 6-12 months from now.
In addition to knowing how many leads are coming from different channels, you’ll also want to know if they were a part of any specific campaign. Were they drawn in from a promotional offer? Did you host a semi-annual campaign that attracted more attention? Knowing where—and why—leads are coming to you will help you replicate success.
If you use marketing automation to track at the campaign level, investigate the following metrics for your most successful campaigns:
- Lead Conversions
- Individual Visitors
- New vs. Returning Visitors
- Form fills/Data entry
Again, if you’re unsure how leads are entering your database, you’ll want to focus on closing those gaps for a more accurate predictive marketing strategy.
Segment Your Audience
Once you’ve gathered as much data as you can, you’ll want to segment your audience in a way that makes sense for your business. Whether this means looking at buyers by product or the number of leads generated by each channel, the key is to walk away with 3-4 “big picture” takeaways for each segment to inform your digital strategy.
The key to successful digital marketing is personalization. Study and learn your current customers, and you’ll create more direct and personal touch points with maximum effectiveness.
Glance Into the Data Crystal Ball
Once you gain a better understanding of your customers and your organization’s past campaign performance, the next step is to focus on identifying triggers that signal changes in a prospect’s likelihood to convert to a customer.
Map your existing base of existing customers to your database of leads. If you’re using a marketing automation platform or working with a demand generation agency, you may use predictive scoring to prioritize known prospects, leads and accounts based on their likelihood to convert into a customer.
For new leads, you can also use scoring to identify and acquire prospects with attributes similar to existing customers.
Evaluate Your Customer Data Against Marketing Channels
There’s no sense in spending time or money on marketing tactics that simply don’t work, so evaluate which marketing channels bring you more bang for your buck. If you’re using a tool like Google Analytics to track your website traffic, you’ll first want to rank which channels are leading to the most conversions. After that, you’ll want to dive deeper into:
- Paid Search: Figure out which ads bring in the most (and least) traffic to determine where to spend PPC dollars and identify which keywords customers search for.
- Organic Search: Inform your future SEO strategy by learning what customers have searched for in the past.
- Social Channels: Evaluate whether certain social channels perform better with different demographics. There may be opportunities to optimize messaging to target a different demographic on each platform.
- Referral Sources: Do your homework to figure out where your referrals are coming from, and if there are any partnership opportunities that would increase the referral rate.
Test Your Campaign
A small difference can make a huge impact. Plan to perform A/B testing with landing pages, content, color schemes, CTA buttons, images and other elements.
Your choice of what to test ultimately depends on your specific goals. For example, if your goal is to increase the number of people that that fill out a whitepaper download form, try testing the length of the form, type of fields in the form, or experiment with progressive profiling for known visitors.
A few things to remember: First, only test one element at a time so it’s clear which marketing strategy is yielding the greatest results. Also, be sure to test the original and the updated versions simultaneously so they are subject to the same market conditions, holiday traffic, and other variables. And, finally, run the campaign long enough to make sure you get results that are statistically significant.
Measure, Optimize and Repeat
Once you’ve discovered what works for your organization, rock it! Increase your effort and marketing spend in the areas that yield the best results. Enjoy the fruits of your labor as you begin to see increased conversions and higher marketing ROI.
Now for a word of caution: Don’t get too comfortable. Optimization is an iterative process that is necessary for a successful digital marketing campaign. Plan to update the your audience analysis quarterly if you have a high volume of visitors to your website, and bi-annually if your audience tends to be smaller. This will help you determine areas where your marketing strategy may need to be adjusted to adapt to the market, new products, emerging social platforms or even new purchasing methods.
Follow that Hunch, Detective Data
In order to stretch marketing dollars, it’s important for marketers to estimate how their target audience will respond to a certain campaign or medium based on previous data related to campaign interactions or audiences with shared characteristics.
Remember, Rome wasn’t built in a day and your predictive marketing strategy likely won’t be, either. Depending on how your organization houses customer data, it will take some time to build a report and manipulate the data to analyze trends—and this has to take place before you start testing new concepts.
Need a shortcut? Check out our guide, 5 Steps to Marketing ROI Stardom to learn how to set the foundation for marketing ROI success with better campaign management.