Three Ways to Leverage Deep Data in 2017
in Digital Marketing Strategy & Trends | by Emily Joseph
In 2017, there’s no excuse for generalizing your audience. Today, advanced technologies allow marketers to gather intimate customer details, making broad demographics a thing of the past. Surface-level personas? That’s so 2016.
As noted in Sparxoo’s 2017 Digital Trends Report, deep data is a hot industry trend for good reason. Unlike big data, going ‘deep’ refers to collecting high-quality information—not necessarily collecting a large quantity of it. By focusing on the smaller information-dense data streams that are applicable to your campaign, you can save time and money while gaining valuable insight into your audience. Win, win and win.
Ready to join the modern-day marketing world? Here are three ways to leverage deep data this year.
Cozy Up with Attribution Modeling Methods
You may know what the attribution model does, but do you know how it works? Didn’t think so. To recap, this model identifies which touch points lead a customer to purchase. From there, it assigns value to each channel and calculates your ROI. Sounds easy — but unfortunately no particular model reigns supreme. So what’s a busy marketer to do with all these options? If working with a marketing automation agency isn’t in the cards for you, start by learning the different models:
Last-Click or Last Interaction Model – gives 100% credit to the last source before conversion.
First-Click Model – gives 100% credit to the first channel within a search session.
Linear Model – credits each channel within the conversion path equally.
Time Decay Model – gives more credit to the channels closest in time to the conversion, but assigns some credit to every stage.
Position Based or Custom Model – allows marketers to split the credit between the Last- and First-Click Models.
There’s a lot of data out there, but not all of it leads to the big bucks. This year, focus on the models that have proven to be successful and reinvest accordingly. Just think of all the time you’ll save analyzing irrelevant (big) data. *Cue the celebration.*
Make Lead Scoring & Grading Work for YOU
In a dream world, everyone would be a target customer. In reality, it’s a waste of time trying to appeal to the masses. Remember: more isn’t always merrier. Your ideal customer is specific. For example, your ideal customer may be a female who’s 25-40 years old, has a college degree, and is in a human resources position at a mid-market company. The next step is to set and assign point values that classify her (and those similar) as sales team-worthy. Let’s dive into lead scoring and grading:
Lead scoring is the process of assigning value to a lead based on their activity as it relates to your brand. In three words: engagement, engagement, engagement. Every move a prospect makes carries a numerical value. The larger the interaction, the higher the number. For example, opening an email about your product might be five points while starting a free trial might be 50 points. To determine what actions are worth, analyze how your previous leads converted to customers. A proper lead scoring model can accurately separate the prospects that are just visiting or researching from the ones that actually want to do business with you.
Lead grading is the process of assigning value to a lead based on demographics, firmographics and psychographics. It’s one thing to be interested in a brand, but it’s another to be qualified to purchase. Lead grading is your chance to rate a prospect. After all, this is a two-way relationship. All leads begin with a low letter grade, such as a D, D-letter grade but can work up to an A+ based on individual markers such as position, authority and location. Those markers indicate a perfect fit for your product or service; the closer the fit, the higher the grade.
A successful marketing automation strategy will combine both metrics and determine the ideal time and place for the sales team to take over — but only if you’ve mastered the art of a customer profile. Have a chat with your sales team and hash out the details. That way, you can combine past results with current trends to create your perfect prospect persona.
Embrace the Predictive Modeling Method
Much to your disappointment, the predictive modeling method does not predict the future. We can read your horoscope, though, if that helps. However, when used correctly, it can predict customer buying behavior. By identifying existing patterns in current data, the predictive model provides insight into purchasing patterns. This enables marketers to develop an effective strategy for upsell, cross-sell and next-sell options.
The predictive model is your secret weapon for growing existing customer relationships and bringing home the bacon. The statistical process gathers previous data and forecasts customer trends so you can predict how much they’ll spend and when they’ll spend it. From there, let your strategy gurus go to work. The results from predictive models can be used to create a customer-centric marketing plan that includes recommending additional (or superior) products at checkout or after purchase — all based on the individual’s behaviors and interests.
The data you want and need is much closer than you think — you just have to swim over to the deep end of the (marketing) pool. The treasure chest is yours for the taking.