We know–models can be intimidating. But as the need to add analytics and science to our work continues to increase, models have become one of the primary vehicles every marketer needs to know how to develop and leverage. If you’ve already dived into the deep end on models, congratulations. On the other hand, if you’re just dipping your toe into the water, have no fear, because while there may be a bit of a current, it is time to venture forth.
Mathematical models help us describe and explain a “system,” such as a market segment or ecosystem. These models enable us to study the effects of different actions, so we can begin to make predictions about behavior, such as purchasing behavior. There are all kinds of mathematical models-statistical models, differential equations, and game theory.
Regardless of the type, all use data to transform an abstract structure into something we can more concretely manage, test, and manipulate. As the mounds of data pile up, it’s easy to lose sight of data application. Because data has become so prolific, you must first be clear about the scope of the model and the associated data sources before constructing any model.
So you’re ready to take the plunge–good for you! So, what models should be part of every marketer’s plan? Whether a novice or a master, we believe that every marketer must be able to build and employ at least four models:
- Customer Buying Model: Illustrates the purchasing decision journey for various customers (segments or persona based) to support pipeline engineering, content, touch point and channel decisions.
- Market Segmentation or Market Model: Provides the vehicle to evaluate the attractiveness of segments, market, or targets. More about this in today’s KeyPoint MPM section.
- Opportunity Scoring Model: Enables marketing and sales to agree on when opportunities are sales worthy and sales ready.
- Campaign Lift Model: Estimates the impact of a particular campaign on the buying behavior.
These four models are an excellent starting point for those of you who are just beginning to incorporate models into your marketing initiatives. For those who have already developed models within your marketing organization, we would love to know whether you have conquered these four, or even whether you agree these four should be at the top of the list. As always, we want to know what you think, so comment or tweet us with your response!
Many companies are developing opportunity scoring models which essentially assign a predetermined numerical score to specific behaviors or statuses within a database. The purpose of opportunity scoring is help sales people know which opportunities are sales ready and worthy, and therefore take priority. Often variables such as title, company, and industry, serve as the basis for the scoring model. However, behaviors can be used too, such as the completion of a contact form, visiting a particular page on the website, participating or viewing a demo, etc. Contextual data adds another dimension to the model by weaving in predisposition information that reflects content, timing and frequency-for example what products they currently use, the last time they purchased, their complete buying history, the types of keywords they used in their search, etc.
Keep in mind, timing is everything. To be effective, contextual data must be delivered to the right person, at the right time, within an actionable context. For example, the date of a key customer’s contract renewal is posted in your CRM system all year long, but that doesn’t mean you’ll remember or even see it. Think how much more useful that data becomes when your system automatically alerts you to the fact that it’s the customer’s renewal date. Sending email messages about renewals too early just creates noise at best and at worst suggests you don’t know their renewal date. Customers are more likely to respond to call to action when it is in context of their workflow. Communication that is contextual is more personal and as a result feels more authentic, shows value, and leads customers want to act. As a result, you can reduce the cost of customer acquisition and the cost of sales.
The end goal of contextual data is to connect with the buyer when they are most predisposed to buy. As a result, you can use contextual data to help build propensity to purchase models, for prioritizing opportunities to support opportunity scoring, to develop more personalized messages, and select the best mix of channels.
This same concept of contextual data can be used to build propensity to purchase models. By identifying the winning experiences associated with a particular segment, you can use this information to craft more relevant messages to similar targets to increase uptake.
Personalization is a compelling and challenging proposition. It’s a moving target and therefore requires a test and learn approach. By adding contextual data into the process you can make your personalization efforts more effective and more relevant.