3silverpopNumbers, data, formulas and algorithms are playing a bigger role than ever in modern society. The business world is no exception, with marketers scrambling to get their heads around customer and prospect data.

For many, a key component of understanding this data involves implementing scoring models to help identify, evaluate and communicate with contacts.

Silverpop’s detailed guide explains scoring, and elaborates on building a good scoring model, adding automation, the future and more, so, I hope you enjoy the summary.

Scoring is a method of assigning points to different criteria for each person in your database. In general, the more pieces of data you build into your scoring system, the more accurate it will be. You can then use these scores to rank contacts according to whatever parameters are important to your business. You can play with the values to determine what scores make the most sense. And you can (and should) edit them at a later time based on your learnings.

When developing your system, it’s important to get input from multiple departments. Keep the group small to remain focused and efficient — three to five people is often ideal.

Building a good scoring model

Both demographic and behavioral data are necessary to ensure scoring accuracy.

Here’s a look at some of the attributes and limitations of each.

Demographic / Firmagraphic / Psychographic Information

A customer or prospect’s demographic data refers to characteristics of both the individual and, in some cases, his or her company - often called “firmagraphic” information. Examples include: name, birthday, gender, occupation, email address, company name, company size, or location. Demographic and firmagraphic information can be vital to an overall score, particularly as you’re able to identify key attributes of prospects that generate sales.

Key psychographic information that’s necessary to accurately score a contact encompasses the classic BANT model: Does the prospect have the budget to purchase your product? Does the person have the authority to make a buying decision? Does the contact need the product? What is the person’s timeline for purchase?

Be careful, self-reported information can be wrong and building trust is important, so here are seven tips for collecting demographic data:

1. Ask for minimal information at the start;

2. Offer good value in exchange for information;

3. Use progressive profiling;

4. Gather data through social sign-in;

5. Implement a preference center;

6. Append third-party data;

7. Make it easy for customers and prospects to provide data across platforms.

Behavioral Information

In the behavioral part of your scoring model, you’ll assign points based on the specific actions a person undertakes. Of course, for your scoring model you’ll want to choose only those activities that you believe indicate that a person is highly engaged with you, e.g.: email opens and clicks,  web page visits (including those people are taken to after completing a form), social sharing of emails or website content, visits to blogs and communities, comments on blog posts, “Likes,” “Follows,” or “Adds”, participation in Facebook polls, video viewings, downloads (Webinars, white papers, how-to guides, etc.), custom behaviors (free trial editions, financial calculators, online tools, etc.), newsletter sign-up, shopping cart abandonment, calls to sales or support, demo request and so on.

It is important however to recognize that scoring on behaviors alone isn’t sufficient to yield a truly accurate customer or prospect score.

Recency & Frequency

A recency element enables you to distinguish between behaviors that happened during a contact’s entire history and those that happened a short time ago, and a frequency element — how often a person took an action- should be part of the score.

Properly weighting your model

How you weight your model will depend on your unique business situation.

Here are seven steps to help:

1. Discuss the common signs for conversion and the degree of impact.

2. List the attributes in general order of importance.

3. Select a score range: Determine what your minimum and maximum scores will be.

4. Define possible attribute categories: List each value you’ll use for scoring each attribute.

5. Assign scores to each attribute/value pair: Referencing the priority list you created, assign scores for each possible value of each attribute.

6. Develop a category system: Start thinking about what buckets you might place contacts in and what score levels might distinguish these buckets.

7. Test your model: Once you’ve got the “first draft” of your model in place, try running some real-life examples of prospects or customers through the model you’ve created to see how they score. You can do this manually, using a spreadsheet to model your score. After you run a few test cases through the model, you’ll likely find there are some adjustments you’ll need to make.

The Silverpop guide provides an interactive Scoring Worksheet, Data Planner and Test Planner you can use for your own scoring purposes.

Adding automation to the coring mix

Now, when a scoring threshold is crossed, automation can kick in immediately to act on the new level of interest or engagement, enabling you to deliver incredibly relevant, timely content. Establish the rule sets and marketing automation does the rest. As a result, efficiency and engagement soar.

Besides email, consider using programs to route
hot leads to your call center or sales team and deliver cold leads to your direct mail print house. Or use your scoring model to help customize web content, targeted landing pages or unique pages within Facebook.

Beyond ranking leads - 5 ways to use scoring

1. Engagement score;

2. Inactivity score - set up an “early identification”
scoring model to identify customers whose engagement level has fallen or
is dropping;

3.Loyalty score - a good loyalty program can make your best customers even happier;

4.Profile completeness score;

5. Gaming score.

Revisit your scoring model periodically – perhaps once a quarter – and see if you notice any outliers. Being able to fine-tune your scoring model improves accuracy, ultimately resulting in increased revenue.

Expect the future to bring even more innovative scoring capabilities, as technology evolves to enable the capturing of additional behaviors across mobile, social and even offline channels. That’s exciting news for marketers who understand that engaging today’s sophisticated buyers requires generating more automated “campaigns of one” in which you speak to each individual on his or her terms.

The guide offers a lot more valuable information, so feel free to download the guide.

By Anjum Siddiqi