Modifying Engagement Scoring Models (START HERE)
How Do CaliberMind's Engagement Scoring Models Work?
Out of the box, most engagement scoring models are a combination of three factors:
- An inbound activity's raw score
- A "fit" multiplier
- Time decay
The simplest aspect of engagement scoring is the Touch Score. These are raw points allocated to a particular activity type. For example, meetings may be a better selling signal than a content syndication read, so you may give meetings 5 points and content syndication .5 points.
Unlike traditional marketing automation platforms, we factor "fit" into the score itself instead of breaking it out into a grid. You may use demographic or firmographic data when designing a multiplier, and this Multiplier does exactly what it sounds like and multiplies the raw Touch Score points by a chosen Multiplier.
For example, if a meeting is worth 5 points and I want to multiply any interaction with a VP by 3, a meeting with a VP would be worth 15 points.
Finally, our engagement models all include a linear time decay, and the timeframe for that decay is included in the model's name. For example, Inbound30 gives an event's full "points" on day one, half the points on day 15, and zero points on day 30.
With all components included, our engagement score for a single touch looks something like:
Chain-based models are a completely different beast and are not modified using configuration screens. In short, Chain-based engagement looks at activities associated with won opportunities and learns which activities in what sequence are most likely to lead to a sale. We figure out your fit and touch score for you, then events, people, and accounts are scored according to how closely they fit the gold standard with a 365 decay.
What Can I Control?
Similar to marketing automation platform scoring models, you can determine how many points you'd like to give to each event type. Unlike traditional models, we allow you to incorporate any event, including inbound events, tasks, opportunities, and campaign members from your CRM and data from your advertising platforms, and even intent data!
We also allow you to specify multipliers for demographic or firmographic matches to your ideal customer profiles.
Keep in mind our best practice guidelines:
- Test first. Before you assume an event type has a higher conversion rate or certain titles are best, check your database to verify. Don't let an executive or salesperson dictate something your database may not be able to support.
- Simple is best. Low point scores for event types and a small number of multipliers are less likely to dilute your score and obscure important, high-value touches. For example, a VP from the perfect company downloading a piece of contact may not lead to a sale as quickly as a demo request from a contact that hasn't been enriched yet. Make sure you don't lose the high-converting money tree in a forest of shiny objects (that metaphor really didn't work out as we pictured).
- Stay focused. Engagement with your brand is ultimately what we're trying to measure here. Outbound activities really don't have a place in engagement scoring because it's an indicator of how engaged your company is with the prospect--not the other way around.
- When in doubt, ask for help. It's easy to over-engineer anything customizable. If you're not sure about what you've put in place, we're happy to help.
- Verify & Celebrate. Set it and forget it is terrible advice when it comes to engagement scoring. Particularly when you first role out your model, run the numbers to verify that your scores are lining up with your sales team's successes. Then tell everyone about those successes. Trust us - it's the quickest way to sales buy-in.