Microsoft Dynamics 365 for Marketing: Automatic Lead Scoring
One of the new capabilities in Dynamics 365 for Marketing includes automatic lead scoring enabling marketing teams to instantly identify which prospects are 'sales ready'.
Using this native functionality, lead scoring models can be quickly configured and deployed with the aim of supporting sales in better focusing efforts on the hottest leads at the right time.
Before looking at some examples, lets start with the fundamentals.
Firstly, this functionality is specific to the Dynamics 365 for Marketing app, and as the name implies, this is for scoring Lead records only.
Also, it's important to note that automatic lead scoring will only work for leads that are associated with a contact. Attempting to score a lead that doesn't have an associated contact will fail (in the current release).
You can have multiple criteria within an individual scoring model and the app supports multiple models. As a result, it is possible that a lead record could have a series of gradings that reflect scores across independent models.
Segments are a key part of Dynamics 365 for Marketing as they define the audience for campaign, but these can only include contacts, not leads. As a result, you can only address marketing emails to contacts.
This isn't a barrier to lead scoring because each model will 'hop' across entities.
In many instances, teams will want to score leads based on email recipient behaviour, e.g. how many times an associated contact has opened or clicked marketing emails. Dynamics 365 for Marketing will query a parent contact record to calculate these values and this can be easily adapted to only score email events that occurred in recent weeks or months.
However, email engagement and web behaviour is just one example of the criteria that can be used.
Data from any field on lead records can be referenced to impact a lead score, in whole or part.
For example, the prospective monetary value of a new lead might be a factor in determine its rating.
In setting up a new Lead Scoring Model a condition tile has been dragged onto the design canvas. This is configured to reference the Lead entity which automatically brings up a list of every searchable lead field within the expression field.
Here, we've pointed to the Estimated Value field and set a minimum value of 10,000.
The above screenshot reflects a single expression but within each scoring condition multiple expressions can be added.
In another scenario, a second expression might reference an 'estimated close by' field so that higher value prospective opportunities that might be winnable soon are prioritised.
Further expressions could reference a specific product or service, or this additional criteria could be referenced in separate conditions within the same scoring model.
Where multiple expressions are used in the same scoring condition it's important to note that these are combined using an AND operator. As a result, these must all be set to true for the condition action to be triggered.
For this example, we'll keep it simple and stick with a single expression so the next step will be to attach a score when this Lead Value condition is met.
In this instance we'll award a +10 value for any lead with an estimated value of 10,000 or higher.
More conditions can be added to develop a scoring model that reflects various attributes which contribute to an overall score.
These can be child conditions that belong to the same parent condition. Again, these are combined using an AND operator so all parent and child conditions would need to be met for a score to be triggered.
For the example below, we've added a 'Confirmed Interest' child condition. Any leads in our database with a value of 10,000 or more, AND where interest has been confirmed will be awarded 10 points.
Aside from child conditions, multiple parent conditions can also be included on each lead scoring model and these will be triggered using an OR operator.
In a different scenario, we are running an event in Glasgow for senior marketing professionals. To help target invitations, a lead scoring model is configured to highlight the top active leads that meet our criteria.
For the example below, we are using 6 independent conditions and if any of these are triggered an associated score will be allocated.
Earlier, we showed how condition expressions reference the Lead entity record but as mentioned these rules can traverse / hop across entities (up to as many as 5 hops are supported).
For our localised event we are applying a higher weighted score to contacts who are based in Glasgow. Within the properties for this condition we'll start this journey with the Lead entity and use dot notation which then allows the user to select the Parent Contact.
Once this is set, the fields shown in the first expression, and any additional expressions, on this condition will reference the Contact entity.
For this example, we are referencing the Address 1: City field for contacts so the condition will be triggered where this value is set as Glasgow.
Further conditions in this scoring model inlcude different weighted values for contacts based in Edinburgh and a series of varying scores across job titles. These also reference parent contacts with scores awarded to reflect the target delegate profile for this event, e.g. a higher score will be applied if an associated contact is a marketing director compared to a marketing manager reflecting their likely purchasing power.
Finally, a behaviour condition has also been added to award an additional lead score where recent email clicks have been tracked from an associated contact.
This condition starts off with the Email Clicked entity. From this we've hopped out to the parent contact and in this rule we'll award a score where at least 1 click has been tracked during the last 2 months. This could be set as a lifetime search but by narrowing the behaviour to recent months will help generate a representative score that will help understand which leads are sales ready.
To complete this set-up and help users interpret lead scores, a sales ready score can now be set with a schedule of supporting grades.
Here we've defined our target 'sales ready' score as 30 with any leads scoring more than 45 being graded as 'scorching', below this tier, leads scoring between 30 - 44 will simply be scored as 'warm'.
Once the scoring model goes live Dynamics 365 will apply these rules to score individual leads. Please note, this isn't a real-time process.
On a lead, we can see this record exceeds the defined 'sales ready' score and it has duly been graded as warm so this looks to be a good prospect to prioritize for an invitation to the event.
Unfortunately, Dynamics 365 for Marketing doesn't seem to include a breakdown to show how an individual score was calculated (maybe an early new feature request!) but in referring back to our scoring model and the parent contact we can see that...
The parent contact is a marketing manager working in Glasgow and these independent conditions are award 10 and 25 points respectively in line with this scoring model.
Finally, the behaviour condition in this model has also been triggered.
On the contact record we can see that email clicks have been tracked on this parent contact within the last couple of months (at the time of writing) which contributes a further 5 points to the total lead score of 40 in line with our defined conditions.
In these examples, our conditions have been positive but as part of a lead scoring model selective conditions can be configured to award negative scores. This will be another tactic to filter out some leads to ensure a total score is truly representative of its value within the context of any model.
Once lead score values are applied, this data can then be made actionable through list views and automated workflows.