Grading and Scoring in Account Engagement
As a marketing operations professional one of the questions I often face is how I can make a process more efficient and effective. This often revolves around the lead handoff process between sales and marketing, a topic that will be debated until the end of time. Regardless of what that process may look like at your company, our job is to ensure quality leads are delivered to sales reps in a timely manner. But how do you define quality?
What are grading & scoring in Account Engagement?
Grading and scoring are two out-of-box capabilities within Marketing Cloud Account Engagement1 that help define the quality of a prospect. While they sound similar they actually provide two different evaluations of your prospects.
Simply put, prospect grading is a way for your organization to understand which prospects most align to your ICP (ideal customer profile). The higher the grade, the more the prospect matches your ICP, and vice versa for lower grades. And when a prospect matches the ICP your company has established, the more likely they are to convert to fruitful opportunities. As a business you should want to prioritize follow up for these prospects. Not only is there potential for more opportunities to be created, but this helps maximize the time your sales teams spend following up with prospects.
As a reminder, a prospect’s grade increases or decreases via automation rules or manual updates when you indicate the prospect matches a criteria rule established in a Profile. Additionally, each prospect starts with a D letter grade upon creation into the instance. Each criteria rule can raise or drop the prospect’s grade in increments of ⅓. Think of each ⅓ being a different letter grade you would receive in school. So if you started at a “D” and moved up ⅓, that would bring you to a D+. ⅔ would bring you up to a C- and 1 would bring you up to a C (a whole letter grade). On the contrary, when you indicate a prospect does not match a criteria rule, their letter grade would drop by the amount you specify. In our example above, if a prospect with a D letter grade does not match a criteria rule with a ⅓ value, the prospect would move down to a D-.
Prospect Scoring, on the other hand, demonstrates which prospects are most interested in your company based upon their behaviors tracked by Account Engagement. The higher the score the more involved a prospect has been with your marketing content. The lower the score, the less they have interacted with your brand via tracked Account Engagement activities. While grading will help classify who you as a company want to talk to first, scoring will demonstrate which prospects are most ready to talk to you. Two different data points but one common theme: measuring the quality of your prospects to maximize conversion.
Scoring is also different in the sense of how it’s measured, which is a numerical value versus a letter grade. Scores can increase or decrease via activities tracked in Account Engagement, as well as through Completion Actions and Automation Rules. You can either choose to include Scoring Categories in your scoring model (tracks interest to specific products) or just keep it vanilla with a general score. Regardless of which approach you take, it’s essential to think about how indicative an activity is of a prospect’s desire to do business with your company.
Why should I implement grading & scoring in Account Engagement?
Grading and scoring often go hand-in-hand, and we believe it’s best to implement the two, if you haven’t already, at the same time. But why implement in the first place? Well, per my last email (don’t you just hate it when people say that), these two capabilities help you and your sales team determine the priority of follow-up. Prospects with a high grade and high score (i.e. a Vice President of Information Technology that has been consistently opening emails, attending events, downloading content, etc.) will 99 times out of 100 be more likely to convert to an opportunity than an intern researching your careers page.
From more than just a sales perspective, though, you can vastly improve the personalization in your marketing when leveraging grades and scores in Account Engagement. For example, prospects with a higher score have likely been engaging with your marketing content for a long time. They know your product, they know what pain point you solve in their day-to-day, and possibly even why your product/solution is better than the company down the road. That’s amazing! Kudos to you for producing quality content. Now get them to take that next step by serving up some bottom-of-funnel content, like supplying them with a buyer’s guide or maybe even encouraging a meeting with your sales team. On the other hand, a prospect with a lower score might have just heard of your product/service from a co-worker and landed on your website. Top-of-funnel content would be best for this prospect, so possibly sending them weekly emails about your product/service to educate them is the next best step.
How do I implement grading and scoring for the first time?
Now that you understand the benefits of leveraging grading and scoring let’s dive into how you can go about implementing both of them.
The first step in implementing a grading model is actually identifying all of the active Profiles within your instance. Profiles are a way of organizing your prospects into different groups in the event they should be graded with different criteria. An example would be if you have two completely different types of products you offer and sell to two different types of audiences. Because the ICP of those two populations can vastly differ it may make sense to have multiple profiles. Generally speaking, though, it is best practice to have a single profile within your Account Engagement instance to follow a single grading model. You can still create dynamic lists to separate out your audiences however you need to.
Once you have nailed down how you plan to use your Profiles, you’ll then need to determine which data points you want to evaluate in your grading model. I would encourage you to use 4-5 different data points in your grading model to ensure you see variety in your grade outputs but nothing too bloated. This will also ensure you’re not using too many of the limited amount of automation rules you have available in your instance. And then in terms of deciding which data points to leverage, solicit input from other teams (i.e. sales) on what makes a prospect higher quality from their perspective. Industry, location, and seniority level are a few examples I often see leveraged.
Once you’ve established which data points you want to leverage and ensured they are accurately and consistently populated, it’s time to think about the varying degrees an individual may match a certain criteria rule. Some fields may be a simple true/false boolean field, in which case you would likely only have a single rule to increase the score if the field is a certain value. But there may be some fields with picklist values that aren’t a simple yes/no to your ICP alignment. An industry data point is a good example of this. While companies in the software industry are your main ICP, your company might also have success selling to the hospitality industry. This is where you can implement a high, medium, low matrix into your grading model. In the example above, “software” might be a high match and increases the score by a full letter grade, while “hospitality” is a medium match and would only increase the score by ⅔ of a letter grade.
The next step is to then implement the automation rules that will match records to the criteria you’ve established and increase or decrease their grade values accordingly. Remember what I said earlier about only including 4-5 data points in your grading model due to the limited amount of automation rules in your instance? This is where you’ll see why. Each weight match (high, medium, and low) will require a separate automation rule. So if you have three weights on all five data points you are including in your model, that’s fifteen automation rules immediately. Plus, you’ll need an automation rule for each evaluated field in your model to reset the grade when data values change in the fields you are evaluating. They add up quickly, and if you’re not careful you might not be able to implement the grading model to the extent you desire.
Once you have the automation rules built out and tested, one thing I like to do is run an analysis projection to get an overview of what the grade distribution will look like across all of my prospects when the model is activated. This will help you get an idea of what to expect before fully activating (reach out to us for a template we like to use). Assuming that looks good, now it’s time for the fun part: the backfill!
In order to backfill the grade values you’ll simply resume the automation rules. It’s best practice to resume them during non-business hours. Furthermore, depending on the size of your database it may make sense to separate your prospects into groups and run multiple batches. This is to ensure the other mission-critical automation rules are able to run without issue. Additionally, if you plan to sync the grades into Salesforce, it’s best practice to inform your admin team of the upcoming backfill. Remember to be a good business partner in your projects; your stakeholders will thank you.
One final note: it may take some time for the grade values to appear in your instance. That’s totally normal (I freaked out when I first completed a project like this). Check in regularly on the automation rules and make sure prospects are matching in the manner you expected.
The first step in building out your scoring model is identifying any automation rules and completion rules that include score changes. While you may not have been officially “using” the scoring model prior to this point, someone at some point at your company might have introduced a rogue scoring update in the system. The goal of this audit is to understand where scores are being updated in the instance and centralize as much as possible to eliminate confusion down the road. It may even make sense to deactivate some of the automation rules and delete the completion actions where unnecessary updates are being made.
The next step is to work with your marketing and sales counterparts to align on the score values for the out-of-box tracked activities and at what point threshold it makes sense to send over prospects to the sales team, otherwise known as the MQL (marketing qualified lead) threshold. It is best practice to work backwards and first determine what that threshold is and then assign point values to determine how a prospect can achieve that score value. This will ultimately help ensure only the most interested prospects are being sent to your sales team.
Additionally, Account Engagement will assign point values to default out-of-box tracked activities, but you and your team should evaluate if these point values are indeed what you want them to be. For instance, a form error initially is assigned a negative point value. In my opinion, though, these kinds of errors should not decrease a prospect’s score value. People make mistakes all the time when filling out forms, so it doesn’t make sense to decrease the score for something that isn’t indicative of a prospect’s interest in your business. One key reminder here, though: changes to point values take effect immediately and run retroactively, so don’t make those changes right away. This step is simply to establish the requirements for those point values.
Beyond just the default out-of-box tracked activities, though, there may be other activities you want to score because they indicate a prospect is interested in your business. Page actions are a great example of this. If a known prospect visits your pricing page or some other page with high-intent content you may want to increase their score by a higher amount than other pages on your website. Ideate as a team to determine what other components should be added to your scoring model.
The next step is making sure your scoring model actually means something and can be used by your sales and marketing teams. This is where you will build out the automation rules to assign prospects, either to individuals or a queue in Salesforce. Work with your admin team and sales operations counterparts to determine the names of those queues of what triggers routing in your Salesforce instance. Ideally, routing should take place within Salesforce and not Account Engagement.
The final step is to then actually update the point values. I would recommend doing this during non-business hours, as your Account Engagement Instance will run a system-wide update to retroactively process the point values on all individuals who have taken that action. Once that update is done, your scoring model is live and ready to go! Activate those MQL assignment automation rules and watch the prospects flow in and over to your sales team.
How do I update my existing Prospect Grading and Scoring models?
A lot of the same steps apply when making updates to existing grading and scoring models, so here are some general guidelines and thoughts to consider when making enhancements:
- Evaluate your use of profiles. As I mentioned previously, it may make sense to condense down to a single profile if you currently use multiple.
- If you are going to condense down to a single profile, run an automation rule (segmentation rule won’t work in this situation) to assign everybody to that profile. If needed, make sure to deactivate any automation rules assigning prospects to the soon-to-be deleted profile.
- Evaluate your existing profile criteria and determine if any updates need to be made. Maybe you have an outdated field that isn’t populated anymore or maybe a new field needs to be evaluated. Whatever course of action you take, make sure to involve your stakeholders.
- If you are updating existing automation rules, be mindful that the grades will be re-calculated automatically once you resume them. If you are creating brand new automation rules, make sure to deactivate the old ones.
- Create reports to monitor the grades as prospects are updated and created. This is a great way to get a step ahead of everybody else and track the quality of your incoming prospects, which can serve as an important metric moving forward.
- Regularly re-evaluate your grading model. Personally, I would recommend twice a year. This allows for enough data to come through to understand performance without making too rash of a decision. Again, involve your stakeholders here.
- Identify any automation rules you have running that are dependent upon scores and deactivate accordingly. Any adjustment to out-of-box scoring will retroactively update to the applicable prospects. No need for a bunch of false MQLs!
- Once you have mitigated any downstream effects, pull a report of all the scoring updates that are occurring outside of the out-of-box activities, such as in completion actions or automation rules. Determine if any of these are still necessary. Remember, it’s best to keep your scoring updates as centralized as possible.
- Once you have your scoring updates better organized, think through the existing values attached to each of the tracked activities. Are they still reflective of the prospect’s intent? Work with your stakeholders here to make any necessary changes.
- Once you have the updated point values, think back to those automation rules that reference scores (i.e. MQL assignments). Should the updated score values, if any exist, warrant any changes in those automation rules?
- Are your prospect scores really low or really high and you can’t make sense of how to treat these prospects? Considering resetting every prospect’s score to 0 and starting fresh. Definitely consult your stakeholders before doing this.
Every world-class organization will have a clearly defined mechanism to understand which prospects they want to go after and which ones are most ready to talk to sales. Account Engagement offers this capability, through grading and scoring, and can be implemented in a relatively short period of time. With grading and scoring you can be well on your way to being one of those world-class organizations.
1In April 2022, Salesforce renamed Pardot to Marketing Cloud Account Engagement.