I write to ask for your ideas . we have implemented a lead scoring system and I am looking for ideas to improve the model we currently have. All ideas are welcome!!!
Current flows we have:
1. Score website page visits (more than one in a day)
2. Score webinar registration and attendance
3. Score hot leads from events
4. Score trials
5. Score social media engagement
6. Score email clicks
7. Score form submission
What worked for you and for your SDR team? Any automatic notification/scoring rule that is beneficial for you?
There are a lot of guides on lead scoring and a lot of opinions as well, that's because good lead scoring is unique to your business. Think about a buyer's journey, what is important and weight it accordingly and relatively to other actions while putting it into perspective of how long the journey would take.
Score between 0 to 100
Double score that a lead accumulates before becoming known
Use score fields to create "counters" to score X pages visited in a session (score people genuinely researching your business)
Subtract or negate score for any actions that isn't normally on a buyer's journey (e.g. visited careers page, logged into customer portal etc)
As Jay says - it's specific to every business use case. However, the way to measure and improve is based how strong Sales-Marketing alignment is persistent. To maintain realistic scores, though, defining a score range, a time range to account for visits is a good idea - few years back we used to track an activity and score it whenever it happened, however it's better to say - X visits in on most valuable pages during a duration and scoring it. Notification of the most interesting moments are always beneficial - provided we identify them correctly, which requires the alignment.
I recommend doing in-depth interviews with your inbound sales team (SDR/BDR/Account Manager) to understand what is meaningful to them.
We have some nuances in our event scoring:
- Anyone that Sales marks as "hot" gets automatically MQL'd. Our event manager has the sales reps who worked the event mark their hot leads. We have a custom boolean field "Tradeshow Hot Lead" and for those leads we change this to "True" and have a smart campaign that MQLs them.
- We give different scoring weights to people who did different activities at a tradeshow. If they just came by the booth to get swag, that's very few points. If they got a demo at a booth, that's higher.
We also have nuances in our trial signups that helps limit the noise
- If someone uses a personal email domain (Gmail, etc), they don't auto-MQL
- Everyone else auto-MQL's for trial signups
We recently redid our scoring and we also now have demographic/firmographic scoring that aligns with our Account Based Marketing approach. People get an Interest Score and Qualification Score, which matrix to create a Lead Rating A-E. Anyone C or above gets MQL'd. There are a couple of behaviors that override that Lead Rating and will auto-MQL, such as Contact Form.
People in our Top Target Account List score higher than non-Top Targets, and people with higher job titles (VP, Director) score higher than practitioners. This requires having an agreed-upon Top Target Account list and also a super complex segmentation of Job Title into Job Level. (You can also do the Job Level as a picklist field on your form, like Marketo does. Although you'll need to figure out how to handle Tradeshow lists and other data that doesn't come through your website forms.)
One thing we do is exclude pages that are not very indicative of buying behavior or are less valuable pages.
We also score more for pages that are more indicative of buying behavior or are more valuable pages. "Valuable" may be different to everyone.
Agree! We only score on a limited number of pages, such as our product tour and pricing pages. We also limit the number of times someone can score on web pages in a period of time (which is a good practice no matter how you're defining your web page scoring).
Does anyone have any suggestions or ideas for demographic scoring? I am currently building out our model and right now what we are tracking is very basic - position, company size (high med low value) revenue (high med low value), source of lead, invalid names, job title (high med low value), number of employees (high med low value), email type and country (high med and low value).
It would be nice to extend beyond just the basics but I am struggling with this one!
Here are some points I suggest for implementing a lead scoring model:
Before starting to develop a model its important that you reflect and plan on a few key areas. Lead scoring is much effective when you consider demographic, firmographic and behavior aspects of the lead that are important as per your business requirement.
Among these categories, you need to plan on defining what are positive factors and respective weights i.e amount of score you would give a lead when he/she visits a page vs fills a form vs attends a webinar. Similarly, you need to define all the negative factors and their respective weights.
Below are a few common contributors:
Can be much more depending on your business case. Then comes the decaying aspect of scores, how would you reduce the scores for a lead if he/she inactive. Inactivity can be over email, web visits, webinars/events or other channels and also inactivity period i.e inactive in last 7days, 30days, 90days, etc. Respectively you need to define the weights that you going to assign for these leads.
Now, you need to look at when would you reset a score for a specific lead. For example, if a lead is converted into an opportunity or sales close the deal you might want to reset the score.
For little more advance model, weights of one factor might vary with geography, product line i.e a lead from US region filling a form vs lead from UK region filling the same form. You can choose to create tokens for defining weights for all these factors.
Some common problems:
Few users create independent scores for behavior, demographic & firmographic and use these scores in combination to achieve the same tasks using a single lead score.
You can do much more to create some advance scoring models, above mentioned should help you build a decent lead scoring model.
Below is a link to Marketo definitive guide that can provide you details on lead scoring implementation in greater details.
I think one of the important factor to consider is negative scoring, if there is no engagement for a period of time then score for the particular lead should reduce.
Another thing to consider is the firmographic score, have you decided who are your target companies, is it based on industry, revenue, service etc,? once you have decided it, then scoring based on these factors must be done for quick movement of the leads.