How to build your first Score?
Last updated
Last updated
Getting started with scoring is hard! There's two key questions you need to answer:
what are the relevant properties?
what weight should be assigned to each of those properties?
Ideally, you'd run a correlation analysis to answer those questions but we all know that the data science resources are scarce & expensive. Don't worry though, this guide is all about how to create good scoring even if you don't have data science resources available to you.
The starting point is to identify a set of both good and bad companies. These can be your best clients or ideal prospects (as examples of good) and companies that you know you can't sell to or you've lost in the past (examples of bad).
Our goal is to find a set of properties & weights that will allow us to clearly differentiate between good & bad companies (good companies scoring high and bad companies scoring low).
We'd recommend to start with at least 5 examples of each. With more examples of each though you'll be able to build more robust scoring.
Using Preview's "Search for specific companies by domain's" search you can add all those companies to the preview screen.
Once you have companies added to the preview now it's time to start finding relevant properties. You will typically know them (from your ICP definitions or experience).
Most importantly you can preview what values your example companies have by adding them to preview screen (you just need to add an empty rule with the name of the field for it to show up in the preview screen).
If you're out of ideas, you can also check the Data Directory and look up what values (& coverage) each property has.
You should iterate this process of identifying a field name, value & the weight until your list is clearly broken down into high scoring (good) and low scoring (bad) companies.
Once you're happy how the good & bad companies are ordered you can remove the domain filters & see whether the highest scoring companies are the ones you like (doing same exercise for Worst fit companies will ensure that you're not deprioritising any good looking companies).
If you see any company in the overall list that shouldn't be high or low scoring (Best fit / Worst fit) add them back to the list in company search & iterate further!
The perfect scoring is unfortunately a holy grail - it will never be perfect. But that doesn't mean it cannot get better over time.
Once you've rolled out scoring you'll start getting feedback - whether that's sales team pointing out that there's some high scoring companies that are clearly a bit fit or (more positively) you'll start creating & closing opportunities with the scored companies.
These can be used to further improve the scoring in the same way we've built it for the first time. We just add the examples of good & bad to our preview and keep iterating until there's clear distinction in the scores.
We typically see that done on some cadence whether that's once a month or quarter.
Building your score for the first time can be confusing & daunting exercise. We're more than happy to help you whether that's brainstorming relevant properties or getting into the delicate art of scoring weights. Give us a shout & we'll schedule a call to go through it together!