You'd like to know what you are getting into to build a Likelihood to Buy model with MadKudu?
Well, you're on the right page.
First, you may want to look at these articles to understand
To implement this model and start pushing scores in your CRM to help you surface your most engaged contacts and accounts, we'll need to go through the following steps which include
- Getting your behavioral data from your system
- Mapping the data into MadKudu standard format
- Building the model
- Pushing the model to your CRM.
Step 1: [Pull] connect your system to MadKudu
Getting Behavioral data from your systems.
We have native integrations with
- Segment
- Amplitude
- Mixpanel
- Kissmetrics
- Marketo
- Eloqua
- Salesforce Campaigns/Tasks
- HubSpot forms
- Amazon S3
If you are storing your behavioral data in a data warehouse like Snowflake, Redshift, BiqQuery, or tracking your product usage in another tool like Heap, you can dump this data into an Amazon S3 bucket and share this bucket with us. Learn more about the Amazon S3 integration.
Step 2: [Event Mapping] map all your behavioral data points
The behavioral data from all your different systems is formatted into a unique and standard way, with user-friendly event names like "Signed Up", "Requested a demo".
Our team will work with you to understand how the data tracked in your system corresponds to relevant actions performed by users valuable to include in the model.
Step 3: [Model building] Build the Likelihood to buy model
We'll build for you the model in our Data Science Studio which you'll be able to edit yourself.
Building a model includes the following steps
- Create training and validation dataset
- Analyze which are the most important events
- Adjust the weight and decay of each event to improve the performance
- Review
- Deploy
Step 4: [Push] Sync to your CRM
- Create fields in your CRM to receive the Likelihood to Buy scores following these instructions
- Configure the sync to your CRM. MadKudu will set up this sync upon your confirmation.