Conversion Mapping

Are you more of a visual learner? Head this way for our video!


Video tutorials are only available to users logged in to the MadKudu help center. Click here to learn how to create an account in the MadKudu help center. 

Supported integrations 

  • Salesforce 

  • HubSpot

  • Segment

  • Mixpanel

  • Marketo

  • Kissmetrics

  • Amplitude

What is the Conversion mapping about? 

A conversion is the success criteria your models will be trained to predict.

For example: you want your models to predict when a lead will be linked to an Opportunity that reaches the stage Closed Won.

The Conversion mapping allows you to define what is a Conversion to train a model against in the Data Studio or to look at the performance of a model/segmentation against in the Predictions page in the App. 

MadKudu has standardized 2 definitions: Closed Won and Open Opp. These 2 definitions are used across the platform in various reports. Whatever conversion definition you want to train your model against, it will always be important to check if your model can also predict when an opportunity reaches the Closed Won stage.

The 'SQO' definition is here to contain any custom conversion definition you would like to use. You can add as many custom conversion definitions as you need.

A frequent custom conversion definition built is : Closed Won above > $ amount.

We recommend that you train your model on the most exclusive conversion definition you can, i.e. the furthest down the funnel your volume of data allows you to reach.

You would start with a conversion defined as Closed Won above > $ amount, and if that is not enough data, revert to a stage mid-funnel (probability = 50%, 20%...) > $ amount. And if that is also not enough data, decrease the amount.

Note : We consider having "enough data" if you have at least 30 conversions per month in the audience considered.

Important conditions to create a conversion definition are:

  • Opportunity Type: what type of opportunity do you want your model to predict? New Business? Upsells? Renewals?

  • Opportunity Amount: MadKudu tracks the amount of Closed Won opportunities in various reports so you can make sure your model predicts your revenue correctly.

  • Opportunity Probability: Your model will predict when an opportunity reaches a certain stage, and beyond. We recommend you to create conditions based on the probability field, so you can easily catch all opportunities that have gone through this stage at some point in time. You can find the matching between your StageName and Probability under the Data tab in the App.  

      

What is the Conversion mapping used for?

  • As the success criteria to train and validate a model in the Data Studio. Learn more

  • As a parameter for the Predictions pages reports

  • As a parameter for the Insights reports

Where is this mapping available?

In app.madkudu.com > Mapping > Conversion mapping

How to create a Conversion definition?

Pre-requisite:

  • You are a MadKudu user with Architect permissions

  • MadKudu needs to be pulling the CRM fields you'd like to map first before adding them to the conversion mapping. Typically, the fields where you track the type of the Opportunity and the amount of the Opportunity will be necessary. MadKudu pulls by default the standard Salesforce fields type and amount, but if you use custom fields, you would need to pull them. Please follow the instructions here.

Conversion definition from your CRM

  1. Go to app.madkudu.com > Mapping > Conversion mapping

  2. Click on Edit 
    mceclip0.png

  3. Select the integrations you'd like to map conversions from. Only integrations you've connected and supported would be available.  

  4. Click Save & Continue 

  5. On the next screen, there's no need to change anything, click save & continue again 

  6. At the bottom of the page, click Create New SQO 

  7. Enter a name for this new conversion definition (special characters, numbers, lower and upper case accepted) 

  8. Click on the +  button

  9. In the "type of deal", enter the exact value of Opportunity type field in Salesforce or Deal dealtype in HubSpot you would like to use to filter Opportunities based on the type (new biz, upsell....). You can find this value in your CRM or in app.madkudu.com > Data   


    mceclip5.png

  10. Select the field on the Opportunity object which contains the Opportunity ARR i.e. the amount if the deal (this value will be used to look at the model performance in terms of revenue in the MadKudu Predictions and Insights reports and in the Studio) 

  11. Select the conditions to define the Conversion (usually contains filters on the Opportunity object) based on:

    1. Object (usually Opportunity)

    2. Field (probability, stagename, amount...)

    3. Condition (is more than, is any of, contains...)

    4. Value 

    5. Check 'Case insensitive' (allows to find the value regardless of upper/lower case) 

  12. Add any other necessary condition by clicking on Add a new condition

  13. If you are using multiple conditions, edit the condition logic like accordingly 

    • Example: $1 AND ($2 OR $3)

  14. Click on OK

  15. When you are done, click on Publish at the bottom of the page
     

Well done! What's next?

Closed Won conversion definition from a behavioral data source

  1. Go to app.madkudu.com > Mapping > Conversion mapping

  2. Click on Edit mceclip0.png

  3. To use a behavioral data source you've already connected to MadKudu, select 'Analytics'. 

  4. Click Save & Continue 

  5. On the next screen, there's no need to change anything, click save & continue again 

  6. In the Closed Won conversion definition, click on the +  buttonmceclip2.png

  7. Choose your source system in the list mceclip3.png

  8. Type the name of the event defining a closed-won conversion (a paying customer), exactly as it appears in your source system.mceclip4.png

  9. Click on OK

  10. When you are done, click on Publish at the bottom of the page 

Well done! What's next?

How to edit a Conversion definition?

  1. Go to app.madkudu.com > Mapping > Conversion mapping

  2. Click on Edit

  3. Keep the integration selected

  4. Click Save & Continue 

  5. Change if you need

    1. Opportunity type: this will change the Opp type for Closed Won and Open Opp

      1. Enter one or more values of the standard Salesforce Opportunity field Type 

      2. Leave empty to not filter on any specific opp type. 

    2. Amount field: this will change the amount field where we get the $value of the conversion (at the Opp or Account level), used to compute revenue in the Insights reports and Customer Fit predictions page.   


      mceclip0.png

  6. Click Save & Continue

  7. To edit any definition click on the + button 

  8. Once your changes are done, review 

  9. Click on Publish at the bottom of the page

Well done! 

What's next?

You should receive an email of success or failure within a few minutes.

  • In case of success: Congrats! Your conversions will be available in the App and in the Studio after the next time map has run on the Process page (estimation : in the next 4-8 hours)

  • In case of failure : Contact your onboarding team with a screen capture of the error so they can help you move forward!

F.A.Q

Is there a limit to the number of conversion mappings I can create?

You can create as many conversion mappings as you need. 

Troubleshooting

  • If you get a 500 error, please make sure you are still logged in. Note that by refreshing the page your configuration will be lost.

  • If you get a 404 error, please make sure

    • all input fields have a value (conversion name, logic, value..) 

  • If you get an error that your dataset is empty after reloading the dataset, check that your updated conversion mapping definition actually catches some records in the dates chosen for your dataset parameters. It is recommended that the conversion mapping definition catches at least around 100 conversions. You can check this by clicking on the 'Check Volumes' button in the conversion mapping page in the app. Screen_Shot_2022-10-31_at_6.19.17_PM.png

Still having issues? Please open a ticket here with the configuration you are trying to publish and we'll get to you as soon as we can.