This document describes how to set up scoring and enrichment of bulks of records with Predictive Insights on periodic basis. Record information stored in Anaplan, (e.g., Account Name, Account Website URL, Person Contact Email, etc.) is passed to the Predictive Insights platform. This integration returns any of the following information in Anaplan:
Company information (e.g., Company Name, HQ Location, Website, etc.)
Predictive Insights’ Score (0-100) and Ranking (A, B, C, D) as defined on Predictive Insights’ platform
Indicators for enrichment (e.g., Company Technology usage, Firmographics, Hiring information, etc.) in multiple formats (Boolean, Text, Numeric)
The enrichment can be done for any market built within Predictive Insights platform. Per market and upon request, Predictive Insights will provide the list of Input/Request fields needed for creating the Anaplan Export Action and the Output/Response enriched fields that will be sent back via the Anaplan Import Action.
The data flows as shown here:
Confirm that each platform is ready for integration.
Predictive Insights Platform
The applicable Predictive Insights Market is selected.
The Market should include a Code field that uniquely identifies Account or Lead records in Anaplan. Navigate to the selected Market and go to Other ➡️ Market Settings to view the fields included in the Market. For example, if “SFDC Account ID” uniquely identifies Accounts in the Anaplan main platform, “SFDC Account ID” must be included as a field in the Predictive Insights Market.
The relevant fields for enrichment are enabled for Export.
Fields can be enabled for export by toggling ON the slider next to the field under the Enrichment tab for that Market.
Users typically select Predictive Score and Rank, firmographic information, as well as relevant Anaplan Indicators.
For more information about how to select the relevant indicators for export, reference the “Selecting Indicators” Training Guide.
The instance of Predictive Insights is enabled for integration to the Anaplan platform.
If the Anaplan logo is NOT shown on the External Platforms screen, contact Anaplan Customer Care or submit request to Anaplan Support.
The user has permission to set up the integration.
Depending on the existing settings in your Predictive Insights instance, an administrator may need to grant you additional permissions or create a new user.
The applicable Anaplan Model is open.
Create a new Model for the purpose of the integration or use an existing model. Confirm the Model contains the Account or Lead records that will be scored and enriched in PI. Confirm that the same unique Code field used in the PI Market is available as a list in the Anaplan Model. Also confirm that other record fields like Company Name, Company URL, Contact/Lead Email Address that appear in the PI Market are available.
The Anaplan model may be the original source of the data used to build the, in which case you may reference existing Modules to bring data into the dedicated PI Export/Import Modules outlined in this guide.
Note the Anaplan Workspace ID and the Anaplan Model ID, found in the text of the Anaplan model URL. The IDs are also found in the “?” ➡️ About section.
The user is enabled for non-Single-Sign On access through the Anaplan “Front Door”.
Note the username and password for later entry into Predictive Insights platform.
If you do not already have a Front Door log-in, uncheck the box for “SSO user” next to the user email address. Log out and log in through the Anaplan Front Door, then click forgot password and create a password.
Configure Integration in Predictive Insights Platform
First, enable a connection between the Predictive Insights instance and Anaplan.
Then, configure the connection for a specific Predictive Insights Market.
Log in to the Predictive Insights instance containing the desired Predictive Insights model.
Navigate to the Settings tab on the left side of the screen, then select External Platforms.
Toggle ON the slider under the Anaplan logo.
Enter the Anaplan email username and non-SSO password you created earlier. An error message will appear if the credentials are invalid. Then click Check Connection and Save.
Schedule the ongoing integration. Select the time and day of the week to run the integration. This can be changed at any time, with the exception of the Refresh Period which is set to be weekly and is the default for all integrations.
Click Save. Later in the integration process you will return to this screen and click “Run Integration Now” to test the connection.
Navigate to “Markets” tab on the left-hand side of the screen and select the Market that will score and enrich records in Anaplan.
In the upper right-hand corner of the screen select Other Market Integrations and navigate to the Anaplan tab.
Input the Anaplan Workspace ID and Anaplan Model ID noted in Section 1: Pre-Configuration.
Enable the applicable input fields for enrichment by toggling ON the “Anaplan File Field” next to the Field Name.
For most models the minimum required fields are Company Name plus Company Website and/or Lead Email Address. Required fields are marked with an asterisk*. Additional fields such as Lead Name, Lead Job Title, Company Country, and others may also be recommended to include if those fields were included as input data fields in the original PI market.
Select the fields that will be sent from Predictive Insights to Anaplan in the “Exportable MI’s” box.
These are the additional Indicator data fields that will be appended to records in then Anaplan platform. These are the data fields that were selected for enrichment in Section 1: Pre-Configuration.
This box should not be left blank. At least one Indicator must be included.
In the example below, Cloud Technology, Software Suite: Microsoft, and other indicators will be enriched to records in the Anaplan platform. These indicators should be selected based on their relevance to the end user for filtering and segmentation.
Select the fields that will be sent from Anaplan to Predictive Insights in the “Record Columns” box.
Include the minimum required input fields enabled above, in addition to the Predictive Score and Rank.
It is strongly recommended to include a Code field (e.g., Account ID, Contact ID, etc.) as a unique identifier. This is to ensure that data is mapped accurately between the two systems. If such a field is not available, use Lead Email Address, Company URL, or Company Name.
In the example below, Company Name, Company Website, SFDC Account ID are all included to accurately match the source record in the Anaplan Model. Predictive Score and Rank are the results of the Predictive Insights market on the likelihood that a given record will convert.
Click Save. Do not exit the screen.
Prepare to set up the integration in the Anaplan platform by selecting “View Export Module Columns”.
Leave this screen open because you will return to this page to copy input field names and enter Import and Export actions created in the next step.
Configure Integration in Anaplan Platform
Action to Export to Predictive Insights
As noted in the previous step, log in to the Predictive Insights platform and navigate to the applicable model and select Other Market Integrations and in the Anaplan tab click on the ‘View Export Module Columns’.
These fields must exactly match the Line Items in the Module and Export Action that exports data from the Anaplan main platform to Predictive Insights.
Note that the field names appear in two different formats. The Module and Export Action include the “Anaplan Module Field Names” as Line Items.
Make sure that the same Code is available as appears in the Anaplan Model. In the example below SFDC Account ID is the code that uniquely identifies records. Depending on the use case other fields such as Account ID or Contact ID may be used as the Code.
Log in to the Anaplan platform Front Door using the non-SSO user confirmed in Section 1: Pre-Configuration.
Identify the existing list or create a new list of records to be scored and enriched. This list will serve as a dimension in the PI Export Module.
This list should correspond to whichever Code field is used in the PI Market.
If the Predictive Insights Market is Account-Based, this list should contain unique identifiers for each Account record, such as Account ID, Company URL, or Company Name.
If the Predictive Insights Market is Lead-Based, this list will contain unique identifiers for each Lead record, such as Lead/Contact ID or Email Address.
Create a new module named “PI Export Module” or similar.
Select the “Accounts”/ “Leads” list as the rows and “Line Items” as the columns.
Copy and paste the Export Module Fields displayed in the Predictive Insights platform into the Line Items box.
Navigate to the Blueprint view for the PI Enrichment Export module
Navigate to the Blueprint view for the PI Enrichment Export module
Change field settings to match input data from Predictive Insights platform. Verify the data type of Predictive Insights input fields by selecting Other ➡️ Market Settings under the Field Mapping tab.
Most fields in this module are likely to be text. Email and link options are OK.
The ID field that is the code for the Accounts or Leads list can also be inputted as a List type.
Verify the fields are identical between the Anaplan platform and Predictive Insights Platform.
Navigate to the Blueprint view for the PI Enrichment Export module
Placeholder data can be used. Actual data can be imported from at a later time
Select Data ➡️ Export
Set File Type to Comma Separated Values (.csv)
Check the “Save Export Definition” checkbox
In the Export Name field, type: PI_Enrichment_Export
Return to the Predictive Insights platform and enter PI_Enrichment_Export as the Export Action. Click Save.
Action to Import to Anaplan
In the Predictive Insights platform, navigate to the Market Integrations ➡️ Anaplan tab. Click on the ‘Download Field Mapping’ button to download a CSV Blueprint file that includes all the columns that need to exist in the Import module.
This CSV Blueprint will be used to create the Module and related Action that imports PI enrichment.
Every column name in this file must be included as a Line Item. These fields will include the exact the Line Items from the Export Module, plus the PI fields selected for enrichment.
Create a new module named “PI Results Import” or similar.
Select the “Accounts”/ “Leads” list as the rows and “Line Items” as the columns.
Copy and paste the CSV Blueprint file column names into the Line Items box.
Navigate to the Blueprint view for the PI Enrichment Export module
Change field settings to match input data from Predictive Insights platform. Most Indicator fields are Boolean although certain categorical fields such as Annual Revenue bands are text. Predictive Score is a number. Other fields are text.
Verify the name and data type of Predictive Insights input fields by in the CSV Blueprint file.
Open the CSV Blueprint file from the Predictive Insights platform. Save the file under a new name and delete all rows except for the header, leaving only the field names.
Return to the Grid view and select Data ➡️ Import.
Click “Upload New File” and upload the file containing the list of field names
Set Default File to Admins Only and click Next.
Set up the column mapping.
The “Accounts” or “Leads” target column should be the Code field used in previous steps, (e.g., SFDC Account ID, Lead Email Address, etc.).
The “PI Results Import Line Items” should be “Column Headers”.
Verify that the Line Items have mapped correctly. The system automatically maps the fields so long as the file headers match the module Line Items.
Click “Run Import” and verify that the action was saved in Anaplan. Name the action “PI_Enrichment_Import” or similar.
Copy and Paste the action name into the Import Action field in the Predictive Insights platform. Click Save and confirm that Market Integration is toggled ON.
Initiate Integration and View Results
In Predictive Insights ➡️ Settings ➡️ External Platforms ➡️ Anaplan Integration Settings and click Run Integration Now to initialize the integration immediately.
Monitor the process via Model ➡️ History in Anaplan to track the execution of the import action.
Verify that the integration was successful by looking at the Results Import Module.
Multiple Predictive Insight Markets
It may be useful to integrate multiple Predictive Insight Markets to the same Anaplan Model. For example, there may be different models for different customer segments (e.g., Geographic regions, Account Size, Product Type, etc.)
Each new market is integrated via a dedicated Import Action that populates its own module. Follow the steps outlined under Section 3.b and 4.a above to configure the Predictive Insights Market and Anaplan Import Action respectively.
Export Action can be reused between different markets provided that input fields for enrichment are identical, otherwise, Export Actions from Anaplan should be created to reflect differences between the markets.
Some Markets in Predictive Insights may be “Multi-Product” and provide multiple Predictive Rank and Score results in a single scoring and enrichment action.
The Import Module and associated Action must include columns for each products’ Predictive Score and Rank fields
For example, a Multi-Product Market will have an Import Module with multiple fields for Predictive Score and Rank:
Add or Remove Fields after Initial Set Up
An existing integration can be modified to include additional Anaplan Insights after the initial set up.
In the Predictive Insights platform, navigate to the Integrations ➡️ Anaplan tab. Search and select the additional fields in the “Exportable Indicators” box. Click “Save” when finished.
Download the updated blueprint CSV file with ‘Download Field Mapping’ button.
In the Anaplan platform, add the new fields to the PI Import Results module.
Run integration once using ‘Run Integration Now’ option in Predictive Insights ➡️ Settings ➡️ External Platforms ➡️ Anaplan Integration settings.
After integration has finished once, the integration should be able to match the new fields automatically. If this does not happen, you need to modify import action (for example the PI_Enrichment_Import action) and, if needed, to configure mapping between new column in the action and new module field.
Save modified import action.
... View more
Market Analyzer is a functionality within Predictive Insights that enables users to build out specific account segments within their prospect account database , as well as identity greenspace or Net New A ccounts . Within these segments, users can leverage the propensity to buy model rank/score, as well as specify desir e d indicators that must be present for each account within the segment. Once the segment filters are set , users can export the segment and upload to the ir desired platform to action on the list of accounts ( e.g. marketing automation, Salesforce, or other m arketing tool s ) . To get started, select the “Market Analyzer” icon from the blue Anaplan menu on the left side of the platform.
Creating a New Segment
To generate a new account segment, select the building icon , highlighted in red below . For the segment to save properly, you must select at least one filter criteria . For example, select the desire d propensity to buy model. You will be prompted to “Refresh” with a blue button. After clicking the refresh button , the segment will then be saved as “New Account Segment". R ename this segment so it explains the purpose of list , as described below .
Naming a Segment
When opening Market Analyzer, you will see a pre - built “Mintigo Re commended List . ” After selecting the pre - built segment , or a new segment that you have created , rename the segment to describe what the segment includes. Note : after renaming, make sure to click just below the title to ensure the segment name has changed. For instance, in this example, the segment is named “NA ABM Q2 Account List Retail . ”
TIP: Avoid using special characters, dashes or hyphens in your segment names as the system will give an error and not allow you to save the name. Also, if many people will be creating segments, a standardized naming convention and time stamp will better enable you to manage the segments.
Whether a segment is requested for a marketing, sales or general efforts, it ’ s important to select the right criteria with the help of the Market Analyzer filters ; s egments can be broad or very granular to meet business needs. In addition, you can specify whether your segment should include “Known Accounts” which have previously been scored, enriched, or exported from the Predictive Insights platform, and/or “Net New,” which are companies in the Predictive Insights database not previously associated with your own.
In this section, select propensity model you’d like to use. Remember, each model/market is built to solve something different, so be clear on what the segment should be referencing.
In this example, the ABM model was built to predict closed/won accounts globally. In other cases, models are built for propensity to buy or reach different areas of the sales funnels, purchase different products, or even focus on specific geographical regions.
Record Type Selection
In this section, you can build a segment for “Known Accounts" , “Net New Accounts ”, or both.
“Known Accounts” accounts from your database used to build your models and any records that have been scored by these models , via in - platform list - scoring or integrations .
Note:The Predictive Insights platform does not have visibility to all accounts in your CRM.To ensure all of your accounts are accounted for in the Predictive Insights platform, a suppression list must be added. See“Add suppression list” section below.
“Net New Accounts” are accounts from the Predictive Insights database that are not within the known accounts detailed above or in the suppression list at the time it was added .
Note: Net New Accounts, when selected and exported from Market Analyzer, can only be considered net new at the time of first export. From that time on, these accounts are considered “Known”. There is a “Source” column in the export file to clarify which segment the account was first associated with.
Propensity to Buy Rank/Score Selection
In this section, select the desired rank/score threshold for the segment.
When hovering over the “Account Score”section, a pop-up appears to show rank/score thresholds. In this example, selecting an Account Score of 60 will results in a segment including accounts that score 60 and above, encompassing both A and B ranks.
Marketing and Intent Indicator Selection
Identifying what rank/score of accounts is desired for a segment is the first step. In addition, if there are firmographics (number of employees, industries, annual revenue, etc.) or additional fit or intent indicators you wish to apply to this segment, you can select the desired indicators in this section. Note: if you select an indicator for a segment and you want that indicator to be appended to an account when the segment is exported from Market Analyzer, the indicator must also be turned on within the model/market on the Enrichment tab.
There are two sections to search for indicators as a filter. Either or both of these sections can be populated to further define a segment.
1. Top box = AND condition. Add desired indicators to the top box when all must be present for each account within the segment. For instance, you only want accounts that are in the retail industry, in North America, and have a specific number of employees or annual revenue. This box is not limited to firmographics; however, it can include technology fit indicators or intent as well.
2. Bottom box = OR condition. Add desired indicators to the bottom box when one or more must be present within the segment. In this example, we can see that the account must be showing intent for marketing automation tools OR intent for Predictive Insights. The account can have one or both of these indicators present.
Once the indicators are added to this filter, be sure to hit the “Apply” button for the segment to save criteria selected.
If Buying Stages is activated, segments can be refined by selecting the desired Buying Stage flag. To activate this capability, a Buying Stages pixel must be placed on your website.
Segments can be refined by any of the Buying Stages. Not only can this filter be applied to your existing database, it can also identify net new accounts.
See the Buying Stages guide for more detailed instructions.
To ensure Net New Accounts are truly net new to your database, you will need to upload and use a suppression list. The Predictive Insights platform does not have access to all records or accounts in your CRM, marketing automation, or other systems, rather it only has visibility into those records used to build model, those scored against a model, or those previously exported from Demand Center (Market Analyzer).
This filter allows clients to select the most recent list of accounts in their database. This requires a client to upload a suppression list to the Predictive Insights platform (see next section.) Once the list is uploaded, please select the list of the drop-down menu.
Adding Suppression List to Platform
To add a suppression list to the platform, please take the following steps.
Step 1: Prepare the file
The file should include: Account Name, Account Website/Domain
The file must be in CSV format
Step 2: Login into Predictive Insights platform
Step 3: Upload file to platform
After logging in, go to Settings → Suppression
When in the Suppression section, select “Add Suppression File”
Select the file from your computer by hitting “Browse” to upload the file.
Then hit “Next” and map either the website/domain to the Predictive Insight field. If you do not have website/domains in the file, yo u can select email addresses from the file.
The file will upload and you will see the name of the file appear in the list of suppression lists. A confirmation email will also be sent to the user that upload ed the list to the Predictive Insights platform.
Activating Net New Accounts
If w hen pulling segments and selecting the Record Type “Net New”, this option cannot be selected or you see zero net new accounts the first time you run a segment, the market/model has not been activated to pull net new accounts.
To activate :
Go back to the “Markets” screen
Select the model referenced in the Market Analyzer segment
On the upper right-hand corner, select “Activate Net New Accounts” from the “Other” drop down menu
Once completed, the platform will let you know the Net New Accounts have been activated
An email confirmation will then be sent to confirm completion as well; this usually takes a few minutes to run and email to be sent
Cloning a Segment
If a segment currently exists and there is a need for a similar segment, the platform allows cloning segments.
For instance, to clone the NA ABM Q2 Model:
Select the desired segment
Select “Clone Segment” from the “Other” drop down on the upper right
After selecting clone segment, a new segment will appear prepended with “Copy of” in front of the name of the segment that was cloned. Be sure to rename the segment and make applicable updates to the filters
In this example, the cloned segment will be the same aside from region of focus
Archiving a Segment
If a segment is no longer needed, Market Analyzer allows users to achieve the segment . Similarly to cloning the segment, use the upper right hand “Other” drop down to select “Archive Segment ”.
These archived segments can be viewed later with the filter ic on in the list of segments .
After you have created or archived segments, there are options to filter visible segments. Select the "filter" image that is next to the “Search Segments” section. This will allow visibility or sorting of segments by Name, Last Updated, Date created, Type of model Lead or Account, and Active vs. Archived.
Reviewing and Exporting Segments
Once a segment is created, the segment accounts can be previewed by selecting the blue "expand" image at the bottom of the segment screen.
After selecting the tab, review the account basic firmographic information , rank, and score . If the list of accounts shows the desired types of acco unts, you can export the full list of accounts as illustrated below.
If the segment is missing something, go back to the segment and adjust the filters.
... View more
Predictive Insights ’ solution for Marketo utilizes the power of AI and predictive analytics to score and enrich your leads within Marketo in real - time, as well as push lists of leads from the Predictive Insights platform into Marketo. Predictive Insights leverages your data and combines it with P redictive Insights’ vast database to enable multi - dimensional predictive scoring. As a certified LaunchPoint partner, Predictive Insights’ native integration with Marketo extends the value of Predictive Insights AI - driven intelligence directly into your Ma rketo - based Marketing workflows, smart campaigns, and lead management processes. This guide will walk you through the required steps to set up the integration between your Predictive Insights platform instance and your Marketo instance. Following an initi al configuration to establish the connection between Marketo and the Predictive Insights platform, the guide details the integration phases in detail. In practice, this integration is executed through a synchronized REST API, using Marketo Webhooks which a re connected to Predictive Insights Markets (i.e., “predictive models”). Each Webhook is connected to one Market.
Setup Process Outline
Create an API user in Marketo to be assigned all API permissions
Create a new custom LaunchPoint service and assign the API user to it
Configure Predictive Insights
Turn on Marketo integration in the External Platforms menu and follow the connection instructions in the pop-up window
Create fields in Marketo
Set up the outgoing and incoming fields for integration in the Market Integration menu for the chosen Market in the Predictive Insights platform
Create a new Webhook for each Predictive Insights Market in Marketo
Populate the fields by using the data in the Market Integration screen in Predictive Insights
Map the insights coming in from Predictive Insights
Create and set up a Smart Campaign that calls Predictive Insights Webhook(s)
Test by sending a few leads and checking in the logs to see if the data passed
Create a “fail-safe” Smart Campaign that picks up failed requests and runs them again
(Optional) Create a Smart Campaign for Batch Update
Several configuration actions that must be taken before the Predictive Insights - Marketo integration can be set up. The se configurations should be done by an administrator in both systems, preferably the person who will own the integration process.
1. Create a user in Marketo:
This creates an API User that enables Anaplan Mintigo Predictive Insights LaunchPoint activities. This User does not require an active email address.
Go to Admin →Security →Users & Roles →Invite New User
a. Step 1: Info
i. Provide indicative details. Note that the provided email address will not be used for mailing purposes.
ii. Click Next.
b. Step 2: Permissions
i. Select a role that includes all of the API permissions (e.g. API role with ABM)
ii. Check the API Only box
iii. Click Next
c. Step 3: Message
i. Click Send
2. Create a new LaunchPoint service
Go to Admin →Integrations →LaunchPoint →New →New Service
Provide an indicative display name and description
From the Service dropdown menu, select Custom
From the API Only User dropdown menu, select the user created in the previous step
* The next step will require you to copy two fields from the LaunchPoint page, so leave it open.
Predictive Insights Platform Configuration
1. Connect to Marketo
Go to Settings--> External Platforms
a. Turn on the Marketo toggle button
b. A pop-up window will appear asking for Marketo ID information. Follow the instructions described in the window.
c. Click Check Connection and Save.
2. Enable Insights for Integration
Each market transmits only the set of insights enabled for enrichment.
Go to the Enrichment tab for the desired Market and turn on each Insight you want to integrate by clicking Enable for Export at the bottom of the Insight tile.
Through the triple-dot menu at the top right corner of the Market screen, go to the Market Integrations screen. In this screen click on the Marketo RT tab
In the Exportable MIs box select insights to send to Marketo as part of the API response. Note that you can add all the insightsenabled for export for a given Market using the Add All Exportable MIs button.
In the Records Columns box select any desired fields related to lead information or metadata, including score and rank.
Marketo Integration Setup
After establishing the connection between the two systems in the Configuration step, set up the fields to store Predictive Insights data within Marketo and indicate how and when to transfer information.The integration setup includes the following steps:
Create New Webhook.
Set Up Smart Campaign.
Create a "Fail-Safe" Smart Campaign.
Create a Smart Campaign for Batch Update (optional).
Following the selection of the Predictive Insights , first in the Enrichment tab and then in the Market Integrations screen of the Predictive Insights Platform, then create equivalent fields in Marketo. This will permit Predictive Insights data to be accurately populated in the corresponding Marketo fields.
Important Notes Before You Begin:
The Marketo field types should align with the Predictive Insights field type. While most Predictive Insights are Boolean, some are categorical (text). Also, the Mintigo - Predictive Lead Score field is numerical.
Many Mintigo users opt to create new fields in SFDC and then automatically sync to Marketo to ensure consistency between both systems.
It is best practice to add a prefix such as “MI -” or “Predictive Insights-” to all MI fields created in Marketo to quickly identify which fields contain Predictive Insights data.
The Marketo field name need not match exactly the name of the Predictive Insight. Any name that communicates the meaning of the insight is fine.
Up to 95 insights can be configured per Webhook. Please contact your CXM if there are issues with this limitation.
Marketo RT Tab Setup
Configure Fields for Enrichment
1. Map Input Fields for Enrichment
This will define what fields Predictive Insights will be expecting to receive from Marketo when sending a lead for enrichment. Best practice is to map all the fields that have the asterisk next to them since that means these fields were used when the Market was built. It is also possible to map any fields used for matching even if they are not in the original data such as-Email Address (even if account Market), Company Name, Domain, Country, State, City, Zip code, and company phone. Note this could lead to slightly different matches than what was matched to these companies in the model.
2. Exportable Insights
This box shows all the Insights that will be sent from Predictive Insights back to Marketo through the single lead integration. (See Fig. 5) By clicking on the open space in the box, a dropdown list of MIs will appear. Choose all the Insights that you are interested in having sent to your Marketo system (those should be the ones you created fields for in Marketo). The list of available MIs to pick for this box is populated only by the MIs that were activated as Exportable in the Enrichment tab. In other words, this group of MIs will be a sub group of all the exportable MIs for this Market.
You can choose more than one at a time by holding the Ctrl key while clicking on items on the list.
It is possible to type a name of an Insightto find it straight away.
Add All Exportable MIs: By clicking the button that is just above the box, you will be able to avoid clicking all of the Mis if you wish. That can be very useful, just be sure you only have MIs you want in the list of Exportable MIs.
Copy Exportable MIs and record columns-Allows you to copy MI setup from a different type of integration in the same Market (one of the other tabs in this page). For example, if you have already set up Insights in the Async Bulk API and now you want to have the same MIs in your Marketo setup, all you need to do is to select the right integration in the dropdown field and then click the Copy button.
3. Record Columns
Similar to the top box, you can choose Insights to be sent to Marketo for the enrichment. In here, you can pick the indicators that are related to the record itself and not to the company for which the record was matched to. This information is related to the data sent for enrichment, so most of these Insights are on the metadata level. Common fields used in this box: Predictive Insights Account Score, Predictive InsightsAccount Rank, Webmail Email address, Predictive Insights–Company ID, Created On, Valid Company Domain
4. Persist Data - ON
Keeping this on means that leads that are sent to Predictive Insights are also kept in a specific list under this Market. This list is exportable, so it allows for a list of all previous enriched leads to be always available. In addition, data from the records is used to improve matching and coverage in the Predictive Insights database. Predictive Insights does not add the data sent to it from customers to the Predictive Insights database, just uses it in order to direct its own data collection.
Once the data is added, click the Save button in order to keep it. Without saving, changes will not be committed.
6. Market Integration - ON
T o ggle this switch on to enable the integration between the systems for this Market. This will not actively start any integration actions but will allow Pr edictive Insights to respond to future enrichment calls to this Market.
After the setup is complete, you will be able to access the information needed to setup the Webhook.
Configure API Connection
1. Download Field Mapping - clicking this button will initiate a download of a CSV file holding the mapping information you need for Marketo. The table in the file has three columns:
MI Name-The name of the Insight that will be exported from Predictive Insights. This is the same name that in shown in the Enrichment tab
Fields Type-Describes the type. A field can be Boolean, Text, or Number
Response Field Name-This is the internal field name that will be used in the API response, and the value you will need to paste in the Webbook’s Response Mapping section in the Response Attribute column (further explanation below)
2. View Payload - will lead to a pop up window with this information:
API URL-holds the URL that Marketo will use in order to call Predictive Insights. Use this field for the URL field in the Marketo Webhook setup
MarketID-No use for this field in regular setup since it is already part of the URL
Request Payload-This box describes the structure for which Marketo will fill a Webhook call with a lead’s information in order to send it to Predictive Insights. Use this field for the Template field in the Marketo Webhook setup
Create New Webhook
Market o uses Webhook s in order to communicate with third party services such as Predictive Insights . Webhook s support the use of GET and POST HTTP request methods to push or retrieve data from a given URL. In Market o, a campaign will use a Webhook for the action of sending lead information and receiving the scoring and enrichment for that lead.
1. Enter Connection Information
In order to create a new Webhook in Marketo, follow the steps listed below:
Go to Admin --> Integrations --> Webhooks --> New Webhook
a. Webhook Name: Name the Webhook
b. URL: The WebhookURL value should be copied from the Predictive Insights platform-In the Market Integrations screen, go to the Marketo RT tab and click the View Payload button
c. Request Type: Select POST
d. Template: The payload template value should be copied from the Market Integrations screen-go to the Marketo RT tab and click the View Payload button. Copy the text box under Request Payload
e.Request Token Encoding: Select JSON
f. Response Type: Select XML
g. Click Create
2. Field Mapping
Next, the newly created Marketo fields needs to be mapped to their corresponding Predictive Insights . Make sure you have the table from the Predictive Insights platform that describes the fields names, their types and their sync names. This can be downloaded from the Marketing Integration menu, by clicking the Download Field Mapping button.
a. Pick the Webhook you just created in the menu to the side (note that Marketo has the tendency to jump back to the first Webhook in the menu) and edit the Response Mapping for the Predictive Insights field
b. Add a field and choose the corresponding Marketo field.
The Predictive Insights field format is data.field_name. The Marketo fields names are shown in their API format and not in the Friendly Label (normal name) format.
Save frequently! If you encounter an error, you will be required to go through the process again.
Make sure that all field types match. The field types are detailed in the Field Mapping file which can be downloaded from the Market Integrations screen in the Predictive Insights platform.
In case where more than one Webhook is needed (for integration with more than one Market for instance), using the Clone Webhook will allow an easier setup.
This concludes the setup of the integration connection . The following s teps describe how and where records are enriched within Marketo .
Smart Campaign Setup
A Smart Campaign is required in order to call the Webhook. In practice, the Smart Campaign initiates the activation of the Webhook.
1. Create a new Smart Campaign.
a. Name the new Smart Campaign.
b. Click Create.
2. In the Smart List, choose which events would trigger the Webhook call.
Depending on the use case, the common triggers are Lead is Created, Fills Out Form, Data Value Changes.Consult your CXM for additional guidance if needed.
3. In the Flow tab, choose which actions will take place once the campaign is triggered.
Add the Call Webhook action and choose the Predictive Insights Webhook.
This step is taken in order to assure that the integration has been set up successfully and is working as expected.
Activate the Smart Campaign that was just set up from the schedule tab.
Check trigger: make an action that will activate one of the triggers.
Depending on the triggers, create a test lead/fill a form/change a value
After ~10 seconds, go to the results tab and verify that the lead you created/changed shows up in the list. The activity type should say Call Webhook and the detail will mention the Webhook name. If there was an issue, the detail column will specify a type of error
If there was no issue, click on the link in the lead name column to open the lead’s activity log
In the log, look for activity type Call Webhook. The records under it should show all the Insights that were updated. Insights that to did not update the data in their field (for example-data was False and the new data value is also False) will not show any update. If no issues are found there, the campaign should be kept on for ongoing use. The integration will update any lead activating the triggers.
Create a Fail-Safe Smart Campaign
It is possible that errors will occur in the communication between Marketo and Predictive Insights. These errors can happen because of Marketo-related issues, Predictive Insights-related issues, or connectivity problems. It is highly recommended to set up Marketo so that in case a Webhook call fails, Marketo will retry sending the call again. In most cases, these errors will be fixed in a short time period (a few minutes to a few hours). Therefore, the best solution is to create a Smart Campaign that will collect failed calls and have them sent again in within a couple of hours.
There will only be one re-try for every failed call. Marketo does not allow a simple continuous retry using a Webhook. If a group of leads was found not to have been enriched, the Batch Update can be used to enrich them assuming that the integration issue is gone.
The following steps detail how to set up such a Smart Campaign:
Create a new Smart Campaign or clone the previous Smart Campaign that was created.
In the relevant Smart List, add two actions:
Webhook is called- Webhook: Choose the Predictive InsightsWebhookGo to the corner and add a constraint for Error Type. To add types, click on the green ‘+’ sign and add three HTTP Errors: 401, 403, 404.
Member of Smart Campaign-first field is ‘not in’, the second field will have the name of the fail-safe campaign.
In the Flow, put a 2-hour wait actio n, and then put the same action as in the first Smart Campaign - A Call. Webhook activity that calls the Predictive Insights Webhook. If something was failing the call, we want to wait a little for it to get fixed and only then re - send the call.
In Schedule, edit the Smart Campaign Settings to run each lead through the campaign flow only once every day. The idea here is to allow leads to be enriched even if there is no availability for a longer time period.
Activate the campaign. Any failed calls will be automatically re-tried once after 2 hours.
Create a Smart Campaign for Batch Update (Optional)
If there is a need for lists of leads to be enriched (for instance: enriching all leads in the system right after first integration or when a newly purchased list comes in) there is an option to create a “Batch Update Smart Campaign” in Marketo.
Marketo doe s not allow a batch list call to a Webhook (a Webhook is designed for single lead triggers only). In order to overcome this, it is possible to create a Smart Campaign that receives batches (lists) and sends them to another Smart Campaign that has a Webhook . The first Smart Campaign would receive a batch and send it to the other Smart Campaign as a one - by - one list of leads, which Marketo will allow to send to the Webhook.
We are first going to create the Smart Campaign that uses the Webhook, and then create the Smart Campaign that will send leads to the Webhook Smart Campaign.
Create a new Smart Campaign (Batch Predictive Insights Enrichment2). This campaign will be identical to the regular Predictive InsightsSmart Campaign, except for in the Smart List menu where it will have the trigger Campaign is Requested. The value for the trigger will be Marketo Flow Action. Activate the Smart Campaign. It will only enrich things that are specifically sent to this campaign.
Create another Smart Campaign (Batch Predictive Insights Enrichment1) . This campaign’s Smart List would have a list that needs to be sent for enrichment. If you are creating this campaign for future uses, leave this part empty.
Under Flow for the second campaign (Batch Predictive Insights Enrichment1) put the action Request Campaign . The campaign that will be requested is the campaign created in Step 1. When you want to enrich a list, go to the second campaign, and adjust the Smart List. Once you activate this campaign to run once, the campaign will take all the leads in this list and send them one-by-one to the first campaign created. This campaign will send them to the Webhook for enrichment. In case of an SFDC integration:-Create the fields in SFDC first and let the fields be created in Marketo through the sync process-Verify that the fields have been successfully created in Marketo.
Q: How can more than one Market be connected through the integration?
A: Yes, but a separate We bhook needs to be set up for each Market. Once done, the Smart Campaign will call all the created Webhooks in its Flow step. Note that each Market will generate unique scores and ranks. For this reason, set up a unique score and rank fields for each Market. This is not required for equivalent MI fields since MI values are the same regardless of the Market.
Q: I’d like to replace the integrated Market with another Market. How can the switch be done?
A: There are two options:
Create a new Webhook (refer to the Create a New Webhook section in the guide).
Edit the current Webhook according to the differences between the Markets. For instance, remove mapping fields that are not necessary or add new MIs that are expected to be returned. Note that the URL has to be edited since every URL points to at a specific Market.
Q: How do I know the data types of insights for which fields need to be created in Marketo? How should those fields be named?
A: The Marketo fields should suppor t Predictive Insights data. Once the Insight s have been selected in the Enrichment tab, they will be available for export as a list in the Market Integrations screen/ Download Field Mapping button. The exported list includes the MI names and data types. The Marketo field names do not have to match the MI names in order for the integration to work, yet it is recommended to use something similar to the original name and add some Mintigo identifier before the actual name. For example - “ MI – Hiring a Large N umber of Employees ”.
Q: I tried to send a Smart List through the Webhook and nothing seems to be happening. How can this be solved?
A: It is most likely that Marketo is preventing your leads from passing to the Webhook. Marketo does not allow lists in the same Smart Campaign as a Webhook. Please refer to the Batch Update section in the guide and follow the steps.
Q: How can I route leads to be enriched by specific Markets based on Mintigo data?
A: Multiple Webhooks have to be u sed. First, create a Webhook that points at a Market that serves for data enrichment purposes. Then, send the leads to be enriched in other Markets using Webhooks, based on relevant Predictive Insights . Please contact your CXM for additi onal support.
... View more
If you are unable to access Predictive Insights, please review the common issues below and follow the troubleshooting steps. If you continue to experience issues with access, please contact customer support.
I haven’t received the initial login email (“welcome email”).
Check your spam mail folder or email routing rules to track down the missing email.
I am unable to log in through the login page at predictive.anaplan.com.
Do you have an active account? If you're uncertain, check that you received an initial login email from email@example.com.
If you suspect your account is inactive, contact your administrator and ask them to access the Users & Roles screen and make sure the toggle button next to your user name is enabled.
Have you entered the correct Company Name? Make sure to use the company name as shown in the initial login email.
I received the initial login email but haven’t set up a password in the 7 days following the receiving of the email.
Contact your administrator and ask them to resend the email through Settings/Users & Roles. Make sure they check the ‘Force change password on next login’ checkbox and click on the Re-send Welcome Email button accordingly.
Follow the instructions in the (new) initial login email to set up a new password for your account.
I forgot my password and I’m unable to login to the platform.
In the login screen, click the ‘Forgot Password?’ button.
Fill in the company name and email address. Note that the company name should be exactly as it appears in the initial login email you have received.
Click on ‘Submit’ and follow the instructions in the email to set up a new password for your account.
I used the wrong password too many times and I’m now locked out of my account.
Contact your administrator and ask them to reset your password through Settings/Users & Roles. Make sure they check the ‘Force change password on next login’ checkbox and click on the Reset Password button accordingly.
... View more
These are some common questions and issues regarding prioritizing lists in the Predictive Insights platform. Please review the information below and follow the instructions to troubleshoot the issue you may have encountered. If you continue to experience issues, please contact customer support.
What is the List Prioritization flow?
Predictive Insights allows you to score and enrich (append with indicators) lists of records against existing markets (predictive models).
I’m trying to upload my file to the platform and receive an error message stating I have duplicate column headers.
This error message means that you have more than one column with the same name in the file. In order to resolve the issues, simply rename or remove the duplicate column before uploading the file again.
I’m trying to upload my file to the platform and receive an error message stating I have a bad header or a missing header.
Open the file you’re trying to upload and review it. The error message can imply one of the following cases:
If you found a column that contains data but has no header, please add a header name.
In some cases, a column with no data may appear empty while that is not truly the case. This would typically happen in columns that follow “regular columns” that contain data. To resolve this issue, please delete any empty columns before uploading the file again.
I received an email with an error notification following the list submission.
In the field mapping stage, you may have selected an incorrect optional field type. Upload the list again and identify the correct field type to map the fields in your file into, according to the information below.
Generic Date Time Field for dates.
Generic Number Filed should be number only and not include any special characters such as $, commas etc.
Generic Word Field should be strings only.
Generic Varchar Field for alphanumeric data.
I have run my list through the List Prioritization process, but the records in the list weren’t matched (weren’t appended with a company ID) and no indicators were appended to the records.
Open the file that you have uploaded, review the column headers and check whether any of them contain a special character. If so, remove the special character before submitting the list to be prioritized again.
In the export file, I can see that scores are returned even though the records seem to be unmatched (there are no Company ID values for these records).
A record can have a score (score>0) and not be matched if there are lead-level features, custom fields or validators used in the predictive model. In order to check whether such data was used, go to Other/Market Settings; If lead-level features or generic fields were mapped, or if validators were enabled in the market, it is likely these fields are the source of scores for the unmatched records.
... View more