New Account Segmentation Template using Clustering Transform

Data Prep recipes continue to get better with every release! With recipes, you can quickly blend, enrich, and export your data with an easy-to-use visual editor and real-time data preview so you can see the impact of your transformations on the data as you build.

Another key benefit of the Data Prep recipes? The learning-powered Smart Transformations you get out of the box: Detect Sentiment, Predict Missing Value, Clustering, and Time Series Forecasting.

Now, what good is all this no-code ML transform at your fingertips if you are too busy to try them out in your org? In Winter ‘22, we introduced a pre-built app template called “Cluster Analysis: Account Segmentation” so you can start using cluster analysis on your data in just a few clicks!

Getting Started

To make use of this app, you simply install a new app and choose the “Cluster Analysis: Account Segmentation” template from the template gallery in Analytics Studio. You can search by cluster, segmentation, etc.

The app comes with a recipe that uses Account, Opportunity, Opportunity Line item, and Product objects, and two dashboards that consume the datasets that will be created by the recipe.

This is the Segmentation Overview dashboard where you can compare the clusters and explore key differences.

This is the Whitespace Finder dashboard where you can compare a specific account with the aggregated metrics and product purchases from the segment it belongs to.

Once you click on “Continue” to install the app, it’ll validate that your system meets the minimum requirements on the required objects and fields. Then you just give it a name, and you’re off to the races!

Isn’t that easy?

To make sure the app installs correctly, your org needs to meet the following requirements:

• Data Sync must be enabled, and Account, Opportunity, OpportunityLineItem, Product2 must be connected.
• The following fields for each object must be enabled in Data Sync:
  • Account: Id, Name, Industry, BillingCountry
  • Opportunity: Id, AccountId, Name, StageName, Amount, CloseDate, IsClosed, IsWon
  • OpportunityLineItem: Id, OpportunityId, Product2Id, Name, Quantity, TotalPrice
  • Product2: Id, Name, Family, ProductCode
• Data Sync for those objects must have been run within the previous 7 days

If your app installation fails, please review the requirements above to make sure that you have Data Sync set up correctly and run successfully before trying to create the app again.

App Overview

This app creates a recipe that reads data from Account, Opportunity, Opportunity Line Item, and Product2. It builds sample metrics from Opportunities such as average deal size by Account, win rate by Account, total account lifetime value, etc, and then it clusters Accounts using those metrics.

The recipe then applies the generated cluster labels to the products/opportunity line items. This is one example of identifying product preferences in each of the clusters.

Segment Comparison Dashboard provides a few ways to compare the clusters based on the sample metrics in the recipe. The app is built with orgs that use Accounts/Opportunities/OpportunityLineItem/Product in mind, so if you don’t use those entities, the results in your dashboard may vary.

In the Whitespace Finder dashboard, you get an automatic, visual comparison of an account’s list of opportunity products against the account’s segment’s list of opportunity products.

And that’s it!

The app is ready for customization to meet your company’s business needs. For example, to change the number of segments being generated by the cluster node or make any other adjustments, simply click on the “Clustering Accounts” node in the recipe, edit the Cluster transform, and update it.

To change the sample metrics, add new ones, or to apply additional filter logic, simply go to this section of the recipe and adjust as needed.

Next Steps

I would love to hear your feedback on this template and any new template ideas! You can share your feedback with me directly on the Trailblazer Community or join us on Slack at DataTribe! I want to put this power at your fingertips but I need your help to make it work for you – please share your use cases and feedback with me so we can make it better!

To learn more about clustering transform under the hood, please go to the Deep-Dive on Clustering blog post.

Written by