Understanding the Idealized Customer Profile

For Untitled customers with a Shopify store connected as a Source and over 100 historical orders, Untitled has released a machine learning model that you can train on your own customer data to find website visitors who are most likely to become loyal, high-value customers.

What is an Ideal Customer Profile (ICP)?

Most business owners have an idea of what their “perfect” customer is like. Maybe it’s a very sophisticated buyer who understands what a difference the high quality materials of your items can make. Maybe it’s an entry level hobbyist who will benefit from your easy-to-follow instructions. When you run your business every day, recognizable patterns emerge about who “walks in the door” and is most likely to become a happy, long-term customer: that’s your ICP.

Obviously, you’d like to seek out people that fit your ICP, rather than wait for them to stumble across your store. That’s where Untitled can help.

How is an Untitled ICP generated?

When you sync your Shopify store as an Untitled Source, you’re pushing your customer data into the Untitled platform. When you click “Create ICP” within Untitled, the platform looks at that customer data and matches it against our robust identity graph with over 300 million people records, adding more information to your existing data. This additional information can include over 100 attributes that are not captured by Shopify.

Next, all of that data is fed into a machine learning algorithm that is basically trying to answer one question: Which combination of attributes is most common across the highest-value customers in this customer data?

The result of this model, which may take hours to train depending on how much customer data you sync, is a list of visitors to your website who match the customized filters the machine learning algorithm found. This list is automatically created as an Audience in Untitled.

How should I use the Untitled ICP Audience?

The results of the ICP model might not be intuitive. In fact, that’s the point of using machine learning to sift through tens of thousands of data points: to find patterns that are invisible or overly complex for us to draw on our own.

Treat your ICP Audience as VIPs. Lifetime value doesn’t mean they are necessarily ready to buy today, but that they are likely to love your brand overall.

Email content that offers discounts is usually welcome from customers, alongside some brand story or overview of your product offerings, rather than a specific product being retargeted.

You can also use your ICP Audience as a custom audience in Facebook/Meta or Google Ads. By uploading this list directly to those platforms, you can target only those people in your ICP and have confidence to possibly raise the bid and frequency of your messaging above what may be a baseline for a ‘cold’ audience.

How will my ICP change over time?

Each day, the ICP model looks at your new website visitors and includes the visitors who fit the profile into your ICP Audience. The ICP Audience will always contain the top 15% of lifetime value customers, so roughly each time a new member joins the Audience, a member will leave.

Each month, the ICP model looks at your new Shopify orders and uses that new data as additional training. Depending on the attributes of your new customers, the ICP may adjust its selection criteria for the Audience.

Summary

Untitled is bringing custom machine learning to your Shopify store with the ICP model. Use your data, plus our platform, to bring new insights to light, creating a whole new Audience for you to target.

Once you have over 100 Shopify orders, create an ICP Audience within Untitled. Remember that ICP Audience members may be at different points in their buyer journey, so try to cultivate them into a long term fan of your brand, staying connected without driving immediate purchase urgency.

Questions about how to create an ICP model within Untitled, link up your Shopify store to Untitled, or something else? Email us at [email protected].