Predict the likelihood of any conversion for your online business to drive more value from each visitor

Edwin Halim

Apr 24, 2019

Edwin Halim

Apr 24, 2019

As a marketer, you know that targeted engagement increases the possibility of a conversion to a significantly higher level. When machine learning can help you predict which user segment is most likely to convert, you can streamline your efforts in just the right direction and hit your marketing goals.

When we use the word ‘conversion’, the first thing that comes to our minds is purchasing a product on an online retail site or completing a booking for a ticket or a hotel on an online travel site. However, there are several micro conversions that build up to a final conversion. Depending on your industry and KPIs, a conversion can be any sort of goal you want users to achieve on your website. It could be signing up for a newsletter, reading a blog post, requesting a test drive, completing a loan application or even requesting for a demo. Every business has custom conversions which drive value across different lifecycle stages of their users.

Table of Contents
  1. A New Predictive Model: Custom Conversion Prediction (CCP)

  2. Driving More Value in Ad Channels with CCP

  3. Let’s See CCP in Action Across Industries

A New Predictive Model: Custom Conversion Prediction (CCP)

We have been working on a powerful predictive model to help an online business predict the likelihood of any goal conversion they may have. Knowing how likely a given segment is to make a purchase, complete a subscription, or book a test drive, any online business will be able to improve targeting precision across channels. Custom Conversion Prediction (CCP) makes it easier for marketers to run multiple machine learning models for different conversion goals that are relevant to their business.

Driving More Value in Ad Channels with CCP

Custom Conversion Prediction (CCP) helps target users who are more likely to convert in the next 7 days in advertising channels. This helps marketers across industries design more relevant advertising experiences, while driving their Return on Ad Spend (ROAS).

Let’s See CCP in Action Across Industries

The best part about CCP is that it gives you the flexibility to set multiple custom conversion goals. Here’s how you can leverage CCP for various types of conversion goals based on your industry and business KPIs. Let’s look at a few examples.


While completing a purchase or a sale is the primary goal tracked for most eCommerce websites, there are several other steps in the funnel that lead to a final conversion. With CCP you can predict users’ likelihood of adding an item to their cart or viewing products. Predicting these micro conversions that lead to the final conversion like a sale or a purchase, can help you target users early in their decision making journey. You can additionally use CCP to track the likelihood of a user to engage with any other services that you might be offering like payments, wallets or tickets that are usually different from your primary conversion funnel.


You can use CCP to help you predict users who are likely to buy your ancillary products and drive your ancillary revenue. For travel websites, predict the likelihood of a user to rent a car or book a hotel along with their flight. For airlines, CCP can predict which users are likely to buy additional services like an upgraded seat or insurance.


One of the primary goals for publishers is to convert users into loyal readers or subscribers. CCP can help predict the users who are most likely to become paid subscribers. Use this information to personalize or show relevant offers on your subscriptions to users across Google/Facebook Ads and your website.

Lastly, using CCP is simple. Just send these segments to either Google Ads and Facebook for remarketing or use it to personalize user experience on Desktop and Mobile Web.

Insider’s predictive technology provides insight into which customers are going to engage and where you can improve in order to ensure conversion. The best part is, with predictive segments, you can pinpoint the smallest groups of customers according to their overall traits.

Implementing predictive modeling to your customer experience strategy can be a game changer in how your brand communicates to your users. There’s so much more to discover when it comes to Insider’s predictive tech. Read more here or connect with us by dropping a quick line at We’ll be happy to hear from you and help you reach your most pressing KPIs.

Edwin is oversees Insider's customer success team in Indonesia and the Philippines. He has 6+ years of experience in digital marketing, with a special focus on multichannel CRM strategy, growth hacking, and A/B testing. Before Insider, Edwin was a technology consultant at Accenture and co-founded his own Digital Agency.