fbpx

Real-time customer profiler and its application

When you start any marketing effort, you either have no data or mountains of it. If it’s the latter, all customer data is usually stored across different platforms, making it almost impossible to unify without advanced tools or knowledge. As a result, you can’t rely upon it to understand your customers.

Verfacto ingests all that data and provides you with a source of truth about the customer journey, their experience, and dedication to your brand.

You can cross-share this data with other tools you use. Based on the first-party data only, you can empower the whole marketing department by allowing them to integrate this real-time data with any marketing tool running on your shop: Klaviyo, Google Optimize, Hubspot, Meta Pixel, Google Pixel, etc.

For example, Verfacto enables you to build a segment of customers who were noticed on a website with an “unfulfilled buying intent” after returning to you for the 3rd time in a row from a “Direct” channel last week and pass this segment to Klaviyo.

Further in the article, we will describe how to set up the Real-time customer profiler, how it works, what attributes it can pass to other tools, and what are the most common use cases.

Setup and Installation

We are using a small tracking script that collects behavior data of all your visitors. In a way, it is similar to what Google Analytics does. The main difference is that data collected by Verfacto is not based on cookies. This allows identifying the same user long after the cookie expires and passing through most ad blockers.

A tracking script is generated for every Verfacto account. To let it start collecting data, it must be injected into your website by placing it into the <HEAD> HTML element.

				
					<script src="https://analytics.verfacto.com/entry-point.js" defer onload="window.VerfactoEntryPoint.loadTrackingScript('{ENTER-TRACKING-ID}')"></script>
				
			

If you are using GTM (aka Google Tag Manager), it’s better to use a modified version of the script to ensure it doesn’t affect the loading of other resources on your website.

				
					<script type="text/javascript">
    !function(){
      var t=document.createElement("script");
      t.type="text/javascript",
      t.defer=!0,
      t.onload=function(){
        window.VerfactoEntryPoint.loadTrackingScript('{ENTER-TRACKING-ID}')
      },
      t.src="https://analytics.verfacto.com/entry-point.js",
      document.head.append(t)
    }();
 </script>
				
			

Side note. If you use Shopify or WooCommerce, there is no need to add the script manually. Verfacto plugins for these platforms—with almost one-click installation—will cover the tedious work and inject the tracking script automatically.

Step 1 is complete and we are good to go! From this very moment data starts piling in and your customer profiles are building up—cross-device and without cookie lifetime limits of Google Analytics.

How it works

After the tracking script is set up, each and every session at your website is not only feeding raw data into Verfacto but also retrieves aggregated data as soon as the session is identified. This identified and aggregated data is the customer profile.

Whenever that happens Verfacto a user object to a global javascript variable “vfUserData.” This user object will be built and maintained for the rest of the customer journey—from the first visit to becoming a loyal customer. The global variable that contains these user objects, “vfUserData,” can be picked up by other tools and leveraged for their advantage.

Here is a sample JSON object of a “window.vfUserData” variable that is returned for each visitor coming to your site.

				
					{
  "attention_rate":-34,
  "atc_rate":-34,
  "recency_score": 4,
  "frequency_score": 3,
  "c_aov": 48.2,
  "c_days_since_last_purchase": 19,
  "c_ltv": 189,
  "session_count": 31,
  "days_since_last_visit": 2,
  "is_returning": true,
  "last_channel": "Email",
  "cap_scores": {
    "affiliates": 0,
    "direct": 0.87,
    "display": 0,
    "email": 0,
    "organic_search": 0.09,
    "other": 0.03,
    "other_advertising": 0,
    "paid_search": 0,
    "referral": 0.01,
    "social": 0
  },
  "most_frequent_device": "mobile",
  "s_discount": true,
  "s_fullprice": true,
  "s_loyal": true,
  "s_refunders": true,
  "s_sales": true
}
				
			

All attributes are divided into several logical groups:

Behavior in the current session

A group of attributes that are mainly based on the data from the current session:

  • attention_rate—as the name implies, this attribute reflects how much user’s attention you currently have [-infinity … +infinity]
  • atc_rate—a real-time prediction of an “add to cart” event based on all users’ past activity.

Customer historical data

A group of attributes that identify the historical information about purchases and the monetary value of a customer:

  • c_aov—average order value of a customer (AOV)
  • c_days_since_last_purchase—days since last purchase
  • c_ltv—lifetime value of a customer (LTV, or CLV).

Behavioral historical data

A group of attributes that identify the historical information about a visitor:

  • session_count—the amount of sessions a visitor made across all devices
  • days_since_last_visit—days since last visit
  • last_channel—the channel identifier of the previous session
  • cap_scores—acquisition profile scores that identify preferred or most frequent channels that a visitor returns to your website from
  • most_frequent_device—a visitor’s device preference.

Segments

A group of attributes that identify predefined segments that a customer/visitor belongs to:

  • recency_score—a score from 1 to 5 based on your customer interaction history
  • frequency_score—a score from 1 to 5 based on all customer profiles that are present
  • s_discount—true/false value showing whether a customer prefers discount goods
  • s_fullprice—true/false value showing whether a customer usually buys for full price
  • s_loyal—true/false value showing whether a customer profile falls into the definition of high value customers
  • s_refunders—true/false value identifying customer profiles that tend to make returns of ordered goods
  • s_sales—true/false value identifying customer profiles that tend to wait for bargains.

Use cases

Knowing detailed information about any individual customer may be interesting but yet rather useless when analyzing data retrospectively. On the contrary, aggregated data leveraged by machine learning algorithms enables you to see trends and patterns hidden from human eyes.

Building lookalike audiences on Facebook

If you just started advertising on Facebook and you don’t have a lot of customer data yet, interest-based targeting is a good solution to start searching for your audience.

You can begin by testing different interests to see what works and gain some traction. Once you start building your customer base and gather enough Facebook pixel data from your website, you can upscale your targeting methods and use lookalike audiences.

For example, you could set up a custom pixel that fires an event when a user with a higher LTV comes back to visit your website.

				
					<script type="text/javascript">
  if (vfUserData.c_ltv >= 150) {
    fbq('track', "ViewContent", {
      highValueCustomer: vfUserData.c_ltv,
    });
  }
</script>
				
			

This enables you to create a Custom Audience on Facebook and set up a lookalike audience. The beauty of this approach is that your audience will always be up to date.

Email marketing and Klaviyo metric-triggered flows

Klaviyo is a great tool that not only solves a number of email marketing and SMS marketing problems out of the box but also has great integration capabilities. One of them is the metric-triggered flow—an option that allows you to queue people for a flow when they take a certain action.

For example, you could track a custom event “Recency” and later on configure the flow in a way that a special welcoming email would be sent to your old customers who abandoned browsing your shop.

				
					<script type="text/javascript">
    var _learnq = _learnq || [];
    _learnq.push(['track', 'Recency', vfUserData.recency_score]);
</script>
				
			

Key takeaways

If you advertise on Facebook and Google and actively do email marketing, you can do much more to target the right audience. Do not miss the opportunity. With Verfacto’s Real-time customer profiler, you can collect and leverage advanced customer data with most of the marketing tools available on the market.

In this guide, we have covered the basics of the Real-time customer profiler. But it has many more capabilities to boost your marketing and business—we will cover them in future articles.

To try it out, simply sign up for Verfacto. In case of any difficulties or product-related questions, contact our support team or book a demo with one of our experts.

Table of Contents

Ecommerce tips and news right to your inbox

Enter your email and stay into the industry trends and Verfacto news