*** Adobe Analytics has been chosen to represent all Web Analytics tool references in this article, due to its dominance in the user personalization field ***
There are many reasons why a company would love to understand what each individual user is doing on their website. It can be the desire to provide better web experience for users, to offer personalized deals, to look for potential leads, or simply out of human curiosity.
Whatever the reason might be, it is undoubtedly good to know more, since information is king. The more we have, the more informed we are to make better decisions.
Having said that, is Web Analytics the right tool? Yes and no.
Sure, the only way to obtain user web browsing behavior data is through Web Analytics, but don’t expect the reporting/analysis to be done using a Web Analytics tool. Simply because, Web Analytics specializes in aggregated data, and moreover, PII data are not supposed to be in the Cloud!
Before we go into the how, let’s take some time to understand what are the possible benefits from tracking individual users:
- Generate potential leads through the understanding of individual user’s product interests
- User X has looked at 5 baby products in the past 3 days. Let’s send her an email offer on baby products.
- Personalize offline services
- User X has been checking on her insurance policy ABC quite frequently this week. Let’s inform her agent to follow-up.
- Personalize communications
- User preferences are ever-changing (and changing fast!). But, changes in user preference will not be reflected in the outdated online profiles we have. With Web Analytics, we can find out her current preferences through her browsing activities.
These are just some of the low-hanging fruits. The potential to reach out to customers proactively in a timely manner is endless.
Let’s take some time to get ourselves familiarized with the core ingredients for tracking individual users on our website.
On the Web Analytics front, we would minimally require:
- Visitor ID
- Customer ID
- Event Tracking
To understand exactly who is doing what on our website, below are a couple more things we need:
- PII Data
- Data/Customer Analytics team
Visitor ID is simply an unique ID provided by Adobe to identify each and every user on our site.
This ID is stored in a persistent cookie found in the user’s browser.
Whenever a new user visits a website that has been tagged, Adobe will generate persistent cookies and send them to the user’s browser. In one of those cookies, you will find the Visitor ID which is used to identify the user for the current and subsequent visits to our website.
For repeated visitors, Adobe will first look for such cookies. If the cookies are there, Adobe will simply use the Visitor ID found in those cookies.
In the case where such cookies cannot be found (e.g. user clears cache and cookies, surf anonymously, etc), Adobe will treat the user as a new visitor and issue a new set of cookies.
If you want to know what other Adobe cookies do, here’s a link for you:
Below is a screenshot of what a Visitor ID looks like:
Customer ID is an unique ID that we, as website owners, give to a customer that has logged on to our website.
Typically, this Customer ID sits in a database within the company, where all our precious and private details (PII) are stored.
PII (Personally Identifiable Information) includes our name, date of birth, identification ID, address, contact number, and of course, the mother’s maiden name (Don’t ask me why, the banks want what the banks want).
With Customer ID, the company has all your personal information (obtained through consent) and is able to serve you information that are privy to you. That includes your order history, parcel tracking, savings account information, credit cards, etc.
Below is an example of how Adobe stores my Customer ID:
Tracking is basically divided into 2 parts:
- Page-Level Tracking
- Event Tracking
For Page-Level Tracking, we are capturing only page-related information on load of webpages. This tells us what pages did our web users visit, and how often were those pages visited.
Event Tracking is where we can capture each and every single user interaction that is happening on our website. Button clicks, video plays, form completion, you name it, we have it.
Page-Level Tracking gives us a very surface view of what our web users are doing, whereas Event Tracking gives us the in-depth view of what exactly our web users are doing.
If you want to truly understand what each individual user is doing on your website, you probably would want to know far more details than just which pages he/she has been on.
Personally Identifiable Information. As mentioned earlier, PII include our name, date of birth, identification ID, address, contact number, etc.
These information are dear to us (hence the emergence of Blockchain and Bitcoin), and should never be exposed to the public.
If data leak is bad, giving criminals access to your customer information by putting them up in the cloud is heinous.
Therefore, PII should never be stored in any Web Analytics tool, as their databases are most likely in the Cloud.
So… if we can’t pass these information to our Web Analytics tool, how am I supposed to identify the users…?
Don’t worry, we will cover that later.
Data / Customer Analytics Team
There are so much confusion between Data Analytics and Web Analytics. I’ll try to explain it as simply as I can.
- Mainly online data
- Commonly used tools include Adobe Analytics, Google Analytics
- Mainly offline data
- Commonly required skills include R, Python, statistics, SQL
- Commonly used tools include SAS, Tableau, SPSS, Qlik
There are definitely much more differences between these two, but I shall not go into that here.
Why do we need a Data/Customer Analytics team?
Earlier on, we mentioned that PII are stored securely in a database that sits within the company. That is offline data. All the transactions and invoices? Offline data too.
In order for us to marry our online and offline data, we will need to create the bridge, and then pass it on to the Data Analytics team along with our online data.
And by bridge, we absolutely mean the customer ID that is present in both our online and offline data.
This is how we better understand what every customer is doing on our website without putting them at risk.
How Information Flows
To provide a clearer picture on how all these can be achieved, I have come up with a slide to give you an overview of how information flows and how it results in actionable insights.
We will break it down into 3 parts:
- Customer visit
- Marrying of online and offline data
- Actionable insights
In this example, we simulate a customer viewing one of our product pages.
When a customer lands on our website, Adobe will first check for a valid Visitor ID. If there isn’t one, Adobe will generate it for the customer.
When he/she logs in, we can then capture Customer ID (generated by the company) and link it to the Visitor ID we have. This way, we can then link all Web Data back to their individual owners.
When he/she views one of our product pages, we then capture the Product ID, which now gives us the final piece of the puzzle – understanding the current product interest of each of our customers.
Marrying of Online and Offline Data
Online data tell us what individual anonymous users are doing on our website, while offline data help to put a face and a profile on each of the anonymous users.
As you can see, Web Data collected with Adobe Analytics (AA Data) only contain the Customer ID and Product ID.
*** Yes, typically, we do capture more product information than just the ID. Still, those data will not be as comprehensive as what we have in our offline data ***
AA Data only allows us to understand what each anonymous user is interested in.
But with the data marriage, we can now map the Customer ID to the customer profiles we have stored in our secured database. Same goes for Product ID.
We, now have the full picture.
Thanks to the data marriage, we can now generate reports showing who viewed what in the past week/month.
This is highly indicative of which products are each user genuinely interested in. This is especially true for non-eCommerce companies.
In general, there are no incentives/motivations for users to browse for specific products (banking, credit cards, insurance, software, etc) without an intent to buy (or at least research). This piece of data gives us the opportunity to sway their buying decisions in our favor.
Whereas for eCommerce sites, it could be the case where the users are just randomly browsing as a form of retail therapy. Having said that, for eCommerce sites, we can still derive a highly accurate metric to gauge users’ product interests.
It is common for users to browse through tons of products in a single session, but they will not check out the product details for every single one of them. Hence, Product Details View can suggest a much higher product interest over a simple Product View.
Web Analytics is powerful on its own, and it can be made much more powerful when paired with offline data sources.
Remember, Web Analytics is not just about reporting vanity metrics. It is about identifying real-life opportunities and then working towards collecting the data required to generate actionable insights.
Page views, visits, bounce rates mean nothing on their own. Put them into the right context and pair them up accordingly, miracle happens.
That is the beauty of Web Analytics. It is not simply just data collection or reporting, it is Science. Actionable insights are what separate the gurus from the wannabes. So, let’s not settle for vanity reporting but instead, work towards drawing actionable insights, one at a time!