How to think about data
The data types available for personalization & how to bucket visitors into audiences
Data is a scary word for most marketers.
It brings forward a feeling of being overwhelmed — a constant stream of customer and visitor data points that is constantly growing into a rapidly flowing river seeping out of its banks.
Not only is data hard to keep track of, it’s even harder to unlock and utilize. Many experts have even coined a clever name for the phenomenon — infobesity — and claim that it’s causing marketers to suffer from information overload and not execute on stated goals.
At the same time, data also unlocks insights and opportunity that otherwise wouldn’t exist. You can identify cohorts of users that provide more revenue to your business, acquisition channels that you should be investing more spend in, and conversion rate optimization experiments (such as personalization and A/B testing) that can improve key KPIs such as trial and demo registrations. Data collection is all about gathering the knowledge necessary to deploy an experiment thoughtfully and strategically.
We’re believers that data shouldn’t be scary — and more importantly, it definitely shouldn’t be a hindrance in your decision to personalize your website. As B2B marketers, we can follow a simple framework to understand what data to collect from visitors, how to store it in a thoughtful way, and where to utilize it to deploy personalized experiences at scale.
To get started with data for personalization, you first need to understand the types of data points you can use to create an audience and personalize. In this chapter, we’ll cover the types of data points that you can collect and how to create audience for personalization.
Data helps answer the Who and What behind designing a personalized experience. In Part I, we’ll cover the data points you need to understand. And in Part II, we’ll cover how to use those data points to create audience buckets.
Part I: Data points you need to understand
We’ve found that the best way to think about data for personalization is to understand the ins and outs of four basic categories of data:
- Demographic data
- Firmographic data
- Behavioral data
- Contextual data
Demographic data is data that is all around about a person. You’re likely familiar with this type of data from any government, focus group, or university survey (such as the US Census or a BLS labor report) — it’s all about data points around a person. For companies, demographic data is important because it answers the who behind your customers.
Here are a few common demographic data points and why they can play important in your personalization campaigns:
- Name — A pretty obvious point to collect, but people love hearing the sound of their name in messages and on-site. Dale Carnegie, the celebrated author of How To Win Friends & Influence People might have described the power of a name best, “a person's name is to that person, the sweetest, most important sound in any language.”
- Email — Think of the email as the gatekeeper to a whole host of other potential information. It’s usually the first thing asked for in a signup form, and it's a key part of unlocking other channels of contact. By providing an email, you allow a personalization software to pull even more information on a user via an API from a data service such as FullContact, ClearBit or Datanyze.
- Title — Are you targeting a certain job title with your software product? This can be both a great way to adjust your on-site messaging to speak directly to the stakeholder’s you value most. For instance, if Growth Marketers and Content Marketers are two of your target high-value buyer personas, why not call out the benefits for each group when they arrive on-site?
- Location — We’re based in Texas, and if that’s taught us anything, it’s that people love where they are from. Use location-based targeting to offer geographic-specific offers and call out visitors with a greeting that appeals to their region. Or use colloquial language to connect in a deeper way (“howdy y’all” is a favorite in these parts).
Firmographic data are data points all about the company. It’s similar in nature to demographic data, but all of the data is around the business rather than around an individual. Here are some common fields you’ll see in firmographic collection, and how you can start thinking about it for personalization:
- Company name — Just as an employee name can make a person feel special and unique, a company name can do the same or more. Employees generally have pride for their business, and when you are able to show that you know their company before they tell you, they are generally pleasantly surprised by the experience.
- Employee count and revenue — More employees and revenue usually indicates a larger business. Plus, it’s an important qualifier for an MQL. In fact, SMB, Mid-market, and Enterprise are often differentiated in part by headcount and revenue. By identifying these properties, you can provide a unique experience for each group and use personalization to adjust your product offering in real-time to visitors from different segments.
- Industry — There might be no more important data point in B2B personalization than industry. An industry is the data point a visitor often associates most closely with and what a marketer wants to identify immediately from their visitors. When you want to truly wow a visitor — pulling case studies, testimonials, and keystone customers from the same industry can increase CTR and conversion rates.
- Software stack — Users of certain types of software might be more likely to purchase your SaaS product because of direct integrations. Or a current software stack might indicate that a prospect that comes into the queue could be an immediate good-fit customer for your product. You can use this knowledge to your advantage and personalize a site by highlighting an integration with a specific software or adjusting the situation to differentiate yourself from your competitors.
- Stage in the sales cycle — Should you treat a trialing user differently than a paying customer? Of course! Both parties have extremely different intents and goals while learning about your business & product offerings. Sales stage is a key data point that can be used for personalization.
While demographic and firmographic data are related to the who behind a visitor, behavioral data is all about what a visitor does while on-site.
Behavioral data shows everything about a visitor’s self-directed actions while on your site or inside your app.
And every person has unique behaviors when using digital products — just as they have unique behaviors and tendencies in the real-world. I might be inclined to take one route to the Austin airport while my coworker might choose a different group of roads all together. It just depends on how our brains are wired.
The same holds true for behavior online. I can take one set of actions on a page while a marketer at another company might use a site completely differently. Savvy marketers know that humans behave differently (it’s in our DNA after all), and when these nuances in behavior are identified, personalized experiences can be used to deliver & outperform competitors.
Behavioral data helps answer questions for your marketing team like:
- What is the person doing on-site?
- Where are a visitor’s eyes drawn on a page?
- What actions has a visitor performed on-site in the past?
- What content are they reading? Sharing? Bookmarking?
- Are they a first-time visitor? A repeat visitor?
- What’s their time on-site during this session?
- What’s their average time on-site?
- Do they hit certain pages on the site?
- Do they hit certain pages multiple times?
- Do they visit certain URLs more frequently than others?
- How are they getting to the site?
- Are they more likely to do different things on different days of the week?
- Are there certain kinds of events that they have triggered throughout their purchase history?
- Is this person a customer? Is this person a trialing user?
Behavioral data is nearly limitless — the only thing that limits your collection of behavioral data is your creativity.
As you think of use cases for personalization on your site, you can start asking questions in the style cited above and start collecting interesting behavioral insights.
Contextual data is tangentially related to Behavioral data in that it is also related to a user’s unique properties. And as the name suggests, this category of data gives context to a visitor’s session on your site.
Here are some common questions that can give contextual information on a customer:
- Are they on mobile, desktop, or tablet?
- Do they use an iPhone or Android?
- Are they on a Mac or a PC?
- Are they using Safari? Are they using Chrome? Heaven forbid, Internet Explorer?
- Where is the device’s location?
- Where is the actual person visiting from?
- What time of day is it?
- What day of the week is it?
- What month is it?
These data points might on first glance not to seem especially important, but as you dig into data, you’ll see that device type, browser type, location, and other contextual data points can have quite an impact on a customer’s conversion on-site.
Part II: Create an audience
Audiences are the groups of unique customers or visitors that will see a personalized experience in the same way. You create these groups and add rules to define who you wanted to be included and who you want to be excluded. It’s similar to defining an audience on Facebook’s ad manager — you create an audience based on criteria that you select from an audience builder.
For instance, if we want to target current customers with a Welcome Back message when they hit our homepage, we’d create an audience for that group of current customers using the following criteria in Proof Experiences.
We name this audience Users and we use Proof’s filtering to include all people that have hit our Segment event Logged in (count value is greater than 0). That event signals someone is a customer because our product is not freemium — so every Logged in visitor is a returning customer.
You can create an audience like above in a handful of ways, and you can get as refined or as broad as you want when creating one. An audience can be as small as 1 person or infinitely large.
Let’s look at another example. You can create custom audiences by targeting properties using AND and OR statements. For instance, if you wanted to target first-time visitors and the industry SaaS, you’d use the following logic in your personalization platform:
You can also use matching logic exclusions to select audiences that include everyone that match one criteria and exclude everyone that has another trait.
So if you want to target first time visitors in all industries other than SaaS (perhaps these are less valuable MQLs for you), you could create an audience using the following exclusion logic:
How we create our own audience buckets
Here’s how we like to think about audiences when personalizing our own website. The first thing we use to segment our visitors into an audience bucket is industry (a firmographic data point).
From our analysis to date, we have identified five distinct industries where we most frequently see customers:
- Other (a catch-all audience bucket that includes, real estate, health & wellness, brick & mortar, etc).
You can think of your most common industry by looking at your CRM data or introducing a survey somewhere on your site (more on that later).
Second, we want to create audiences to target people off of the lifecycle stage (behavioral data). We can ask a few questions around behavioral data to identify whether a customer is a:
- First-time visitor
- Repeat visitor
- Demo registree
- Demo attendee
- Trialing user
- Activated user
- Churned user
With these audiences in mind, we can tailor our imagery, headlines, and on-page content for a visitor’s unique characteristics.
For instance, we wouldn’t want to position our software in the same way for a first-time visitor as we would to a churned visitor. While for the first-time visitor, we want to focus our messaging for our social proof product around our value prop of increasing conversions by 10-15% — and for the churned user, we want to cater more carefully and incentivize them to give our service another try.
By now, you have a great understanding of personalization and the data points you can use when deploying web personalization experiments. You’ve also learned about how you can create audiences, and how we think of audience creation at Proof.
In the next chapter, we’re going even deeper into the data side of personalization. We’ll cover how to collect data for personalization, the tools you need to do it, where to store it, how to create a tracking plan, and naming conventions.