Why HubSpot’s Director of Analytics Wants You to Make Data More Human
Posted on: 21/05/2020
Reporting and attribution have revolutionized marketing in every industry. Marketing data allows businesses to make more informed decisions about their audiences’ needs, challenges, and interests.
For years, demographic reporting has done wonders for marketers. Data points such as age, ethnicity, gender, location, education, and employment have informed marketing teams and heightened the impact of campaigns across the board.
Unfortunately, there’s only so much that demographic data can tell us about the people searching for and purchasing our products and services.
2020 is the time to make data more human.
Key Takeaways & Expert Insights
Humanizing a historically inhuman component of marketing (such as reporting and attribution) can be tough. Let’s talk about a few ways to do this.
Firstly, start to switch your reporting focus from demographic data to psychographic and behavioral data. This should be a major change your organization makes in how and why you collect data. We’ll talk more about how to do this later in this article.
Secondly, expect democratization of data. Today, data is difficult to gather and is typically controlled by a dedicated marketing analytics team. While this may be a good way to collect and manage your company’s data, it can create a silo that alienates your data from the people at your company who can actually apply and learn from it.
In 2020, expect analytics teams to start making data more accessible for all employees. To jump start this process within your organization, invest in systems that automate your marketing reporting and make it easier for your employees to access that information. Teach employees how to properly apply this data, too.
Strive for efficiency and alignment across your organization. Here are a few ways to do this:
1. Align all departments within the organization around being a data-driven company and agreed-upon goals.
2. Choose and define a home for your marketing data. Where does your data live? Is it currently available? How can employees access it?
3. Avoid common marketing reporting mistakes, including one-off reports and random data pulls. These waste time and produce inconsistent data that doesn’t speak to your audience or contribute to real results.
4. Document your marketing reporting process — this might look something like: gather data, define important patterns, automate, and repeat. Instead of assigning different roles or teams to each part of the process, consider dividing up your marketing reporting workload by data source. This will allow the same people to see your data through from collection to application.
Keep Investing in Attribution to Identify Bottom-Line Impact
In 2020, continue focusing on marketing attribution reporting. When did customers first touch your brand? When did they last touch it? Where are your visitors and customers coming from?
According to our 2020 State of Marketing report, only 50% of companies are currently using attribution reporting — and almost 30% answered “No.”
Attribution tools such as Databox, Looker, and HubSpot’s Marketing Hub are getting better and better at producing more accurate and timely marketing data.
Start Using Behavioral Data as Your Guiding Light
Consider behavioral data king in 2020. If you aren’t already, start tracking behavioral data and define relevant marketing key performance indicators (KPIs).
For example, instead of tracking demographic data like age or location, start paying attention to behaviors like click-through activity, online purchases, search query information, and on-site engagement.
Behavioral data is more impactful data than demographic data, and it helps ensure compliance with GDPR, CCPA, and future data regulations (which we’ll touch on below). It also helps you identify and understand (and then effectively target) people based on their actions.
Behavioral and psychographic data points represent people showing up and taking action. They are a better indication of interest and intention. Demographic data doesn’t adapt to users and behaviors in the same way.
According to our 2020 State of Marketing report, almost 25% of companies reported their top marketing priority for this year was closing more deals. If your priorities are the same, behavioral data-based reporting can help meet those goals.
At HubSpot, we focus on the data from a few tools (our own instance of HubSpot, for example) and on-page site data. We review each order of event sequences (i.e. how visitors came to land on and interact with our website) and take a look at where people landed, how they got there, and if they were satisfied with the results — if they converted, shared, and/or made a purchase.
For example, let’s say someone lands on the HubSpot homepage through a LinkedIn post. They explore the website by clicking on a few pages and finally, on the fourth click, land on a product page. There, they click to request a call with our sales team.
This approach to collecting data better informs our business strategy. It provides a deeper understanding of what content and which site placements help users find the information they need to perform the actions they want … like submitting a demo request.
“2020 is when we’ll get automated insights at scale. Between the continued development of natural language processing and AI-engines trained to understand which metrics are important to different industries, we anticipate being able to get relevant insights sent directly to you so you know where you need to focus.” — Jocelyn Chen, Analytics Lead at Seer Interactive
For those seeking buy-in for behavioral data, I recommend running an initial analysis of your site and reviewing the results. Establishing on-site patterns can actually provide more data to be shared with your team and business leaders.
These patterns can then integrate your business teams across marketing, sales, service, and more.
The top-line argument here is that demographics and segments have never been about the people — your audience and customers. When you understand how people are using your site, you can improve it immensely.
Stop Ignoring Important Data Regulations
Regardless of what types of data you introduce in 2020, you must stop ignoring data regulations. It’s time to get proactive with data regulations, rather than reactive. This is especially true for GDPR and CCPA. Even if you’re not operating in the European Union or California, these regulations will control your access to data — especially demographic data.
This is yet another reason to lessen your focus on demographic data. Not only will it become less available through more regulations, but it also poorly represents your audience. Demographic data is essentially bucketing people into segments that they didn’t choose and don’t make sense for them.
When you stop defining people by their demographic data, you gain a better understanding of their actions and intentions — and you work in tandem with data regulations, not against them.
My Recommended Reporting and Attribution Tool
Marketing reporting and attribution is not a manual process — an intelligent, robust tool is almost always required. But there isn’t one perfect tool — what works best for you is dependent on where your data lives and what platforms you’re already using. This will determine what you need going forward.
If you are dealing with a lot of disparate data sources, I’d recommend Looker. Looker is a very powerful query tool with a lot of flexibility. If you are more streamlined and have a lot of your data on one platform, there’s a ton of solutions — for example, if your data is in HubSpot, we offer many powerful reporting and attribution features.
Take a Deep-Dive Into the State of Reporting and Attribution
Dive deeper into HubSpot’s survey data by clicking the download button on the image below.
Editor’s note: This article was researched in December 2019 and January 2020, and was originally published in early February 2020 in our State of Marketing Report. A lot has changed in the world since then, so keep that in mind as your process these trends and data.