HubSpot’s 6 Steps to Building Out a PQL Model
Editor’s Note: The following article is based on a recent episode of OpenView’s BUILD podcast. You can listen to the full episode featuring Kieran Flanagan, VP of Marketing at HubSpot, here.
Many SaaS companies are taking advantage of a marketing opportunity that is uniquely suited to a product led go-to-market strategy. Unlike the marketing qualified lead (MQL) scoring model, which relies on basic activities such as website visits, webinars and gated content, the product qualified lead (PQL) scoring model is based on prospect behavior as they actually use your product. This hands-on approach provides valuable insight into buyer intent that you just can’t get based on clicks, downloads, or email opens.
One company that has used PQLs with great success is HubSpot. As VP of Marketing at HubSpot, Kieran Flanagan has deep experience both with launching a PQL strategy and refining it for optimal performance. During a recent conversation with me, Kieran shared some great insights and details about how HubSpot integrated PQLs into their powerful go-to-market arsenal.
Getting Started: Launching a New Idea
HubSpot has a history of innovating by giving people within the company the room and resources needed to take a big idea and run with it. In practice, this often takes the form of launching a micro company within the company. And, when one of those “side experiments” works, it is ultimately integrated back into the larger organization. Kieran traces HubSpot’s work with PQLs back to their initial freemium model experiment, which was a product called Sidekick.
Sidekick was an email notifications tool that was developed and launched by an internal HubSpot team including Brian Balfour, Christopher O’Donnell, and Mark Roberge. Designed for salespeople who needed to know if people were interacting with their emails, Sidekick was a freemium product that used touchless upgrades throughout the user experience. When Kieran joined HubSpot, Sidekick was being phased out, but his team was able to learn a great deal from the project and apply the insights to HubSpot products.
Building on this foundation, Kieran and his team developed three different kinds of PQLs:
- Hand-raise PQLs, which began as call to actions that were inserted at different points within the product to invite people to reach out and learn more about HubSpot’s paid products.
- Usage PQLs (the most common type), which provide enough free product use to deliver value, but do not allow for ongoing use.
- Upgrade PQLs, which require upgrades for access to gated features.
Building the Foundation: Getting the Basics Down
In the beginning, HubSpot leaned heavily on hand-raise PQLs. After a while, however, they transitioned to focus mostly on “triggered” PQLs like usage and upgrade models. These models allow freemium users to access and use various features until a certain trigger is met. One successful example of such a feature is HubSpot’s email templates for sales people. In this case, a freemium user was provided with a certain number of templates for free, but once that maximum was reached, the user needed to upgrade.
The process of determining which features to gate and which to set up with triggers is iterative and involves a lot of testing. There isn’t really a one-size-fits-all mathematical equation to apply. There is, however, a trend toward moving from gated to freemium/triggered PQLs, and it’s not just because this approach tends to convert better. “In a world that is moving toward a product-first mentality, it’s better to give value before you try to extract value for the company,” Kieran says. In other words, with a product led strategy, you have the opportunity to actually get in there and solve problems for the prospect. This allows you to prove your product’s value in real-world situations before you even try to generate revenue.
In addition to this big-picture philosophy of giving before asking, HubSpot also focused on features that could evolve into acquisition channels on their own—ways to bring more users into the HubSpot ecosystem. This included features that had a certain viral factor as well as items for which there was a lot of organic search volume. “We would put the reasons to gate something or give it away into a matrix,” Kieran says. “And we’d also look at how good the feature would be for acquisition and monetization. But, after doing general research, a lot of our choices were based on gut feelings.”
Testing and Iterating: Refining Your Approach
To get from their initial launch to the refined PQL strategy HubSpot uses today, Kieran and his team went through a number of steps. While this process will vary from company to company, it’s always helpful to get a sense of the general framework.
- Step 1: See what’s happening. The first thing Kieran’s team did was build a dashboard that gave them visibility into the individual funnel of each PQL point. The “pretty incredible” dashboard allowed them to see how many times users interacted with each category of PQLs, how many users had converted into customers, and other details such as average deal size and overall MRR from each PQL.
- Step 2: Assess performance. With all the data collected via the dashboard, the team was able to stack rank PQLs based on which ones had the highest conversion rates. It also allowed them to see which ones had the most potential for improvement. This enabled a more intentional approach to further development.
- Step 3: Run a few simple tests. They began with some variable testing on call to actions (CTAs)—placing different versions at different points within the product. While they no longer use such tactics, in the early days one example was inserting a CTA on the dashboard and CRM to invite the user to reach out for a walkthrough/consultation on the CRM. This turned out to be their highest converting PQL point, but after six to eight months, they shifted their strategy.
- Step 4: Make some choices. Even though the CRM consultation CTA was working well, the team decided to do away with the hand-raise approach because it wasn’t a great product experience to have CTAs popping up in the product. It wasn’t an immediate or easy decision, but it was the right one in the long run.
- Step 5: Refine your approach. After the initial tests and decisions, the team implemented their PQL model, which pops up in the app when a trigger has been met. In the initial stages, this model allowed the user to either talk to sales or upgrade through a touchless sale.
- Step 6: Optimize interactions. The next step was to increase the communication options for the PQL model. Today, this includes options that allow the user to schedule a meeting right within the app (instead of going back and forth over email), add a note to have someone call back, or access a live chat for an in-the-moment response. Getting all this up and running was complex because of all the integration between these different parts, but it delivered the user experience Kieran and his team were after.
Throughout this process, Kieran used a few very specific metrics to track progress and performance. “Because we were our own little company within HubSpot, we focused on freemium metrics,” Kieran says. “Our ‘north star’ metric was our Weekly Active Team rate (WAT) because we know that if we can increase WATs, we will have better retention and better monetization.” For their purposes, Kieran’s team defined an active team as two or more people from a company who were using the product in a meaningful way on a weekly basis.
Integrating PQL: Understanding Your Audience
One of the nice things about working product qualified leads into your go-to-market strategy is that it’s not an all-or-nothing proposition. In fact, it’s most common for SaaS companies to employ PQLs alongside more traditional MQLs; and there are reasons to use both tactics. At the end of the day, they are both just different ways to acquire new customers.
“We decided to have a free CRM and some free sales tools because we thought it was the best go-to-market for those products,” Kieran says. “HubSpot’s model has been to generate leads through the valuable content we create, and then turn those leads into MQLs. What we learned from the early Sidekick model is that there’s this other way to layer in different kinds of conversion points that allow us to also acquire customers by letting people use our software for free.”
Over the years, it has become clear that there are different audiences for each kind of strategy. “People are still thirsty for knowledge, and that tends to convert into MQLs,” Kieran says. “But there’s also a cohort of people who want you to demonstrate that you can solve their problems before they make the decision to upgrade and pay you money.” It really boils down to understanding your audience and how you can most effectively solve their problems.