Let’s take a step inside the data-driven demand generation marketing team. The biggest concerns on the CMOs radar are that the acquisition costs are too high and not hitting their pipeline or revenue goals. Now looking at the data, we know that not only are they spending a lot on paid and organic traffic, but the quality of the traffic is good, and it’s not converting.
So, of course, the next question would be – what can they do about it? A common answer is to focus on website conversion rate optimization, which involves running online experiments. That’s something you can put a budget around and prioritize but recognize that your executives are going to want to see impact based on pipeline and revenue and probably want to see it fast.
Back in 2017, the Harvard business review published an important article digging into the power of online experimentation. In it, they correlated successful business outcomes to a culture of experimentation.
Image Source: The Surprising Power of Online Experiments (Harvard Business Review | Link)
The article cited examples like the one below from Bing, who tested multiple different colors on their site, ran experiments. and realized an incremental $10 million in annual revenue from these experiments.
Image Source: The Surprising Power of Online Experiments (Harvard Business Review | Link)
Similarly, Google ran a test with 40 different shades of blue on their site. When they ran those experiments, they achieved $200 million in incremental revenue. Given these results, should we, as demand gen marketers, be running the same experiments?
In our opinion and experience, no, you should not.
You’re not Google or Bing. Leaving aside traffic considerations, you’re trying to influence B2B buyer behavior over customer journeys. And the reality is that groups of buyers that consider enterprise solutions are not going to buy based on the button color or other small cosmetic changes.
This is important because experimentation comes with a cost. Not only do you have people and the technology costs of running online experiments, but also your organizational ability to make decisions. So, focus on the elements that would deliver revenue and influence those B2B buyers when you’re thinking about experimentation.
When we think about the B2B buying journey or the revenue funnel it’s common to conceptualize it as a series of buyer stages. As prospects progress through those stages, they do so through exchanges, in which you’re offering something to that prospect in exchange for something else. The offer could be some content in exchange for their attention, an event, or an opportunity to speak to the sales team in exchange for their contact information. Ultimately those offers are how they learn more about your solution and how it would benefit them.
From our experience and the testing that we’ve done, the highest leverage use of experimentation for the demand gen org is to improve the relevance of those offers and the ease of engaging with them throughout the buying journey. Of course, we always want to ensure we measure the impact of those experiments based on the KPIs that matter – pipeline and revenue.
What does it mean to optimize offers? There are three components to an effective offer. One, of course, is the offer itself. That item you’re proposing to exchange with that visitor or prospect for them to understand your solution. The more relevant it is, the more effective your ability to convert them will be.
The second important aspect is how you frame it. Our primary focus here is the headline and Call to Action (CTA). Your headline is important because a visitor will spend five or ten seconds deciding if they want to stay on your site or hit the back button and go somewhere else. So, entice them to continue reading the content on the page.
Finally, the third element of the offer is the exchange and how they provide what you want. Most likely on your site this is a web form, but it doesn’t need to be. It’s increasingly common to see conversational marketing tools (chatbots) that accomplish the same thing by providing that medium of exchange for the offer.
Let’s look at some examples of how you could optimize your offers.
Landing pages are a great starting point for thinking about your offers. Many of you are probably running traffic to dedicated landing pages and putting an offer in front of the visitors hitting it. But not every visitor is interested in the same offer. In the example below, we recognized when working with a customer that they had three viable offers for those visitors coming through their paid campaigns. And rather than only showing them one, we use data to dynamically personalize the offer itself as well as framing and the page layout to reflect what might be most relevant to that visitor.
When we ran the experiment against the static landing page we saw a 44% improvement in revenue per visitor.
For most of us the most trafficked page on our site is the homepage. And on your homepage the “above the fold” section at the top gets most of the attention. Many of us think about our homepage in the context of welcoming the first time visitor and introducing your solution as in the example below.
For SaaS and Demand Generation websites it’s common to have a lot of returning traffic. Since return visitors are familiar with your solution, it wouldn’t make sense to show them that same offer. In an experiment, we targeted these return visitors and the solutions they showed interest in and presented them on the homepage. In this case, those offers were buried in the site and require additional navigation. By presenting this offer they would likely be interested in and serving those directly on the homepage, we saw almost 55% improvement in conversions coming through this page.
You can also target well-defined buyer stages. In the following example, we have a customer with a freemium model where visitors on the free plan come to the homepage and see a CTA or a button prompting them to “Upgrade Your Plan”. The baseline experience was to take them to a set of SaaS plan tiers where they could select the one that they would upgrade to.
Using this data, we can identify the specific plans most relevant for any individual and offer them directly on the homepage. The framing included the benefits and replaced the CTA with the cost of that specific plan we recommend. Since we recommend a single upgrade plan, we bypassed the plan selection (and the friction it created) and took them directly to the credit card to upgrade. By removing friction and presenting them with a more relevant offer, we saw an almost 70% improvement in revenue per visitor coming through this experience.
The most common mechanism of exchange for the offer is the web form, and as a result, we spent a lot of time optimizing them. It’s important to recognize that there’s a lot of friction for the visitor when they encounter one of these forms.Even if they’re interested in the offer, they face the prospect of handing over their email and other personal information, which often presents a big hurdle. Since it’s common to see drop-offs at this stage, we would like to take those contact forms and reinforce the benefit and the value to the visitor filling them out. In the following example, we tested an updated version of the form page resulting in an 85% improvement in conversions.
If you have the data, you can get sophisticated with offer personalization. It’s common to see pages like the one below. It is a product page that contains multiple offers for different personas within the organization. Unfortunately, when you try to put them all on a single page, they compete for attention and blend in, making it hard for users to know which one is relevant for them.
In this case, we target specific personas visiting the page based on data we had in the marketing automation platform and identify the most relevant offer. By testing variations that replace the default experience with a single focused offer, we see an almost 50% improvement in revenue per visitor.
It’s possible to waste time, effort, and money optimizing inconsequential elements of your website. For demand generation marketers, the highest leverage things to focus on are the offers – specifically their relevance to the visitor and the ease of engaging with them.
Before you undertake this experimentation it’s important to make sure you have solid revenue insights. What that means is, evaluating your existing offers as well as future experiments based on their pipeline and revenue contribution.
Some of the personalized examples above require some segmentation. Our recommendation is to prioritize segmentation based on the differentiated intent and addressable size of those segments. We often find that marketers are running building audiences that can only address 5-10% of their audience, or ones that don’t have meaningfully different intent from one another. Ultimately those aren’t going to have much value when it comes to optimizing offers.
This is why we start with buyer stages as our starting point for segmentation because it a large set of well-understood segments with differentiated intent – buyers at different stages will naturally gravitate towards different offers. The vast majority of the visitors coming to a demand gen site fit into anonymous, known lead, active opportunity or existing customer.
Finally, when it comes to improving offers, start with common sense ideas. If you start thinking about your buyer stages, some opportunities should become apparent. For example, should a known lead see a lead capture form, or can we repurpose those pixels for something more relevant? Similarly, should existing customers see the “Request a Demo” or “Talk to Sales” CTA? Maybe there’s an opportunity to get them to support resources or event upsell them.
What’s stopping you from generating more revenue by improving offers on your website? If you’re a Demand Gen marketer and need help, feel free to get in touch.
Hotjar is a great complement to Google Analytics. Layering qualitative and visual data over the raw numbers gives you another dimension of insights.
But just like with your Google Analytics data, if you ignore key segments, you do so at your own risk.
Imagine, for example, that a heat map shows you that only 20 out of every 1,000 of visitors click on your Product Tour CTA. In fact, the scroll map shows you that only 15% of visitors even reach that section of the page.
You might conclude that the section and CTA don’t matter, and consider removing them.
Now imagine that all 20 of those visitors are leads – visitors who have identified themselves by signing up for a free trial, downloading a resource, or attending a webinar. Suppose that on average 15 of those 20 leads end up turning into opportunities. The Product Tour just went from wasted space to one of the highest-value interactions on the site!
Fortunately, it just takes a bit of work to begin segmenting your most valuable visitor data in Hotjar. Let’s look at how to do this with leads.
While leads might not be your most important identifiable visitor segment, for most B2B SaaS sites they deserve special attention. In fact, they’re already getting special treatment in your nurture campaigns. (Right?) And hopefully you’re personalizing offers and CTAs for them as well.
Still, the steps below will work for any segment you can identify. Target accounts, industry of interest, or existing customers can all be given VIP status in Hotjar.
Before you begin, make sure you have two things in place.
1. Hotjar Plus or Business
The free plan doesn’t support custom tags and triggers.
2. A way to identify leads on your website
Not sure how to do that? This post will walk you through it. And if you’re using Marketo, FunnelEnvy automatically syncs lead status with all your frontend tools – Google Analytics, Drift, Google Optimize, and yes, Hotjar.
Tag session recordings
Watching playback of visitor sessions is a great way to put yourself in your customer’s shoes. It’s also dauntingly time consuming. One day’s worth of recordings could take a month to view.
So clearly you need to prioritize what you focus on. Watching a half dozen leads interact with your website will yield more insight than watching a hundred anonymous visitors land, scroll, and bounce.
All you need to do is execute a single line of code when you identify a lead on the site:
Set this up, and you’ll be able to filter recordings later.
(See the Hotjar docs for more detail on how this works.)
Trigger heat maps
Instead of mixing clicks from anonymous visitors, customers, and leads all into a single heat map, you can create one for leads only.
If you’re using FunnelEnvy for Marketo, it’s as easy as adding a Trigger to Google Tag Manager:
Then create a Custom HTML Tag to fire the Hotjar code:
Create a custom poll for leads only
What page has the highest exit rate? What page do visitors spend the most time on? What are they looking for, and not finding?
The answer is probably different for leads compared with anonymous visitors. The only way to find out is to ask.
Lucky for you, you can trigger a custom poll with the same code that triggers custom heat maps.
Where to start
There’s a lot you can do to better understand (and more effectively convert) leads on your website. As a first step, just tag and watch some session recordings to see how leads navigate your site.
This requires a way to identify those leads in the first place. Solve that problem once, though, and you open up deeper insights in Google Analytics, custom playbooks in Drift, and personalization options in Google Optimize.
If you’re using Marketo, FunnelEnvy solves this for you. No need to bring in the dev team and turn it into a multi-month project. If you’re ready to start giving leads the special treatment they deserve, just get in touch.
You’ve decided to improve on your one-size-fits-all website content by serving personalized content to leads. You figure that a free trial user will literally never click “Start Free Trial” … but they very well might click “Buy Now.” Especially if you give them clear reasons to do so.
Great! So, how will you target these visitors?
It’s a straightforward process of identifying “leads only” behavior, then ensuring you’re able to activate this data on your site.
What do leads do?
The answer is unique to your product, but it’s not a trick question.
Here are visitor behaviors you can use to identify leads:
- Sign up for a trial
- Opt in for a lead magnet
- Click through on an email message sent to leads only
- Trigger a domain or company match to an account that’s in the pipeline
- Click “Log In” on the homepage
If you’re only looking to segment out leads in your analytics reporting, this might give you everything you need.
Your “Leads” segment is the set of all visitors who carried out any of the above actions. Even if you’re not tracking “Log In” clicks or using a firmographic data provider, you’ve got pageviews on /app, or /dashboard, or /whitepaper-download-thank-you. That’s enough to define a segment.
But to take the next logical step of serving a more relevant experience to these visitors, you’ll have to have this data available not just in your reports, but on the frontend of your website.
How to activate experiences for leads
Once you’ve narrowed down the list of actions that define “leads only ” behavior on your site, you’ll need to attach some sort of identifiable metadata to the user across your website.
If you can spare a few developer cycles, setting a first-party cookie is a good option. Whenever a visitor starts a trial, or signs up for a webinar, set a cookie you can use to identify that they’re a lead. All your dev needs to know is the exact trigger (or triggers), the name and value you want to use for the cookie, and when it should expire.
Once this cookie is set in the visitor’s browser, you can use it to activate personalization campaigns, experiments, customized lead magnet offers, and whatever else you think might get leads to convert.
If you’re using Marketo, you already have a source of truth for a visitor’s status in the sales process, along with useful metadata about their site behavior, lead score, and more. All packaged up into a cookie that’s already on your site.
In that case, the easiest path forward is to use FunnelEnvy for Marketo to activate this data, which you can then integrate with Google Analytics, Google Optimize, Optimizely, Drift, and whatever else you’re using. No custom code required.
What to do next
You can start scoping this project right now. Write down the actions that identify visitors as leads in your pipeline. Forward this along to your dev team, and ask them what it will entail to set a custom cookie for visitors who complete these actions.
Or skip the back-and-forth by signing up for FunnelEnvy for Marketo. We’ll solve analytics, targeting, and activation. You can move on to designing a higher-converting experience,
Your website receives visitors in different stages of the buying process, who have varying needs and priorities. You recognize this, so you’ve installed a personalization platform. Where to begin?
First, a word on what not to do. Do not get click-and-drag happy with your platform’s audience tool, and end up creating a monstrosity like “Returning visitors from Texas using Firefox on Mobile.”
These audiences are easy to target, but hard to reason about, painful to maintain, and impossible to extract value from.
Instead, start with leads.
Why leads? (And what’s a lead?)
The exact definition will depend on your customer journey, but broadly speaking a lead is any visitor who has identified themself on your website.
This might include:
- Free trial users
- Whitepaper downloaders
- Webinar attendees
Put another way, leads are visitors who are neither paying customers nor anonymous.
As for why you should provide a personalized experience for them, there are three main reasons:
- You can. They’ve signed up, so you know something about them.
- They’re your second-most-valuable visitor segment. (Customers are #1, but that’s a topic for another day.)
- Your current website is probably dominated by top of funnel content that they’ve already seen, and no longer find valuable.
Given how important this group is, it makes sense to provide an experience that’s relevant to them. But how do you do that without rewriting your whole website?
Where to personalize
Meet your leads where they’re already spending time. Finding out the answer to this question is as simple as segmenting your analytics data by leads, then looking at top pages.
The answer is probably “the Homepage and the Pricing page” but don’t take my word for it. Let the data tell you where to focus, and what kind of reach you can achieve.
Once you’ve identified the top pages visited by leads, you can further prioritize by focusing on the elements they see and interact with.
For example, Hotjar allows you to trigger heat and scroll maps with custom code. That means you can create a “Leads only” heat map of your home page. (If that’s too hard, just keep your focus above the fold.)
How to personalize
This step is where the magic happens. What unique questions do leads on your website have? What tasks do they prioritize? What does activation look like?
To help structure your ideas, look for chances to do three things: Educate, remove friction, and nudge.
Does your homepage hero heading tout your product’s core value proposition? That’s great, but your leads probably know it by now. Can you change it to outline an important differentiator?
Does the homepage hero CTA still say “Free Trial”? You definitely don’t need that. Does it make sense to link to your knowledge base, or a quick start guide?
A simple improvement you can make is to show your free trial leads a more prominent “Log in” or “Visit My Dashboard” button. There’s a good chance that’s what they came to click.
You can also disable widgets and popups focused on lead generation. Those elements, by definition, can’t provide you with a new lead in this context. All they can do is annoy an existing lead.
What steps does a visitor have to take before obtaining value from your product? Configure an integration, view a dashboard, import contacts?
When they were new to the site, pushing them toward this would’ve been overwhelming. Now that they’re more familiar with your product, though, they need this guidance.
Does your chat widget still ask “New here? Got any questions?” Why not start an onboarding-related conversation instead? A simple script along the lines of “Have you imported your contacts yet?” can transform this chatbot from a nuisance to a touch point for upgrades.
Stuck for ideas? Here are a couple of suggestions for websites we at FunnelEnvy know and love.
What you can do today
The first step toward obtaining value from a personalization strategy is convincing yourself that it’s worth the effort. So, start there.
What is the single highest-traffic page for existing leads? How many visitors does it get each month?
What’s the single most impactful element on that page? If you’re not sure, start with the hero heading copy and CTA.
What’s the current experience for leads? Is it relevant at all? Can you think of a message that would easily be 10 times more helpful?
If so, you’re onto something.
The good news is that the technical hurdles involved in making this change are solvable in any number of ways.
Your personalization tool might support targeting based on past visits to the /dashboard page. You might convince a friendly dev to set a cookie for new signups. If you use Marketo, FunnelEnvy lets you target by Smart List.
So take your newly acquired vision for a better lead experience, share it with the team. You’ll be hard pressed to find someone willing to fight to keep redundant CTAs on the page. I doubt anyone will argue that converting leads to sales is a wasted effort.
Starting personalization with a well-defined, high value, high reach, and observably different audience segment will make the difference between real ROI and a cringe-inducing vanity metrics report. So let’s go nurture some leads!
The tools that Optimizely gives you out of the box are excellent for lead gen or e-commerce sites. And while conversion metrics can and should be used for media sites, more often than not you’ll also want to be able to track the effect of any given experiment on pages per session.
The problem is that Optimizely, in most cases, will only let you track differences in whether or not a single event happened — a conversion. Optimizely also has built-in support for tracking the difference in revenue per visitor, but not for the difference in the average of any other metric that you might be interested in.
For the purposes of this post, I’ll be writing about how to get around that specifically for pages per session, but this can be used for any metric that can be tracked in Google Analytics.
Tracking Pages Per Session in Google Analytics
To get started, we’ll need to start tracking pages per session as a dimension in GA. Google Analytics will already track Pages Per Session as a metric, but since we’ll need the distribution in order to perform the analysis of the data we’ll need to track it as a dimension.
First, we need to create the dimension in GA.
Open up your Google Analytics account and click on the admin section.
Next, go the “Property” panel and click on “Custom Definitions” then “Custom Dimensions.”
Analytics plays a critical role in improving conversions in several ways including exploratory analysis and ongoing reporting. One of the main factors that can constrain your ability to optimize your site for conversions is the quality of the data you have available and your ability to act upon that information. There are an amazing array of web analytics solutions available to the marketer these days, but too often they’re underutilized in the face of other business priorities or technical and setup complications. That’s a gap in the analytics architecture, which is just as important as the product architecture for online business.