How to Drive Website Conversion

A fireside chat with Jaleh Rezaei at Mutiny

Websites have been around for decades. But the truth is, most websites don't do a good job at converting visitors to customers. This is hugely problematic for driving efficient growth. Product-led companies in particular rely on their websites for their top-of-funnel, and later as they layer in a top-down strategy, they need to use their site for account-based marketing. I sat down with Jaleh Rezaei, who recently raised a Series A from Sequoia (and me) for her company Mutiny, to learn about specific best practices for optimizing your website.

Website marketing has historically seemed like a branding activity. But I think you consider yourself a growth engineer. Why is growth engineering the right approach to driving website conversion?

Before starting Mutiny, I led marketing at Gusto and was tasked with driving fast revenue growth. 

What I realized when I took over the marketing team was that we were spending a lot of money generating top-of-funnel demand: things like content, marketing, advertising, direct mail -- you name it. If it was a channel, we were investing money into it. However, that investment hadn't materialized into revenue at the rate that we expected it to.

Our spend was going up, but revenue wasn't scaling with it. We knew we had a problem to address.

To learn from best practices, we spoke with all the marketers we knew. It was clear that personalization of the website to each visitor was the gold standard for driving conversion. So we created 80 different landing pages, showing each visitor a customized view of case studies, quotes, coupon codes, and other content. We ended up growing 100x during the time I was at Gusto.

But we had to build all this personalization manually, since there was no better way. Many of the exciting things we wanted to do were bottlenecked by engineering. Whether it was improving the website for different audiences and bifurcating our traffic to show restaurants a perfect experience, or to show founders the perfect experience for them -- everything required an engineer.

Ultimately it was clear to me that every company had to personalize their website well. But no software vendor made it possible to run a marketing team like an engineering operation, in the way that we did -- even though all the required data on website visitors was now available. So my co-founder and I left Gusto to build Mutiny, which is the no-code way to build mass personalization into your website.

What are the steps to personalize your website to each visitor in this way?

First, you need data on who is coming to your website: for example, their industry, size of company (SMB or large enterprise?), role (executive? team member?), use case (which of your solutions is most relevant?), and whether they're an existing freemium customer. 

Amassing this data requires you to integrate different data sources. You can get information about people who have already interacted with you from Salesforce, Segment, Marketo, and Hubspot, to name a few sources. That population tends to make up only 10-20% of a company's website traffic, but it's valuable because these people are already engaged. Another 30-40% of the data comes from IP data, such as Clearbit and 6sense, which provides data on the company where the visitor works. Then you should layer on top of that certain behavioral data (which we have our own data source for) that shows how they are navigating your website, and based on that, what might be their role and use case.

Next, you'll figure out which audiences to prioritize. We use machine learning to automate this analysis in our product, to automatically surface the best segments based on their conversion rate, and share what playbooks work for those audiences. 

Third, you'll trigger personalizations within your website. You can staff engineers on this, but we have a no-code editor that does this and then helps you see the degree to which the changes actually impacted conversion. When you launch a test, you need a hold-out group within the target audience that does not see the experience that you launched. 

For example, let's say you're looking to improve your website for healthcare visitors by changing the customer logos on your site to reflect healthcare companies. You'll show some of those healthcare visitors the new logos and show the others the generic version, then measure the delta in conversion rate and see whether it's statistically significant. Then you'll continue to iterate to improve the rate. 

Of course, you'll want to collapse this entire process so you can improve your website really quickly. I believe in leveraging speed as a core competency of marketing. 

I often see companies "uplevel" their website when they start selling to enterprise, but they risk alienating their SMB customers in the process. How can you cater to both segments?

You definitely need to create different website experiences for each segment. One size doesn't fit all.

To simplify your website for startup visitors, delete content from your website. Less is more; there may be one entry pain point you need to highlight. You can mention the word "startup" directly in the headline. The call to action can be "get started."

For the enterprise segment, there are a few changes you'll want to make:

  • Evolve your descriptions: migrate from "easy-to-use" to "built for large teams", mention that you can manage multiple teams using roles and permissions, and otherwise speak to compliance questions.

  • Uplevel the logos to appeal to enterprises, initially using the largest logos that you have, but over time as you acquire more customers, you can customize them to the industry. 

  • Pull in your Salesforce data to see if a visitor came from a target account and let your lead scoring model influence change the call to action, since a visitor's level of engagement will influence what next step you want them to take. 

  • Remove your prices, so that your prospects don't anchor to your entry-level pricing when your sales team is looking to sell them the enterprise version with a broader set of features. 

  • Personalize your site to the visitor's role. For example, if you sell to line leaders but require participation from IT in the buying decision, you'll want to change the messaging to cater to the IT buyer when they visit your site. 

A webinar and conferencing software company in Europe that we work with saw a 200%+ increase in enterprise leads by changing their website in these ways. 

How should one's growth strategy (product-led vs. top-down) impact one's website strategy?

The interesting thing about PLG companies is that when most of them achieve a certain size, they layer in a top-down motion. And many top-down companies try to add a self-serve component later. So most companies have two funnels by the time they reach about 200 employees, and their website needs to be useful in each funnel. 

Particularly in a PLG context, it's important to know if a website visitor is already a user. If they've already signed up for your free product, the call to action shouldn't be "sign up"; it should be "upgrade." One of our customers personalized their messaging in this way and drove about 80% more upgrades.

What kind of resources are required to do this kind of mass personalization well?

Companies already spend a lot of money to generate web traffic -- both in terms of the per-lead cost, as well as the salary of the person who sets up the AdWords or Facebook ads. But historically they haven't harvested every lead that these efforts could generate: those visitors come to the website and don't convert. We've made it so that with an additional four or five hours per week on average, you can personalize your website to each of those visitors.

***

If you’d like to learn more about Mutiny, check them out here. They've made it easy to try out their product on your own site for free.

If you'd like to hear from more founders like Jaleh about once a week, sign up for my newsletter here: