Deep Dive with Matt from Drift: Why they have over 20 different versions of their homepage, how they created the conversational marketing category and more
Welcome to the 2nd deep dive interview on MonkHD. We have Matt Bilotti, Senior Product Manager on the Growth Team at Drift, one of the hottest SaaS companies right now, and an expert on MonkHD. In the first deep dive interview, we talked to Dandrew from Sift Science on how to get 3–5x the conversions of traditional landing pages with LinkedIn lead gen forms and how to use intent data to power account-based selling.
Matt was previously a product manager at HubSpot where he helped drive millions in revenue for them. In this interview, we talk about why Drift has 20+ different versions of their homepage, how they created the conversational marketing category, and how to use conversational landing pages to get 3x the conversions of traditional landing pages and more.
Tell me more about Drift. What do you guys do? Who do you sell to?
Drift is a conversational marketing platform that helps sales and marketing teams connect with their best leads now.
Over the years, it’s been quite the journey to get to where the product is today. In the first year of Drift, we were constantly changing what the product was gonna be and who we would sell to in search of a market and a way we could make the business <> customer relationship more human. We changed that every 2 months for the first year. First, we were an email support tool, next, we were trying to sell a next-generation CRM, then we were trying to be the world’s best in-app product marketing tool, that didn’t take off either. We kept going and going till we found something that worked which was live chat for marketing.
In the market, live chat was mainly used as a support tool and we kind of reframed it and said it can be a marketing tool to talk to your customers and build your brand. This is where things started to take off and we evolved from there to the conversational marketing platform we are today.
Drift is generally purchased by marketing teams and people using it day to day are the sales people. It’s a widget you install on your website and helps you connect with your best leads in real time via chat and chatbots. Sales people can jump in and chat and go right to a demo or a call as opposed to the old school way of leads filling out a form and the sales team coming in the next day and calling all those people. So overall, Drift is a way to help businesses generate more leads and connect with those leads now.
Who’s your typical customer?
A B2B SaaS company. We are used by many other industries but B2B SaaS is our sweet spot. Companies that have a sales team of more than a few people often get the most value out of Drift.
We have 2 core offerings, one of which is a free product. People using free are generally startups — 1 to 10 people. They probably have a sales person or two but the company is using it more like a channel to talk to whoever comes to their website. The freemium offering helps the bottom of the market where it’s used by everyone, not just B2B SaaS.
Paid offerings are positioned towards companies that have a sales team. Companies with 100 or more employees, sales team of 10–20 people. That’s when you get the most value from Drift.
At the stage where you have a full sales team, the value you get often comes from a chatbot booking meetings for your team 24/7, automatic lead routing, targeted messaging, and more. It takes away a lot of the overhead you would otherwise have. It is useful for one sales person to be using it all day, but even more powerful when you have a lead flow, you already have a business generating millions of dollars. We help you make more of the funnel that you have already built.
What’s your role at Drift?
I was the first product manager on the team and my role at Drift has changed a lot over the years. I helped build and launch a lot of the core offerings — the chat widget, the chat bot, the CRM, the integrations, the conversations view, the mobile app, a lot of those core pieces.
Over the years, I have also done customer support, built out our remote customer support team, did a lot of the early product marketing, I was in all the product videos and produced them. My role has taken a lot of shapes and sizes and forms over the years.
Earlier this year as the team was growing, we were spinning up a growth team and I was tapped to move over to the team where we own most of the lead generation for the business.
We own the free product — how it works, how do we onboard new users, how do we get them activated, how we acquire them in the first place, how we build the viral loop in the product. Our product is inherently viral, which is a quite fun thing to work on. Customers install the chat widget on their site and it has a “⚡️by Drift” button and that drives their visitors to come to our site.
We also own the demand generation channels — bots, email and paid. We also use bots to set up conversational landing pages, which we drive traffic to.
And then we do experimental lead generation based on 3rd party sources. We believe that there is intent out there that exists before someone comes to our site and puts their email address in. There are things that signal that that company or person is going down a path where our solution will be a great fit for them. I work on a team that built that vision out.
Then we have website experiments where we have 20–30 different versions of the Drift homepage depending on the type of visitor that we also work on.
Bots for conversational landing pages sounds really interesting. Tell us more about that.
Conversational landing pages are a really great way to create a personalized experience with minimal work where a bot does the heavy lifting for you. People click on a paid ad and instead of sending them to a typical landing page, you send them to this conversational landing page where a bot talks to them and qualifies them and then books a meeting. You are extracting what you would put on a website or a landing page and making a lightweight conversational version of it.
We have a customer who says that a normal landing page converts 10% for them whereas a conversational one converts 30% for them. We are talking about a huge huge jump with these landing pages.
You can use these conversational landing pages for the traffic you are already getting from paid ads, emails, partnerships, and drive more value out of the activities you already do.
Do these conversational landing pages work only for simple products or also for complex ones?
Even complex ones. One of our customers who is seeing some of the best success with conversational landing pages is Rapid Miner, which is a really technical product, and you can see tweets from their CMO on how well conversational landing pages are working for them.
The power of the bot is that it’s a multifaceted way to work with your leads. You can ask your leads — “Hey how much do you already understand about our product?” Not much / A lot.
If they say not much, you can can educate them more about the product — Great, here are a couple of articles. Let me know what you think.
If they say I totally get the product, then you can say we have 3 kinds of offerings — this, this and this. Would you like to talk to somebody about which one might be a better fit? If they say yes, then you can connect them to a sales person.
How easy or difficult are these bots to set up? What skills and resources are needed?
It’s pretty simple to set up because it’s all done visually. It’s just like building a flowchart.
And the cool part is once you have set them up, you can go watch people have these conversations. And in the reporting, it says 78% of people clicked on this, 22% on this one. And if 22% is lesser than you expected, then you can reword the question in real-time.
Super interesting. Let’s talk about your free product now.
Drift started out with a freemium offering. I think it’s easier to create a free offering when you start with it rather than being an established company and then introducing a free version. Because then all the people who are already using the paid product are looking at the free version that you’re releasing and asking — can’t I just use free version instead? It becomes a hard thing to navigate. And you have to have the resources to dedicate to building a legitimately useful free offering.
When I was at HubSpot, I worked on one of the sales tools, which was like an email tracking service — when did people open your emails, clicked on your links etc. For people on the free version, we put out a link saying this was powered by HubSpot sales tool.
When you have a free offering, it’s important to think about what channel it gets distributed in. Is there any part of your product that can touch external users? If it’s not something as easy as a product built into a channel like a chat widget or email, are there things you can do to get your users to share something on social media, for example?
It’s harder to do virality when it’s not a part of a channel. For example, Gusto which a payroll service has a referral program. If you invite someone and they start using Gusto in their company, you get $100. But it’s really hard because you have this link which you send to your friend and the friend has to sign their company up using this link. The person receiving it hasn’t gotten any value yet. He has just got a link someone asked him to use for signing up.
For a product like Drift, the end user experiences high value because they used it to talk to a company, so they experienced the entire thing themselves. That makes them far more likely to sign up.
High value experienced by the end user correlates with the likelihood to sign up to create a viral loop.
If you can find ways to fit a distribution loop into a channel, great. If not manufacture ways to fit into a channel, like sharing a report.
To give you another example, in the US, people have these housing alarms, so if someone tries to break into their house, the police gets alerted immediately. House owners put out a sticker in front of their houses saying they are using so-and-so security alarm, so it keeps burglars away. But other house owners also look at it and say — oh I should be using these alarms too. The house is the channel here so to speak.
Similarly if you are a cybersecurity company, you could get your customers to put a badge on their website saying — verified or secured by so and so company. Payment processors already do this, so payments becomes the channel there.
If you offer a free version of a product, is there any change in how your product or brand is perceived?
I think it changes a lot. It makes the product and brand seem far more accessible. It lowers the barrier people feel to sign up for the product. A lot of the people who sign up for the free product could be a really good fit for the paid product, so not only is it a means for people to install the thing and refer other users, but now that they are trying it out, it’s easier to talk to them because they are already semi-invested.
I also think of a free product as a means for people who are a good fit to express intent similar to how they can express intent by clicking contact sales.
You had also mentioned that you have 20–30 different versions of your homepage depending on the type of visitor. Let’s talk about that.
Yes, there are 20–30 different versions of the Drift homepage we have based on… do we know what company you are at, have you been to the website before, what industry you work in, what size of business you are, etc. The CTAs will be different, the formatting of page will be different, the headlines will be different.
What a lot of companies try to do is create one core message that works or they A/B test and have a one-winner-world for everyone. We build our experimentation differently where we have a best version for that type of visitor. We have 20 different versions not only of the homepage, but 20 different versions of all of the top 10 trafficked pages on the website.
This sounds great but also pretty sophisticated. If a company had to start off with this, how would you advice they do that?
We have a great relationship with Intellimize, which is the tool that powers all of that multivariate testing on the website. A tool like that is necessary — building it on your own is not a resourceful endeavor. These guys are world class at the Bayesian statistics behind it and building out these kinds of systems.
Then you need to get a designer and a person who owns the website, the two of them can then cook up dozens of different versions of the website. It’s also important to break the mould of how you would think about making websites today. Today, your process probably looks like this — you have a designer and you say to her, “Hey we need a new version of the homepage.” The designer will design a few different versions and then ask, “Here are 3 different versions I am working on. Which path feels right?” A product manager or marketer or whoever owns the website will say, “This one feels right. What if we could do this with it?” And then you build on top of it. But the whole point if you can go and test those three, you don’t have to go through the process of picking it at all.
Remove the notion that you and your team can intuitively pick the one that will work the best. What could happen with the 3 versions the designer made is that one might work really well for qualified traffic, one for unqualified, another one works for this vertical that you haven’t really tackled yet. Rather than trying so hard to iterate in the design mode, you can iterate with it being live on the website.
And you are not doing it with the risk of losing out on your conversions because it’s not like you are taking half the traffic and putting it in this other version like you would do with A/B testing. You are putting it in there and the system is going to give 1% of the traffic to the new version and if it looks like it is doing good, then it will give it 2% and so on. If it’s not performing, then the system will stop giving that version any traffic at all.
How different are these versions?
We are talking about completely differently designed pages. One version of it where you can’t scroll at all, there is a photo of a person, a big headline and the only CTA is to enter your email address. Another version is a long page that scrolls and has 2 videos and a bunch of product illustrations and all that.
What we have found that works pretty well is that you start out by testing massively different versions — make 3, 4 or 5 completely different versions, test them and find the one that works best. Then you test headline changes, then you test swapping out the person’s face, making the CTA different.
Bigger changes will lead to a version doing 1% and another one doing 12% as opposed to changing a button which can only get you from 3.1% to 3.4%.
Let’s move on to something you guys are famous for — your brand and the conversational marketing category you have created and owned. What does it take to create a category?
It’s funny because we were trying out a bunch of different terms early on and finally found that conversational marketing was just the most self-explanatory and then we just started calling ourselves that. Instead of a marketing automation system, we became a conversational marketing platform. The term is relatively easy to understand, it’s a part of the old category which is marketing and conversational is a pretty straightforward thing to understand.
I think it was a mix of simplicity and us just repeating it again and again. When our executives would get interviewed for articles by publications, we would call ourselves a conversational marketing platform. When we would launch new products, we would put the sub-text as Drift: conversational marketing platform.
We also worked with companies that run these kind of category systems, like Gartner or G2Crowd. On Gartner, Drift got called a conversational marketing cool vendor. G2Crowd added a conversational marketing category. We reached out to them by saying hey we have been building this thing for 2 years and we notice you don’t have a category for it and here are some other companies also in this space.
How did you know conversational marketing was the right term? If you have to repeat a term over and over to create a category, so how do you know what’s the right one to get behind?
I think it’s hard to know. But there are some signals that show you that it is working. One of the things we noticed was new startups calling themselves conversational marketing platform 3–4 months after we started calling ourselves that.
Also, with the other terms, we tried them for a couple of weeks and realized it just wasn’t clicking. I would explain what Drift does to my friends and they would look at me — what the hell are you talking about? But now when I tell people I work at Drift, they are like, oh yeah yeah you are building out the conversational marketing platform right? The moment when other people start using terminology and get behind it and it’s not just you saying it and having to explain it is when you know you have something.
We also had to create content on what is conversational marketing or why conversational marketing is the marketing automation of tomorrow either by ourselves or by partnering with other companies or blogs to create that movement.
Does it take a lot of resources and funding to create a new category?
The resources it takes to create a category is a following and you can create a following with very little resources. It’s less about the resources, more about the coherency. Does your marketing use that message everywhere, does your sales team align with it and carry it through in the conversations that they have, does your support and service team embody that in the interactions they have, does the social media team exude that type of business, does the product deliver on that?
What kind of results have you seen from creating a category?
I see a lot more people coming to us because we are the leaders of the space and we are the first result to show up when they search for conversational marketing. When you end up in the hearts and minds of people because we have built an amazing brand, then people start to pick up on that. You have created this goodwill, you are the first person to be associated with it and it all comes back around full circle. Then, you need to work to deliver the promises you’ve given and delighting your customers on a daily basis
— — —
That’s all from Matt. If you would like to get advice from him on demand gen using new-age methods, virality with freemium offering, product messaging or branding, he is available as an expert on MonkHD and you can get on a call with him at $95/hour. He’s really helpful and incredibly nice to to talk to :) You can know more about him on his LinkedIn.
If you are interested in scheduling a call with him, please get in touch with us here.
