Beyond Land: Atlassian's Guide to Customer Expansion - A Conversation with David Yuan and Atlassian’s Cameron Deatsch

“Land and expand” has rapidly infiltrated tech investor parlance. It’s a catchy slogan that implies once a vendor makes an initial sale, expanding that customer account is predictable and highly profitable. Indeed, customer expansion can drive a high revenue renewal rate, which is one of the single biggest drivers of growth and profitability for a recurring revenue business.

This is easier said than done, however. Making the transition from a single product to a multi-product company can be one of the toughest challenges of scaling. Few companies can compete effectively in multiple end markets and execute a go-to-market model that successfully sells multiple products.

Atlassian’s approach to customer expansion is unique, and they do it really well. Cameron Deatsch ran Growth and now the Enterprise business at Atlassian, and he shared a master class on all things customer expansion at TCV’s 2017 Customer Acquisition Forum. Topics covered included:

• Product portfolio
• Three paths to expansion
• Land vs. expand products
• Building a testing culture
• Onboarding
• Personalization and segmentation
• Naming
• Pricing

***

DAVE YUAN: Cameron, most tech executives have heard of Atlassian, but can you ground us in some basic context of the business?

CAMERON DEATSCH: Atlassian was launched in 2002 in Australia. Our mission is focused on “unleashing the potential of every team.” We believe that great accomplishments require great teams, and we focus on helping teams work remarkably better together. We went public in 2015 and now have 90,000 commercial customers and a broad portfolio of products including Jira, Trello, Confluence, Stride, Bitbucket, and more.

When we think about our customers, we don’t focus purely on the large Enterprise and biggest spenders. Instead, we think about the 1,000’s of teams that make up those companies. Atlassian is all about enabling them as distinct teams. This forces us to think of companies as a combination of multiple customers while providing us with many different entry points into those large businesses.

All told, there’s a huge market out there for our technology.

DAVE YUAN: A lot of folks are obsessed with customer acquisition. Why is customer expansion so important? Explain the difference.

CAMERON DEATSCH: First, I will say that customer acquisition is important and more important than expansion. If you don’t secure a customer in the first place there is nothing to expand. And everyone knows that acquiring customers is really difficult. If you’re lucky, you build a great service that attracts customers purely through word of mouth and owned channels like your blog and social media connections. However, as you grow this only gets more difficult and more expensive.

Expansion, on the other hand, is the key to long-term financial growth for a company. Keeping and expanding a relationship with a customer is much easier than acquiring a new one. Thus, at Atlassian, we attract new customers through compelling offerings and either free tiers or low-cost entry packages. Once they’re successful, we build paths for them to grow either through user growth, new products, or premium versions of the products they already own.

DAVE YUAN: A broad product portfolio is core to your current strategy, but have you always had a multi-product approach?

CAMERON DEATSCH: I realize that when you’re an early start-up with a very small customer base and low monetization, telling your product teams to go build a second product might sound like sacrilege, right? But that’s what we did.

Our co-founders, Mike and Scott, did this back in Sydney two years after starting. In 2004, we launched our second product—Confluence—and every year or two afterward Atlassian has brought a new product to market. It was probably a little more organic at the time based on customer needs, but it is now a core strategy and it has since worked very, very well.

DAVE YUAN: It does run pretty counter to conventional wisdom of “Focus, Focus, Focus.” How did you avoid getting spread too thin?

CAMERON DEATSCH: First, you have to figure out whom you’re building products for. For starters, all of our products are built for teams. Across all products, we’re trying to make teams more productive and more collaborative, and we also have common capabilities, features, and experiences across them. Whether it’s Bitbucket for code collaboration, Jira Software for agile project management, or Confluence for documentation, the common theme is to help your teams organize, discuss, and complete work. And since teamwork is complex and different for different teams, we need multiple products to address this inherent breadth and complexity.

DAVE YUAN: Ok, let’s jump into how you put this product portfolio to work to drive customer expansion.

CAMERON DEATSCH: My first piece is, build your paths towards expansion. Literally write down where and how you are going to grow a customer. If you have multiple products today, the best option is to look backwards at the customers that have actually grown really large and see the path that they took over time. Was it because they added a bunch of users and bought a premium offering? Or did they buy multiple products up front? Write it all down and analyze those customers. I guarantee you’ll see trends. If you’re just introducing a new product, build a path but don’t fall in love with it. Customers will go where they want to go and you need to adapt to that fact.

We really think about expansion in three basic ways:

1. Upgrades: A customer is adding more users to your product. Fundamentally, that’s a customer success project and a product experience challenge. As a marketer, there’s very little that I do to actually get customers to add more users. We can do some experimentation with different offerings, but in general, banking on “we’re going to upgrade tiers” is quite difficult in an organic model.  
2. Upsell: This basically means getting customers to upgrade from a base product to our premium offerings. This is something that I spent a lot of my last year on. Different price points, different capabilities, different conversion points. It’s still the same product, just a lot more capability in the premium offering. That’s a major part of the Atlassian strategy today, specifically for our enterprise customers, and I could spend all day talking about it.
3. Product Expansion: This means we add another product into our portfolio. It’s adjacent and it might be the same buyer or a different buyer. Whatever it is, it’s something different. How do we make sure we’re getting our customers to try it out and hopefully, purchase?

DAVE YUAN: Absolutely, how do you manage cross-sell across the portfolio?

CAMERON DEATSCH: My number one rule here is to distinguish your land products from your expand products. Your go-to-market strategies for each will be fundamentally different. This is something that took us a long time to discover. We had all these individual product teams, and each individual team wanted to grow as quickly as possible. Each wanted to email all the customers all the time. Each wanted all the advertising budget and all the marketing channels. Everyone was asking for everything all the time on all the channels. Parsing out land products and expand products can really help prioritize those demands.

Making the “Land vs. Expand” distinction also helps align channel strategies. Land is the world of paid media, PR, dedicated sites, etc. These channels drive awareness. At the tradeshow booth, land represents the big product logos that we have: “Come talk to us, if you’re not an Atlassian customer, these are the products that we’re going to talk to you about.” These are the products we built brand recognition around. These are the products that we apply our advertising budget to.

With expand products we focus more on our owned channels and existing customer touch points. Our customers come to us all the time. We have a variety of customer touch points including our community, user groups, newsletters, support, and most importantly our products. Yes, we run marketing campaigns into our customer base. Don’t get me wrong, we do a lot of traditional cross-sell tactics, but some of the best tactics are the least conventional.

DAVE YUAN: You mention traditional cross-sell tactics. What conventional cross-sell techniques should marketers avoid?

CAMERON DEATSCH: My first recommendation when you’re thinking about a cross-sell strategy is to take email completely off the table. I’m not saying you shouldn’t use email but it can be a crutch. We become way too dependent on email and it’s easy to abuse your customer list. Build your entire strategy and communications plan and address email last.

My second big recommendation is test all the things!

DAVE YUAN: How do you build testing capabilities and a culture that supports them?

CAMERON DEATSCH: First, start with the team. If I were going to build a team from scratch, I’d start with a kick ass product manager, some devs, a designer (definitely get a designer), and an analytics person or multiple analytics people. By the way, the number one thing we missed when we got started was trying to turn devs into analysts. Trust me, that doesn’t work out so well.

And even before you test, invest in your tools. This is not something you will build overnight, but there’s a ton of great off-the-shelf software now for A/B testing. We can run A/B tests on just about anything: landing pages, signup flows, trial length, in-product messaging, you name it. Of course, doing all this at the same time and coming to a strong conclusion on a test is difficult.

This is why analytics are so important. Honestly, if your analytics aren’t ready, stay the hell away from A/B testing. All you’re going to do is add more confusion to the mess. I was not an analytics expert by any means, but it’s so important that you get all of your pipeline analytics correct. That’s where I would start. Reality is—A/B testing or not—adding good analytics to your funnels is just good business practice.

Realize testing is a tool, not a strategy. You can’t be like, oh we need to do a bunch of expansion and get lots of growth with multiple products so let’s test our way into it. It is not going to work. You need a strategy behind it. Testing is nothing more than a tactic.

My last piece of advice is to move fast. The majority of tests are going to fail. The reality is that you spend all this time building the world’s best test, and it likely isn’t going to work. All of a sudden, you’re only getting one success every two years, and it’s a complete waste of time. Move fast and empower, enable, and protect your team. And yes, break things. If you’re not moving fast and getting multiple experiments out in a month or a quarter, it’s a completely failed exercise. If you're doing one test a month or a quarter, you’re not running tests anymore. You’re rolling out features.

This fundamentally requires executive backing. When you’re routinely wrong, everyone can get mad in the wrong environment. You might have someone say: “I told you that’s not going to work, and you ran that test anyways,” or “You broke my experience.” Or you might have a designer get mad at you because he or she thinks they built a beautiful experience and you went in and messed it up with A/B testing. You have to have trust and a culture that supports testing. Our CEO, Mike, really drove the testing culture in the beginning providing both direction and air cover for the growth teams. Basically, when a test broke or provided a horrible customer experience, Mike would have our backs. If we didn’t have that executive backing from the beginning, we would have been shut down, never ever to test again.

DAVE YUAN: What else? What other ideas are helpful with product expansion?

CAMERON DEATSCH: Onboarding is a huge opportunity. We’re constantly iterating to see what the optimal upfront packaging is. We have this thing called the app switcher in the top left on our product. When someone first starts using our products it actually pops up and shows the different products we have. The user can make it go away really quickly, but at least it shows them, “Hey, there’s something else there.”

Product notifications are critical. Most companies have these as part of their onboarding flows. Details such as what type of text you use, what type of buttons you use, what’s your call to action, those are all opportunities to introduce other products. As an example, we created a product notification where someone logged into Jira, and the company already had Confluence integrated with Jira, but the person had never logged into Confluence. It would say, “By the way you’ve got a bunch of people working in Confluence, come join your team.”

What you name things in product is incredibly important. That is the cool thing about A/B testing: It takes multiple iterations to find the right names and call to actions. But, a 5% improvement in front of a million users ends up being a whole lot of incremental users, and that’s what these little tests will do. It doesn’t cost you anything except for the development and the time. There are no ads to buy, no email to send, and you can run this by a small subset of users, get a statistically significant result and then release it to your entire customer base.

Personalization: We use personalization to drive individuals to events. If we have an event in NYC, and if someone in the NY region comes to the website, we can actually say, “Hey you’re in New York, come to our financial services event” for just those users. That simple tactic made our click throughs go up 30 percent.  

If you really want to get sophisticated with testing, end-to-end journeys are by far the most successful. Essentially, it’s building a custom ad to a custom landing page message to a custom onboarding experience. This is not an easy exercise. These are the things that take a little bit longer and we call these the mega experiments. These can take six to eight weeks to build, run and test, but these are the ones where we’ll see a 10% increase in activation if done correctly. Don’t underestimate how difficult it is to stitch that many pieces of technology all together.

DAVE YUAN: I did a talk with the Evernote CMO on pricing, and he had a great point around optimizing not just creative but the “moment.” Does that apply to cross-sell as well?

CAMERON DEATSCH: Absolutely. Timing is critical. We found that when a product administrator is doing any sort of installation and configuration management, this is an optimal time to provide expansion paths. They are already in the game of, “I am dealing with this product, I am trying to make it work for my organization, what else can I do to make this work better?” Some people just go to Google, search for “JIRA DOWNLOAD,” find the file and that’s it. They don’t even come near our website. How do you cross-sell to these people? Well, we added parts to the actual flow of the installation.

The lesson is, think about the place where people are actually working on your products and doing something other than just using it for the day-to-day. For example, opportune times are when they are renewing, upgrading, or configuring new features or versions.

The counter to this is nobody can actually set up multiple products in an intelligent way when they land on the site as a new customer. It’s just too complicated. Make sure they’ve landed before you even think about expand. More products are not the key to success when you are onboarding new customers. We make sure customers have been successful, renewed a few times, and then we wait a few months before we provide them paths to our other offerings.

DAVE YUAN: Speaking of pricing, how do you think about portfolio pricing?

CAMERON DEATSCH: From a pricing perspective, your land product sets the bar. Atlassian competes in different markets, and we strive to be highly competitive in the markets we serve. But that really doesn’t matter. All the customer cares about is how much they spent on your land product, and they will reference all future purchases based on that original transaction.

Remember, you are selling technology solutions, not fast food. “Give me fries with that” doesn’t have a place in enterprise software. Simply adding products to sweeten a land deal is a poor strategy that ruins your margins and ends up with a bunch of shelfware for your customers.

DAVE YUAN: Well, Cameron, thanks for an incredibly rich talk on expansion.

CAMERON DEATSCH: Happy to do it! Thanks Dave.

***

The views and opinions expressed in the transcript above are those of the speakers and do not necessarily reflect those of TCMI, Inc. or its affiliates (“TCV”). This transcript is not an offer to sell or the solicitation of an offer to purchase an interest in any private fund managed or sponsored by TCV or any of the securities of any company discussed. The companies discussed above are not TCV portfolio companies and are not necessarily representative of any TCV investments. For additional important disclaimers regarding this document, please see “Informational Purposes Only” in the Terms of Use for TCV’s website, available at https://www.tcv.com/terms-of-use/.