Discover, try, buy, repeat: a radically empowering approach to B2B sales

This article is part of our series on using the flywheel model to grow your business. Check out the complete collection here.

Let’s start with an exercise. Take out your phone and start the timer. Now Google your company’s name and see how long it takes you to sign up for and start using your own product.

Now, reset your timer and get out your credit card and try to buy that same product. How long did that take?

If this whole two-part process only took a few minutes, then there’s really no reason for you to read the rest of this article, because you’re already well on your way to building a business based on the flywheel model. In fact, you’ve mastered two of the most difficult pieces. For the rest of you, keep reading.

Growing a flywheel business starts with making it easy for people to discover, try, and buy your product completely via self-service. This is a large part of what enables Atlassian to spend just 15 percent of our revenue on sales and marketing and 35+ percent on product development. For most software companies, those numbers are flipped. Imagine how much more value you could deliver to your customers if you doubled the size of your R&D teams.

But the reality is, from an organizational perspective, building a seamless self-service customer experience is exceedingly complex.

I’m going to walk you through the entire customer journey so you can see just how elegant it can be on the front end, and how all the sausage gets made on the back end. I’ll be talking about how we do this, which is exactly what I’d advise other B2B companies to do. So as you read our story, imagine what it’d look like for you to take the same approach.

An inside look at the self-service customer journey

Let’s walk through the entire process, from discovery through to dedicated customer.

Discovering the product

Some of this first section will probably sound like “Marketing 101.” Even so, it requires continual maintenance and investment to stay relevant. That’s the hard part.

It’s safe to assume that the first place people start when looking for a solution to a problem is Google. So that’s where the journey starts for both our customers and our business. At Atlassian, ownership of this experience falls on two different disciplines within marketing: performance marketing, i.e., online ads, paid social posts, and search engine optimization (SEO).

The performance marketing team controls a significant portion of our marketing budget and is continuously optimizing our ad campaigns and the keywords we bid on. They’ve even built a machine learning engine to automate parts of this. On the other side, our SEO team works with other marketing teams to make sure our content pleases Google’s algorithms, whether we’re bidding on a keyword or trying to rank organically. All of this leads to our prospective customers finding Atlassian in their search results.

Depending on which result a searcher clicks on, we might give them a tailored experience related to their search term. This is where our website and web personalization teams come in. Product marketers work with user experience (UX) design, web development, and e-commerce teams to create these custom pathways. However, we’ve found that more often than not, it’s most effective to land people on a sign-up page prompting them to try a product related to their search or the social post they clicked on.

Trying the product

Once the customer signs up to try a product, a few other teams spring into action. Our identity team creates a new user ID for the customer. Our provisioning team spins up a dedicated product environment for them. Our email team sends onboarding emails with just the right amount of information to help them get started. (How do we know it’s the right amount? We don’t always. But we test, iterate, and test again. We never consider it “done.”) And before any of that was even activated, our product teams have already built onboarding flows right there inside the product.

We want to track every step of that customer’s signup and onboarding journey, so now the analytics team gets involved. And we need to optimize each step of this journey, which means including our growth-hacking team so we can A/B test each step.

This entire flow has to be seamless for customers. An experience that takes just a few minutes for them is the result of thousands of hours of hard work from dozens of teams across our business. Again, the trick is not to expose all that internal complexity. Our product marketers and designers also create self-serve resources such as demo videos, training content, and use case-specific microsites with the goal of making customers successful in our products without speaking to an Atlassian team member.

To be clear, it wasn’t always this way. In our early days, it was so hard to install and get started with our products that we would send you a T-shirt if you proved you made it all the way through. But that’s where cloud-based SaaS products really have an edge over legacy on-premises installations: you can transfer all that administrative pain from the customer onto your systems and automate the whole journey for them.

Buying the product

Now we’ve got a prospective customer using our product. Job done, right? Wrong. We have to get them to actually pay us for it.

This is where I see a major divergence in how B2B software companies think about their businesses. Many companies will treat this new product sign-up as a lead for their sales organization to chase down. Based on the information they’ve gleaned about the customer so far, sales development reps or account managers will reach out attempting to arrange a meeting and drive the sales process. While this is a tried-and-true approach, it inherently limits the number of sales you can land to the number of sales reps you can hire.

Atlassian takes a starkly different approach. Instead of throwing a sales rep at every prospect, we approach this as an e-commerce experience.

We have a whole matrix of products and editions (Free, Standard, Premium, and Enterprise) that inform what price customers pay. But everything starts off free. Once customers bump up against seat limits (10 users for most of our products) or reach the end of their trial period, they usually pay with a credit card.

Enter our commerce team, which is focused on optimizing and streamlining the complex set of activities around quoting, ordering, billing, and invoicing customers. And the web team again, which builds the front-end purchase flow. The product team is back in action as well, because they built the in-product notifications that prompt users to pay. And of course, the analytics team, so we can track every step along the way.

Adding more users

Now that we’ve landed a customer, how do we expand inside their organization? For Atlassian, this always starts with getting more users to use the product.

Getting products to essentially go viral inside a customer’s walls involves a ton of A/B testing. So here come the growth hacking, product, email, and analytics teams again. They look at things like which in-product and out-of-product messages are most effective at inspiring existing users to invite their colleagues. At what point in the user journey should we surface those messages? How difficult is it for a brand new user to create an account and join their colleagues?

Then we have to be able to satisfy the demand we’ve just generated, but in the most frictionless way possible. With most software, customers have to go through a whole process to add users. They need to contact the company, get a quote, and make a payment, which delays their ability to get value from the product.

But in a flywheel business, “easy for the customer” means their admin can add seats with a few button clicks, without even leaving the product. And because we’ve already got their credit card information on file, we start billing them for the additional users automatically. Of course, many enterprises forbid their employees from purchasing software with a credit card. We have processes for that as well, which I’ll discuss below, but you’ll be surprised how often customers will start with a credit card well before they run into their organization’s procurement policies.

The lesson, in a nutshell, is this: if you really want your product to spread like wildfire inside a company, make it easy to add users.

Adding more products

If customers are actively using one of our products and getting a ton of value from it, there’s a good chance they’ll be interested in additional products or services. This might include add-ons that integrate with the core product to satisfy niche use cases, premium editions of the product they already use, premium support services, premium training services, and, of course, additional products.

Discovering and learning about other offerings has to be as easy and low-touch as it was with their first purchase. SEO, helpful content, free trials, click-to-purchase, rinse, and repeat. If you can embed try-and-buy functionality right into the product like Atlassian does, even better. A Jira Software admin can try Confluence with a few button clicks in the admin panel. If they like Confluence, they just keep using it. Here also, all the billing updates happen behind the scenes.

Troubleshooting

In a perfect world, software would be so intuitively designed that nobody would need help using it. But we live in the real world, and real customers will have questions. Sending each and every question to a customer support rep isn’t an option for a volume-based flywheel business. The economics just don’t work.

Instead, Atlassian provides robust information online. Help docs and knowledge bases cover a lot of ground in terms of figuring out how to perform operations in our products. We also publish “tips n’ tricks” content on our blog.

But with millions of end users across many industries, there’s no way we can anticipate every question. That’s where our user community comes in. What we and many other companies have discovered is that end users are incredibly knowledgeable and love to share that knowledge with other users. So we built the Atlassian Community website as a peer-to-peer Q&A forum. In fact, for customers of our free cloud products, Community is the only place they can go to ask questions that they couldn’t find answers to on our other web properties. (Although, if their question hasn’t been answered in a certain period of time, we’ll invest some time from a customer support engineer in answering it.)

The role of sales reps in a flywheel business

I realize we haven’t covered every nook and cranny along the self-service customer journey, but these are the roads most traveled. If you’ve lost count (and I wouldn’t blame you), it involves at least 15 different teams, each with unique skill sets.

The one team we haven’t talked about, though? The sales team. For many years, Atlassian was known for not having a sales team. While that was true a decade ago, we do have a sales organization now. We just approach it in a very unique way.

One thing I want to make crystal clear is that the sales team is not involved in landing new customers in the context of the flywheel model. We optimize that entire experience for self-service. However, if a new prospect has a question, our incredible customer-advocate and product-advocate teams are ready to assist. Our sales organization is largely focused on assisting our biggest customers to align their complex challenges with the many products we offer. And even then, the sales team only engages once a customer meets a strict threshold of annual spend with Atlassian.

That said, there are downsides to this approach. If we threw sales reps at every single customer, our conversion rates would go up. Some customers would be happier. We might understand our customers better. It would probably help our brand’s reputation. But it would also mean jacking our prices way up, or sacrificing profitability, or both. And in general, we can solve customers’ problems faster through self-serve web-based capabilities than by making them wait to interact with a human.

But self-service doesn’t work in every case. Let’s say a customer needs to pay with a purchase order and a check instead of with a credit card. We have a team to help facilitate that.

Or, let’s say we land a 100,000-employee enterprise as a customer. They’re probably used to getting the white glove treatment and might balk at this whole self-service thing. Sure, we could bend the rules for them and have a sales rep reach out right away. But we don’t. Rule-bending is a slippery slope, and ultimately, a trap. (And worthy of a whole separate article. Stay tuned.)

Atlassian’s solution to this dilemma was to build out a network of sales channel partners. Their role is to provide high-touch sales and service to newer and/or smaller customers who want that experience and are willing to pay a premium for it. This works so well that channel partners now account for a significant percentage of our sales.

Let your values guide the process

Nothing about the self-serve model is magic. It’s blood, sweat, tears, and passionate debate between all the teams involved. Without meaningful values to guide your decisions, it can turn into a big, stinking mess.

For us, building a beautiful self-service experience ties into three company values, all in the service of making the customer successful on their own:

• Be the change you seek.
• Build with heart and balance.
• Don’t #@!% the customer.

We’ll keep challenging ourselves to get that experience a little closer to perfect every day.

The last thing I want to do is discourage you from adopting this approach. But I want you to go into it with your eyes open. Developing a top-notch self-service game requires a lot of teams that do nothing but tweak, optimize, and evolve your self-service go-to-market machine.

As with anything related to your business model, the whole company has to be focused on optimizing the self-service experience. If you can pull it off, it’s a huge win – from a spend perspective, a product development perspective, and, most importantly, from the customer’s perspective.

Hungry for more? Check out the other articles in this series.

Thanks to Sarah Goff-Dupont for her contributions to this post.