Overview: Improve your website relaunch by delivering many small releases instead of one big release.
About the Speaker: As the director of brand and digital experience. Sarah Fruy leads the strategy, goals, and road map for Pantheon’s public-facing website, branded assets, and experimentation program. Fruy is a ScrumMaster® and Certified Agile Marketer. Who joins Pantheon with over 15 years of experience in the marketing, digital publishing, and online advertising industries. Along with marketing strategy and digital marketing certifications from Cornell Johnson Graduate School of Management.
Most recently the VP of Marketing at Glasshouse (acquired by AAA). She lead branding, sales strategy, PR, content marketing, social media, event marketing, and design for the company. As VP of Marketing for AdsNative (acquired by Walmart). She was responsible for launching in the brand. Managing the marketing department, and driving sales growth for the company.
Hello, and thank you for joining me today to discuss how to kill the website relaunch. So you can start iterating on your digital experience. I’m Sarah Fruy, Director of brand and digital experience at Pantheon though webops platform built for agility. I currently oversee our brand as well as a website and our growth marketing initiatives to improve our user experience. And drive conversions for the Pantheon business. I am a certified Scrum Master and agile marketing practitioner.
And I’ve also worked for heritage brands like Hearst and the San Francisco Chronicle and SFgate.com along with emerging media companies, like say media to launch or relaunch digital properties and help monetize those websites for our customers. And in addition to that, I am a frequent contributor on many marketing sites as a guest blogger and I write a lot on the Pantheon log. So feel free to follow me on Twitter, if you’re if I peaked your interest, and you want to hear more about some of these things that I’m talking about, and you can also connect with me on LinkedIn, I’d love to get any feedback you have on this presentation.
And so let’s dig in the agenda for today we’re going to discuss problems with relaunching your website, how to build a web ops organization, how to implement web apps, how we’ve done it at Pantheon. Look at this from the agency perspective, and then you know, give you some pointers on how you can do this for yourself.
Today, site owners serious challenges. Keeping a website fresh and its functionality, visual design and substance is expensive. It’s also time consuming, complex and often painful process, so people tend to avoid it. Most organizations delay substantial changes until it’s time for a redesign or a replatforming project, which happens at most once a year. However, websites that drive results for businesses don’t come from giant relaunch projects. Big Bang approaches make it impossible for websites to keep the pace the demands of the market. It’s time to get serious about iterative web design to boost customer engagement, because everything leads to your website. If you are spending budget on display, social, SEM or any of these channels, I can pretty much assume that your main calls to action drive people to your website. Do you aren’t investing resources in optimizing the experience for your customers, you’re failing them.
There are two main schools of thought when it comes to project management and marketing, agile and waterfall. Conceptually, you can think about agile as an approach or philosophy that is designed for context where it’s hard to predict the future and address that by taking an iterative approach to validated learning, meaning that we do something small and MVP, ideally as small as possible while still validating direction. And if we get validation we do more. If we don’t we try something else.
Waterfall on the other hand is the measure twice cut once approach, practitioners try to look at historical data to predict what’s going to work when they launched four months later. And it fails more often than not and rapidly changing markets because most annual marketing plans are out of date in less than a quarter. The result is a few high risk Big Bang releases. As you can see the biggest obstacle preventing marketing teams from being more agile is legacy workflows.
Projects take weeks of planning, they can quickly lose scope. And your team is always on call in case something goes wrong. But in our experience, most website teams really struggling to be agile. Typically we see a pattern that we call water fail. All the budget is pinned to one big thing website relaunch project. After you get the website out the door, the list of things that needs to be fixed keeps growing longer and longer. The marketing plan stopped making sense months ago, stakeholders start complaining, morale sinks. Ultimately people get fired over this. The average tenure of a marketing leader is 18 months and dropping fast. Marketers increasingly have numbers next to their name just like sales leaders and they have to deliver with them. short period of time, waterfail can get them in their teams into serious trouble.
So, how can you pivot from your current workflow to start advancing your team’s agility. Adopting agile means transitioning from a world of a few big releases to one in which you make many small releases. By quickly delivering many small releases, you fundamentally reduce the risk of any individual release. The result here is many smaller low risk releases that validate direction more quickly delivering small wins.
Today, digital experiences that win result from an agile process. In fact, agile teams are twice as likely to succeed. And they do so by having measured goals.
Because relaunching doesn’t get the goods, going beyond all the pain and risk, it introduces a ton of change at once. It means you don’t know what in particular works. And the chance of it of having a perfect launch is effectively zero. You have to keep iterating and optimizing. And this is done through web ops. At a high level, web ops streamlines how work gets done through agile tools and processes. It is a set of practices that brings together developers, designers, marketers, content editors and more. With web ops culture, everyone whose job it is to change a website in significant ways, be it through code, visuals content, thinks of themselves as one cross functional team iterating towards the same goals.
Now I’d like to show you how we leverage web ops at Pantheon. Before coming to Pantheon, I’ve been in the year long enterprise website relaunch projects, as well as smaller startup teams with tight deadlines and no budget. On the enterprise side, you have access to large budgets and endless resources. But all that comes with major red decisions take too long to make great ideas get diluted. And projects quickly lose scope and spiral into a seemingly endless cycle of revisions. At smaller companies, I would go through a different cycle where I’d hire an agency to update my website over the next three months or so and accumulate marketing debt.
But simply set work when we relaunch no longer fit or go to mark or evolve and go to market strategy. So I would beg our engineers to make updates, and they would push back saying that they needed to focus on building our product, I’d push back saying that if they wanted to sell their product, we needed to update the website. Eventually I go to the CEO, ask for budget to hire the agency back again, and we rinse and repeat the process. I’ve learned a lot from these experiences, and I’m able to manage my website and team differently at Pantheon. Thinking back to the core elements of web ops as being people tools and process. I first like to introduce my team.
Here’s what my core web ops team looks like. We have our designer Stacy, Stephanie and Eduardo who partner with our developers, Nikita, Jeremy and Rachel. For any given project, there will be at least one other cross functional team member working with us on the initiative, most often for the product marketing or demand Gen team that Pantheon.
Next I’ll show you some of the tools we use to be successful. We measure user behavior with analytics tools like crazyegg, Optimizely, and Google Analytics. We manage projects with Asana, and designers discussion, Adobe, and build those experiences on Pantheon or Drupal, CMS. Finally, we track customer data communication and revenue through Marketo, and Salesforce.
So let’s take a look at our website real quick. I can show you what I mean by iterative design. On the left is an example of the website that I inherited. When I first joined Pantheon. I felt like it was really dark, there was a lot of copy, the navigation was very overwhelming. There’s tons of information in there. And I had also been asked by the senior leadership team to do some rebranding work in the process. What we came up with is the homepage on the right. You can see there’s like a new look and feel we tried to add more whitespace to it, to modernize the imagery and things like that.
Over the past two years, I’ve continued to iterate on that experience. So again, on the left here that you can see. We changed our hero image played with the messaging and imagery further down on the page. And then earlier this year, I did some more research, where we interviewed a bunch of folks that would have been ideal prospects for our contract side of the business. And some of the feedback that we had gotten was that second slice with the logos. Some people mistook that for the footers, they weren’t scrolling past it, which we validated by looking at things like crazyegg and Click behavior on the homepage.
And so we decided to test out adding a slider bar, which is what you see on the right, like global access for all customers that would move and there were three different value props that we would present there. We kind of updated the design for the data points below that that was something that was really popular with our users. And I’ll get into more later about the results of that. But we did test it and are continuing to iterate on this experience.
So let’s put this into context of your web. items in the backlog will prioritize work against what we’ll refer to as growth lovers, growth lovers, or changes whose effectiveness can be measured with statistical significance. For example, how does presenting two CTAs next to each other compared to presenting a single CTA growth levers are tied to a Northstar metric that you care about such as the overall conversion rate for your site, but small change to growth lever won’t result in statistically significant impact on that Northstar metric.
You do know, however, that growth lever winds are moving your Northstar metric. It’s like looking at the hand of an hour clock, you won’t see a move. But if that second hand is moving, you know the hour hand is moving. So the key is to do a lot of iteration on growth lovers where you can measure the impact to move the overarching metric set another way, lots of small changes result in big change.
That as I mentioned, our homepage is one area of the website where I’m constantly experimenting with growth lovers. First, my team wanted to test the hypothesis that adding a second CTA. Or call to action to our homepage hero would increase the conversion rates for the page. First test off 53.5% improvement on our demo CTA and a 10.8% improvement on our free trial CTA. Then, we decided to test our homepage tagline. As a company we’ve made a shift last year in our go to market strategy. And I wanted to test our new tagline before fully committing to it. We ran sem social media and display campaigns in addition to a test on our homepage. Well, the gains weren’t as great the test was successful and that the change or conversion rates were minimal.
Next, we have tested personalization or tagline it’s on increase of 3.3%. In our demo, CTA and 1.5% improvement on our free trial. CTA personalization is tricky, but I’ll leave that topic for another discussion. I also want to point out here that on the last page. There was a big one, this is a small one, you are something sometimes going to nail it and get a home run. And other times it’s going to be a really small impact and you’re going to fail and you should expect too faithful. The important thing about failing is that you learn from your failures. And you figure out a better way to do things next time. I believe in a culture of positive failure.
It’s easy for me to sit here in a webinar and talk about all of the great wins that I’m having. But trust me, there are been plenty of losses along the way. And I think the best left life lesson I’ve gotten from it is that will insulate instincts are great and guiding us in terms of coming up with new experimentation ideas. At the end of the day, the data is going to tell you the truth. And it really takes your ego out of it. I love learning I love being proved right or wrong. It’s exciting. It’s surprising. Sometimes things that you think you’re going to be slam dunk with your audience just don’t work. And so expect surprises. But the important thing is that you keep doing it and the more frequently you do it, eventually you’re going to get an outsized impact,
the last example that I showed you another company that’s had or actually before we get into a case. So yeah, I’d like to kind of go back to the homepage experiment that I mentioned earlier in my presentation. So I’m on the V zero version, we had done some user research, again, discovering that people were not realizing that that slice was more content down the page, but we’re thinking of it more as like a footer, and we’rent continuing to scroll. And so we tried a black version in the V one and a white version in v2 of this slider bar. And what happened here was that the results were inconclusive. So you have losses, you have wins, and you can have inconclusive results.
There were things that we liked about the results, and some we had a deeper scroll rate on one version. We have more click throughs on another some of our KPIs like the demo and the free trial offers. One did better in black and the other did better in white. And so there was no way for my team to sit there and say like, Hey, this is the best way for us to move forward. And so rather than hard coding next To the website, we actually decided to, to not implement the slider bar. Even though I do think it’s a beautiful design, I like it.
But we did move forward with the data points below. Like that was something that tested positively in. In our experiment, and then also just qualitatively in the user interviews that we did. And another bit of information that was really great that we applied was that people wanted to click on those data points to learn more, the original version in yellow v zero, was not clickable. So that was another update that I was able to make. This is where that feedback loop of talking to your customers measuring their behavior is really going to make your user experience better for your site.
If you come back to my website next week, you’re going to see other experiments that I’m running. We have a new homepage hero image that we’re trying out. We got some interesting feedback about that from our research. If you’re paying really close attention, you might have noticed that the labels in our navigation have changed in this example. So we’re playing with some things there. And we’ll see how this these two different experiments work and might be making some more changes to the website in the near future based on those results. And we’ll continue to iterate on that to make it better.
In addition to the look and feel of the website, my team, what is this this is an eyesore. Right now, this is a lucid chart. I don’t expect you to be able to read all the little details. But this just shows you how much effort goes into different types of experiments and, and experiences on our website. So when you land on Pantheon, we have chat functionality, we use intercom for that. And we’ve recently this year started building out a chatbot program. with our partners at ChatFunnels, they’ve been helping us build these out. Now we’ve gotten to a point where we have them on a bunch of landing pages, and we’re starting to experiment with them.
So we’re looking at different conversation flows to say like how can we improve that automated conversation experience. To not only help our visitors get the information they’re looking for. But also to make sure that when sales. Decides to engage with them during chat. That these are the right conversations for them to be having. We’re not wasting their time. And then we can also see the impact of these conversations on our bottom line.
We’re looking to capture emails, we’re looking to book meetings, or push people to a free trial if they’re not the right candidates for our higher level offerings and contract plans. And so there’s a lot of ways that you can test an experiment on your website. It’s not just through visuals or copy or where your buttons are, but also through how you communicate. the sky’s the limit in terms of what you want to optimize for in the digital space these days.
Now, let’s take a look at one of our customers. I’ve talked a lot about Pantheon here. Tableau is one of my favorite stories to tell. They’re an amazing example of the power of agile marketing to drive a business result. They came to Pantheon, at 250 million in revenue. Which is an amazing feat in and of itself. But unfortunately, their marketing team was bottlenecked on it waiting to deploy changes to their site. They were only able to make changes on a weekly or monthly basis.
And within a month, we had them up and running on Pantheon and 10 X the rate of innovation. This actually enabled them to scale as a business from 250 million to $1 billion in revenue in just three and a half years. With 95% of the revenue sources coming from their website, when they do 50,000 leads a month across 7000 landing pages.
So constant iteration on their website enables this marketing team to function like a best in class product organization. It’s really, really impressive. And so for those of you who work at agencies, you might be sitting here and saying like, what does this mean for me, you’re gonna take business away like my bread and butter is these big relaunches. And I would argue that actually this is good for business. especially now in today’s climate, budgets are being cut. Projects are shifting.
There’s a lot of uncertainty. But there’s also a knowledge that we need to be proving our website and proving your digital experience because more people are online than ever before. Looking at people’s companies that can’t go to brick and mortar anymore. sales people aren’t going into in person calls. And so the way we do business and how we communicate is changing and it’s really, truly focused around your website. So there’s going to be work out there I promise you it’s just the form of that might need to change to accommodate the needs of businesses who are dealing with a lot of uncertainty right now.
For me personally, I’ve worked with agencies in the past and I continue to do so today. But how I work with them has changed. So rather than hiring them for these lengthy expensive relaunching projects. I partner with them on more strategic initiatives like improving our user journey on the website. So that visitors can find what they need before abandoning our site. Or up leveling our performance scores so that pages load faster and our user experience and SEO rankings improve.
Agency should also shift their goals from leading these big relaunch projects. Which will still happen by the way. To establishing long term relationships with your clients where your team becomes embedded with others. You want to become their go to resource for experimentation ideas. As well as the team to implement them on the website. These experiments will help prove to the budget holders that your work is adding value to the business. And now a small contract quickly transforms into a lucrative ongoing partnership.
So as I mentioned, relaunch projects aren’t going to disappear completely. There are very valid use cases for this to invest in this kind of project, like your company is going through a rebrand or you’ve outgrown your current infrastructure and need to replatform to a more sophisticated setup. Or for those of you who’ve been neglecting your website, you may be drowning in technical debt. In this last case, I would suggest that you fix that first, you can rescan a bunch of landing pages. But if you still can’t update them, you won’t be any better off in a few months time.
If you do need to relaunch my advice here would be to spend no more than 50% of your budget upfront. Plan your project in an agile fashion so that it will get you an MVP, you can feel comfortable launching, then see how people respond to it and make a plan to continue to work based on the feedback you gather. I guarantee you, there’ll be many more updates to make and you’ll need that budget later. So thank you for joining me today. I really appreciate your time. And I’d love to have any questions. If you want to reach out to me out I’ll try and get back to you as quickly as possible. So have a great rest your day. Thank you.