Everyone’s building a platform.
Sometime in the last decade or so, we entered an era where it’s no longer enough for tech companies to build one or two industry-leading products. Instead, they need a whole suite of products. They need to guide customers through complex workflows spanning multiple user types and business units. They need to provide cohesive user experiences across entire ecosystems of related products.
In short, they need platforms. Just look at the recent string of blockbuster acquisitions.
Intuit acquired Mailchimp for $12B. Microsoft acquired Nuance for $19.7B. Square acquired Afterpay for $29B.
Each of these companies acquired products that are only tangentially related to their core offerings. But when combined with the acquiring company’s existing products, these new products promise a platform UX that could be far greater than the sum of its parts.
Of course, building this cohesive platform UX is no walk in the park. Year-after-year of M&A, design debt, and misaligned product teams can result in a suite of applications that really don’t work well together.
This is really common among our clients here at DesignMap, and was the exact problem our client ExactTarget came to us to solve. And solve it we did.
Over the course of a 7 year engagement, we helped ExactTarget envision, build, and iterate on a first-of-its-kind platform that fundamentally transformed their industry. We took them through several iterations of our 3-phase design & innovation process, helping them build and improve upon the platform that was acquired for $2.5B by Salesforce.
If you’re tasked with integrating a suite of disparate products into a cohesive platform UX, I want to help. To do that, I’ve summarized the 5 key platform UX lessons we learned during our work with ExactTarget.
Let’s dig in.
In 2009, ExactTarget was at an inflection point.
The marketing industry was in the midst of a big transformation. Digital technology had reached a point where suddenly marketers had to coordinate campaign messaging across a multitude of new channels. TV, radio, and print were still there, but marketers now had to also account for email, social media, digital advertising, SMS, in-app notifications, and more. Creating a cohesive cross-channel customer experience in such an environment was a huge problem, and marketers in virtually every industry were desperate to solve it.
Having just raised a $70M series C, ExactTarget’s leadership knew that shifting from a single-channel email provider to a multi-channel marketing solution was a huge opportunity. But how would they do it? And just how big of an opportunity was it, anyway? The first phase of our work was designed to answer these questions.
We started with UX research. A lot of it.
We conducted interviews with designers, product managers, and other internal stakeholders. We audited their product, looking closely at the navigation, information architecture, UI design, interaction design, animations, and the ui/ux across desktop, ios, and android devices. We analyzed competitors, modeled the market landscape, and mapped the product ecosystem before conducting trend analysis & strategic forecasting.
All of this was extremely insightful, but the real aha! moment didn’t come until we got deep into our user research. The team and I had interviewed no fewer than 50 different marketers. They’d shown us countless spreadsheets and presentations and post-it notes they’d used to coordinate messaging. They’d several times recounted the frustration they’d experienced waiting for their IT team to run SQL queries to segment their audiences.
These were some real user needs, and we were happy to find them; but it wasn’t until we interviewed one of the last marketers that the lightbulb really went off. She worked at a national apparel retailer, and had been trying to send emails with different offers to customers based on key attributes: what they’d purchased, how long they’d been a customer, how much they’d spent, or when their birthday was.
She showed us a corkboard she used to model these journeys out. As we looked at the color-coded index cards branching into different pathways at key milestones, it dawned on us. Could we go beyond simply providing marketers multiple channels within one interface? Could we create a user interface that’d sit on top of these channels and combine them into a single cohesive journey?
Journey Builder was born.
This product was the first of its kind, and came to be the crown jewel of ExactTarget’s platform. It coordinated the functionality of ExactTarget’s core products—Email Studio, Contacts, Behavior Studio, etc—within a simple user interface not unlike the corkboard we’d been shown. Marketers could schedule messages from any of ExactTarget’s channels based on different user behaviors or milestones, sending customers on journeys where virtually all messaging from the company was personalized to their interests.
Rumor has it that Journey Builder was a huge factor behind Marc Benioff's decision to acquire ExactTarget for $2.5B. It was a first-of-its-kind tool that completely transformed how marketers approach multi-channel messaging. It’s come to be the de-facto standard in marketing automation, with everyone from HubSpot to Marketo to Eloqua now having their own version of Journey Builder.
None of that would’ve been possible without the deep deep discovery work we did. So the lesson?
Taking the time to do product discovery right could make you billions.
During our product audit, we noticed that ExactTarget’s navigation and information architecture correlated almost perfectly with their org chart. This might make sense on the surface, but when we conducted user research and usability testing, we found that it didn’t make sense to users. They were getting confused and had tons of difficulty in navigating the product, which resulted in many powerful features going unused. On top of that, users simply had no idea how powerful the platform really was.
This problem is called “Shipping the Org Chart”, and is extremely common among high-growth companies. When you’re rapidly scaling, it’s common to end up with product teams that are siloed and rarely communicate with one another - even if they’re solving similar problems for the same personas.
This is no fault of the product teams. Their OKRs are usually focused on their own roadmaps rather than the overall platform UX. They’re often under immense pressure to ship feature after feature while having little incentive to collaborate with other product teams. Over several years of this, decisions that made perfect sense in isolation end up breaking the overall platform UX. You get low feature adoption. You get bad NPS and CSAT scores. You get users who grow increasingly frustrated with your platform UX until they eventually switch to some easy-to-use, dumbed-down version of your product peddled by a competitor.
So as ExactTarget’s churn and sales close rates got worse and worse, we told them exactly what we tell all our clients today:
Structure your platform by what your users want. Not your org chart.
Once we had ExactTarget’s product vision, we were inspired. Our designers were inspired. Our executive champion was inspired. But if we’d’ve taken the vision straight to product leaders and executives who’d weren’t privy to our process or findings, we’d’ve probably been met with a lot of puzzled reactions.
When we started to outline what sort of transformation this vision would require, those puzzled reactions might’ve turned to angry ones. Here’s why.
Transformational change can be really painful. When you envision a platform UX structured around user needs rather than your org chart, executing it almost always involves reorg.
So not only are product teams resistant and sometimes incensed that the products or features they poured their blood, sweat, and tears into might be completely changing shape or going away, but their entire teams might cease to exist. As you can imagine, this involves some very uncomfortable changes.
Leadership gets reorged. New employees get hired. Existing employees get reassigned (or laid off). So when you couple this with all the resources a platform transformation requires, you can see that getting people bought-in is crucial. I’d even go further to say you need people to be excited.
But when your product vision is just a bunch of aspirations and ephemeral words—when people can’t actually see it—it’s virtually impossible to get them bought in and excited. And without that excitement, your platform vision is dead in the water.
This is where Visiontypes come in.
Visiontype is a term coined by Silicon Valley Product Group’s Marty Cagan. Basically, a Visiontype is a high-fidelity prototype of what your platform vision could look like in 3-5 years. The idea is not so much to provide a prototype of UI-ready designs that you can turn around and build right away, but instead to help people envision how your platform can change their customers’ lives.
It’s our experience that once designers or executive champions have the Visiontype in-hand, and people are able to really see what the platform could look like, that’s when the lightbulbs go off.
With ExactTarget, the Visiontype was instrumental in getting everyone aligned behind the new vision. The internal teams and executives were so impressed, they decided to debut in front of thousands of marketers at ExactTarget’s annual Connections conference. When the crowd erupted in cheers after the debut, we knew alignment wouldn’t be an issue.
Once everyone was aligned behind ExactTarget’s vision, our first step towards that cohesive platform UX was to build a living design system.
A living design system is a crucial part of a cohesive platform UX. The term “living” refers to the fact that these design systems are coded into front-end component libraries which allow aspects of design to be easily changed and automatically applied throughout the platform.
Not only does this make it much easier to maintain a consistent visual design and interaction design throughout the platform, but it also significantly reduces design debt. Any time a component needs to be updated, simply making a change to the component within the centralized library will automatically apply the change throughout the platform. This keeps upkeep and development overhead low, while also allowing designers to constantly iterate on the UI components via A/B and user testing.
The process of creating a design system depends heavily on the individual company, but generally goes something like this:
Here, we’re taking inventory of everything that’s been designed so far. What works, what doesn’t work, and why. We’ll develop a list of all existing components, and conduct research into the function of these.
In this step, we meet with key stakeholders about developing the aesthetic basis of their system and discuss branding. Prototypes and design component mockups are created, and we begin to develop a style guide.
You’re going to need a timeline. Both for designers and engineers - who will be working simultaneously to produce design artifacts.
Here’s where those design elements, components, and artifacts begin production. This is often a highly iterative process, with each element going through phases of testing and feedback.
Finally, as each of these sprints falls into place, the living design system is ready to be released to users. Having this system in place is also much more conducive to an iterative design process. The style guides created earlier are easily adapted for new versions of the product.
ExactTarget’s design system followed a similar progression. We’d done much of our design discovery at the same time as our product Visiontype creation, and enjoyed enthusiastic buy-in, so we were able to hit design sprints right out of the gate.
First, we had a series of meetings in which we established the aesthetic principles and branding that would guide our design. We took a thorough inventory of the current design, and met with stakeholders throughout the organization to gather feedback.
This can be a particularly challenging step. People have a lot of feelings about aesthetics, and opinions and pride tend to become wrapped up in the design. Setting those base principles and branding from the start helps mitigate this - but it’s never going to be easy.
With a comprehensive view of the current design system, and a solid direction on where everyone wanted to take things, we set to work on developing the design system itself - the colors, spacing, fonts, sizes, margins, tokens, logos, icons, animations, patterns, images, functions, and countless other artifacts of platform UI design.
While both our design team and ExactTarget’s were wrapped up in all this work, the engineers had a separate, yet complimentary, timeline dedicated to building out and applying the new design work.
With a living design system in place, we had what we needed to bring all of ExactTarget’s products and features together into a cohesive user experience. But we needed to think carefully about how to roll it out, which brings me to my last point...
Product teams always face a challenge with determining how to release an updated product. It basically boils down to the choice between gradual releases, or all at once.
Do you fix the entire thing and create a product that is essentially new, then release it all at once? This approach is satisfying, surely, but can also be unpleasant or confusing for the user. One day, they’re using their old familiar tools, and the next, they can’t find where anything is. The “all at once” method also means you have to wait until everything is ready before you release - during which time, any number of changes could derail the progress.
The other option is to work your way through the platform, fixing the most popular and important tools, and re-working the platform in a series of gradual releases. This gets design work out the door and working for your users much faster, but it can admittedly be less “sexy.”
ExactTarget opted for the “gradual” option, and we set to work making changes immediately.
We started by hacking together some of the platform’s existing functionality into what was essentially an MVP for Journey Builder. Called Playbooks (pre-made customer journey templates), we collaborated with the engineering team to scrap this together using their existing codebase and design system components. The idea here was to give users a small taste of the platform vision to get them to start to see the possibilities we’d be bringing to them over ExactTarget’s evolution.
The user data we got from Playbooks not only validated our vision (marketers loved it), but gave us extremely valuable insights that informed future designs and rollouts. We iterated our way through this process step-by-step, gradually releasing more and more product and feature updates while engineering updated the back-end.
Within about 4 years, ExactTarget’s platform vision had come to fruition, and marketers loved it. We saw product engagement and customer satisfaction metrics improve dramatically, and on more than one occasion heard stories of marketers who’d seen sales increase by hundreds of millions of dollars after adding Journey Builder and its multi-channel capabilities to their arsenal.
Shortly after, ExactTarget was acquired by Salesforce to be the crown jewel in the Salesforce Marketing Cloud. The Journey Builder experience has since been applied to Salesforce’s Sales Cloud and Service Cloud, allowing companies to create unified journeys throughout the entire customer lifecycle.
And as we already mentioned, Journey Builder redefined the standard of the multi-billion dollar MarTech industry.
I hope this story’s inspired you. As you continue on your journey towards your next-gen platform, be sure to keep these lessons in mind, as they’ve proven crucial to our ability to design cohesive platform experiences.
On behalf of the whole crew here at DesignMap, best of luck on your journey!