How to redesign your blog and engage your readers
Senior Content Manager
May 11, 2017
What turns a good mobile marketing blog into a great one?
It's a question we've been asking for months throughout the design of what you now see, and we've faced a lot of roadblocks to get here. Finding (and then implementing) the right CMS, agreeing on the design, and importing our entire catalog of posts with all the technical issues attached was tough - not even taking into account the organizational challenges to contend with. The one in the middle was Esa Hernandez-McGavin, the Project Manager for Web at Adjust, and we spoke to her about the new changes, and what went into the design of the next stage of our blog.
James: How did the decision to update the blog come about?
Esa: Well, it was something we were thinking about for a long time.
Our web team exists as part of a cross-functional growth team that includes performance, email and content marketers, all of whom have a stake in the health of our website. For all of us, it’s important that our site is kept active and healthy, and for us to keep potential and ongoing customers interested in Adjust. We knew that we were underutilizing our blog, so when the infrastructure that would support a big change like this fell into place we were quick to snap up the opportunity.
As Project Manager for Web, it’s my job to focus on the continuous improvement of our web properties, and so I was part of this project from the start. The other person who had a big hand in its design and creation was George, the Marketing Web Developer, who builds what users see and interact with on our marketing websites.
James: How did you start the redesign process?
Esa: In our team, all things start with data. Data – and oftentimes even a distinct lack of data – is what defines the structure and prioritization of the projects I work on. When the project kicked off, the growth operation had been around for about half a year, and in that period we’d incrementally released site updates targeted to specific channels, users and types of content. We tended to release things that were high-impact, which meant it was easy to suss out user and channel trends.
My initial work on this project was especially fun because the research required a lot of resourcefulness for some of the more experimental blog elements or content types that we didn't have existing data for. I spent a lot of time investigating data for comparative users and channels, as well as reviewing the user feedback collected over the last several months, and thought about how we would be able to extend our learnings from past projects to this one.
The biggest questions on my mind were ones that you should always have at the start of any project: What is our criteria for success? And how will this translate this into measurable KPIs? For me, personally, those are the most fun questions to solve.
James: What was the big discovery you came away with?
Esa: After looking at the data and considering our evolving web and content strategies I had two big takeaways: we needed to be fully mobile-responsive and we needed to split product updates into their own feed. They were both major changes to our blog, and they were both decisions supported by the user personas I’d developed and the data we were working from.
James: How did we go about the initial designs, and what kind of process did you bring in to get things kicked off?
Esa: Before we talked about the specifics of the blog UI, we defined and prioritized a series of requirements – we especially considered what we needed from the infrastructure of the blog itself.
The infrastructure often isn’t visible or noticeable to the end-user, but it directly affects so many aspects of the project: URL schemes, how pages will be indexed, the relationship between elements, how users will find and experience content, how we test things, and even how we track the KPIs defined for the project. In short, a content-rich structure like this requires a lot of thought and time dedicated to infrastructural elements before you can even think about the design.
From these requirements we sketched out basic UI elements, imagining how these would translate into things like navigation or indexes.
From the basic elements, George mocked up a series of high-def prototypes. For a lot of people going high-def right away is pretty unusual, but George is really design-oriented and has a knack for synthesizing a number of competing variables into something really cohesive and workable. This is a luxury because the stakeholders who are involved in our design processes tend to be more responsive to having a really realistic, visual model to work with.
From there, we asked ourselves a lot about the user journey and how a user would achieve their goals: How will the user know they can get from Point A to Point B? What specific behavior to these elements encourage? Is this the best way to translate an element to X device? What is the best way to implement this technically?
James: So when you guys found a working design, what happened next?
Esa: Well, outside of the design choices we made for the UI, there was a lot to do – we had to migrate, reformat and QA just over 200 posts and their corresponding redirects. George additionally built the entire CMS that the Content Managers work with from scratch, and then customized it according to the Content Team's needs, which was no easy feat. We also had to decide which parts of the blog would be part of the initial release, and which parts would come in later iterations.
James: What’s your favorite part of the end result?
Esa: It’s something that readers can’t see, but as people reading a data company’s blog would probably enjoy: the analytics setup! Working from the project KPIs and the hypotheses we’re testing, I designed a fully scalable tracking framework complete with custom reporting focused on our most valuable KPIs for the project. Over the next months it’s going to be really exciting to see how people interact with the new properties, especially the Product Updates, which is targeted toward a very specific type of user exhibiting very specific preferences and tendencies.