SPSS Online Community

Maximize Your Data Analytics.


In early 2015, a product manager at in IBM's Analytics unit created the IBM Predictive Analytics Online Community. The community had three components joined by a single landing page: a blog hosted on a Wordpress platform, a forum hosted on IBM's instance of AnswerHub, and an Extensions Gallery hosted on GitHub.

Later that summer shortly after I joined the SPSS Statistics team as a Designer, we ran a 30/60/90 Digital Refresh project effort to improve the Discover-Try-Buy funnel for IBM's SPSS products (Statistics and Modeler), which included leveraging the online community as a marketing and support tool.


In terms of the online community's role in the Discover-Try-Buy funnel, marketing, support, and product had different ideas:

  • Marketing saw the community as a place where prospective users could learn about SPSS products and find trial versions of the programs for download and use, therefore focusing more on the Discover and Try portions of the funnel.
  • Support saw the community as both a way to provide additional documentation for users and a potential mechanism for self-service support, focusing on the Try and Buy portions since this was where the largest user drop-off in the funnel was occurring.
  • Product saw the community as a way to promote the creation and sharing of Extensions, addons for Statistics and Modeler hosted on Github that helped enable analytics methods that weren't native to the software suite. Because we were a smaller team with fewer reseources for conducting user research, I saw the community as an opportunity for us to better engage with and understand our users in service of improving the user experience for SPSS Statistics.


We kicked off the Digital Refresh project with a two-day design thinking workshop to determine key persons, identify pain points, and define goals.

The online community as it already existed was created from one of IBM's Wordpress templates, and I had been tasked with ideating a new visual design for the landing page prior to the Digital Refresh project; this task became my primary focus for the 90 day duration of the project.


In the beginning, I focused on wireframming based on the requirements given by other stakeholders in the project. After gathering sketches from the heads of marketing and support, I created a series of wireframes with alternate versions for review. We had limited control over the UI/UX of blogs and the forums since the former was controlled by IBM's DeveloperWorks unit and the latter relied on a third-party service, but we were able to fully design the landing page and other informative pages.

Around halfway through the project, we lost the developer that had been allocated to this project, and we were starting to slip on other deadlines. At this point, I took on full ownership for the design and development of the online community, taking on a larger role in driving the design process as opposed to wireframing based solely on requirements and feedback provided.


Because we had a bit of a time crunch to complete the community's redesign by the 90-day deadline, I broke the usual design process and begin working primarily in code.

The community's product manager and I looked at other online communities, both internal and external, for visual design inspiration and content strategy references. The community had originally been created with the intention of highlighting SPSS's extensibility, but we decided that success for the 30/60/90 project required a stronger emphasis on predictive analytics in general.

We used our comparative analysis and developed a heirarchal flow for the landing page, and we decided that the visual style of the website should be a cross between SPSS's legacy and IBM's then-new design language.

  • We created a large banner area at the top of the page. We had orignally planed on rotating data science jokes and puns in that space, but, for about a month, we celebrated a the retirement of a colleague beloeved by the broader SPSS user community with a gif in the baller. Later, when newer versions of SPSS Statistics and Modeler were released, we highlighted them in the banner.
  • Below that, we had three primary actions placed roughly above the fold that linked to the other three sections of the landing page.
  • We decided to have a section for a Featured Post in the blog section so that we could highlight certain articles, such as product updates or interesting data analytics methods. We also used this space for various announcements, such as the link to a survey regarding SPSS Modeler usage on behalf of IBM's Data Science Experience team.
  • We hoped that placing the Forums section after the Blog section would prime users to focus on predicitve analytics in general; this ran slightly contrary to Support's preference for the forums being a form of self-service support, but we believed that fostering a "predictive analytics" community, as opposed to an "SPSS" community, bringing more value to the community by bringing in more viewers and, hopefully, contributing members.
  • We placed a preview of the online Extensions Gallery at the bottom. This was intended to reinforce a cascade from the top to bottom; the blog entries were intended to be easily consumed by a broader audience while the extensions were generally the interest of heavy and more-invested SPSS users.


During this project, I learned about taking ownership of a process; even though this was my first major project as a working designer, that shouldn't have been a reason for me to take the backseat to other stakeholders in the project.

We relaunched the redesigned community on time at the end of the 90 days, and viewership increased to over 20,000 hits within the first month.

The community manager and I had also developed a roadmap for the community, which included instrumenting it for website analytics so we could identify ways to better improve the community and redesigning the Extensions Gallery experience to become more social and App Store-like.