UX Research Case Study 1 : Celonis Help Center

helpcenter-01.png

 
Initial hand-drawn concepts for Help Center.

Initial hand-drawn concepts for Help Center.

The Approach:

With Celonis’ transition from an on-premise software to a cloud platform and its expansion of process mining-related services, users would be confronted with an entirely new user flow, options and functionalities. To offer support, I began to research, design and validate a help center solution.

Research and Development:

A general problem statement and complicating issues were compiled. Onboarding and support center methodologies of other softwares were collected and assessed. Proto-personas were devised to define differing levels of knowledge that existed between our users.


Participants were asked to assist a hypothetical intern in understanding  Data Jobs . Heat mapping reveals a composite gaze fixation of all participants on the menu item,  Data Jobs , which was correctly identified as contextually specific in contrast to global help resources provided below.

Participants were asked to assist a hypothetical intern in understanding Data Jobs. Heat mapping reveals a composite gaze fixation of all participants on the menu item, Data Jobs, which was correctly identified as contextually specific in contrast to global help resources provided below.

Discoveries:

Ongoing usertests of our different services revealed that many participants found themselves stuck at intermediary steps of setup within different cloud products, often confused by our technical terminology. An initial hypothesis was that users would feel a conventional help link or structured documentation would be limited in providing specific help as they navigate through the platform. Therefore, three conceptual versions of a help resource were devised and tested: Two that provided a traditional, all-inclusive help center available through an icon in the header, and one version that provided contextually-specific help along with global resources.

Validation:

Usertests featuring high-fidelity click-dummies (prototyped in Marvel) with 30 participants revealed that not only did a help center accessible through a hover interaction perform most intuitively, but it also provided suggested information that before had been unknown to users or would have otherwise been difficult to locate. Implementation could now begin.

Heat maps and ‘Time to First Fixation Mean’ measurements, both obtained through eye-tracking userstudies, quantify the discoverability of the Help Center and its key functionalities .

Heat maps and ‘Time to First Fixation Mean’ measurements, both obtained through eye-tracking userstudies, quantify the discoverability of the Help Center and its key functionalities .

With a working version, I tested users both in Munich and New York City offices to account for any potential cultural differences. I created a user study that employed eyetracking, to be used with 7 participants at each location.

Results confirmed users easily located the tool, with an average discovery time of 4.29 seconds, and could differentiate between global and contextual information.


This is my new best friend.
— – Novice user
This is something I’d use in a sales demo.
— – Expert internal user

The finished version of Help Center, demonstrating interaction and search.

The finished version of Help Center, demonstrating interaction and search.

Conclusion: 

Continued field interviews with customers confirm the lasting success and support that the current help center implementation provides. Tracking analytics confirm that end users regularly utilize the tool as they navigate throughout the software.