Ask the user, JACK!

Its hard to implement user feedback; but its harder to even gauge it — this has been one of the key takeaways of this project aimed at recommending improvements to Kennedy school’s career portal JACK. Understanding Karin and Paula’s perspectives, their current methods of data analytics, usability hypothesis followed by a product tear down, stakeholder map and value proposition template brought to light that the constraints to improving JACK may not be in resources as much as in understanding the user’s pains and possible gains.

I recommend making user (students for the context of this memo) the center of decision making to design better user journeys and making the portal far more user friendly. This could involve better nomenclature, improving readability and reducing the number of clicks to accomplish critical goals like applying to a job or booking a coaching appointment.

This positioning of the user at the center, calls into re-evaluation everything including the name of the portal. By the self-admission of Karin, the primary manager of JACK, herself, we hear that not many students know its full form. The name has no bearing on the job this portal is supposed to do. Maybe putting the user at the center of decision making would help us identify a better name. For instance, career management portal at Wharton Business School is called Careerpath. JACK may have been an easy acronym for owners of the service (jobs and careers at Kennedy school), but for students a name that calls out the functionality (Like Careerpath) may aids better organic recall and could improve usage rate. The issue of comprehension of the user, derived through research should take precedence over administrators hypothesis over what works best for them.

This recommendation hold true for the choice of color palette, font and aesthetics of the portal as well. As a representative user I thought that color contrast could be used more effectively to define different kinds of actions and prime real estate could be optimized. For example, I would reconsider using home page top banner to discuss where does JACK source their job openings from.

Karin and Paula mentioned that they evaluated renewal of contract with Symplicity instead of Handshake ONLY by speaking with other peer schools and testing the portal themselves. Their concern over data privacy, stability and motivation to house multiple functionalities like alumni resources were very placed but it did not sound ideal that user feedback/ usage diagnostics was not a central element in this decision process. I could see very high frictions in user journeys but it occurs to me that the digital team of career office have centered their data analytics to counting coaching signups and base its employer outreach guided by number of applications for jobs posted. Basis the user-centric design philosophy and my intuitive experience as a e-commerce manager in the past I am wary of the number of incorrect hypothesis such lag metrics like above could lead to ; a particular job opening may get more hits sheerly because it was more visible. It is critical to observe user behavior in terms of clicks, bounce rate, dwell-time metrices and collect qualitative feedback from new and existing students to optimize multiple user journeys like applying to a job or booking a coaching appointment in order to arrive at actionable insights about how to improve the portal usability.

A quick skim of the website and attempts to follow through on any of the use-cases of JACK presented a wealth of information on user difficulties. For example the list on the left is not mutually exclusive and collectively exhaustive. This leads to confusion amongst students.

As a new user I could not understand why “signup for interview” is separate from “view jobs” and why are “Career chats” separate from “coaching appointments”. There is in fact no opportunity to signup for interview when user clicks on that option — it merely acts as a cart when a student views jobs and applies through “view jobs” page.

Here’s a representative user journey of trying to book a coaching appointment to illustrate the need for optimization. The user clicks on “Make a career coaching appointment” and receives a ton of information about the modality of this appointment for the 1st 3 clicks. In fact, the button to take the user to the next screen is buried deep under as the multiple announcements take up the whole screen space. This could lead to bounce off as some users would just find the friction of scrolling so far down tedious.

The appointment scheduler is a significantly long form and has multiple friction points.

For example it asks the student to choose a location, and the drop down menu has no options in it.

The student realizes through trial and error that no advisor has an appointment for 45 minutes even though the drop down menu establishes it as a valid option. It may be more user friendly if the 45 minute is auto-locked when no appointments of that duration is available.

From start to finish it took me 21 clicks and 10 minutes to book an appointment with an advisor and it seems that the process could be made far less ambiguous using user centric design. Analysis of bounce rates and dwell times can support or refute our hypothesis if the points in the user journey is in fact complicated enough to warrant change.

The same philosophy could be used to optimize the process of applying for a job or connect with alumni. At the cost of listing a dirty list of issues here are a few things we may want to fix in JACK.

1. Incomplete information about employer contacts
2. Messy placement of blocks

3. The homepage banner inviting coaching appointment application only gives more information and does not link directly to the appointment scheduler.

This list is representative of possibly many other glitches and scopes of improvement that shall continually get better. The recommendation therefore is not to just fix specific glitches but the way the team approaches improvement planning. Current improvement planning seems to be focussed on organizational need, organizational understanding of user need, peer benchmarking with little to no room for voice of user.

In order to build a more user friendly customer journey it is critical to walk the shoes of the user and as we learnt in the Calfresh case — have the humility to know that as admins we are not the user and therefore lean heavily on observational and conversational research to improve user journeys for multiple use cases of the JACK portal. That may require organizational incentive and KPI changes to ensure that those with maximum interest also has significant influence in co-creating the solution.

A few ways this could be achieved may be are :introducing a NPS or user satisfaction metric as a key performance indicator of the success of the JACK project, introducing simple tools like google analytics to continually track and reduce bounce rate and adopting some of the most commonly used performance metrics of commercial platforms in establishing measures of success.

This blog post was was written in response to an assignment for the course DPI-662 Digital Government: Technology, Policy, and Public Service Innovation at Harvard Kennedy School.