The purpose of this project was to optimize the workflow for how engraving templates were made and then assigned. Engravables makes up a substantial amount of revenue for both Stella & Dot and Keep, which is a sister company that focuses mostly on customizable jewelry pieces.
One of the main reasons this project came into the scene is because the company recently acquired a new machine for engraving that would allow different materials to be used (leather bands, semiprecious gems, wood, etc.). Knowing that engravables would be expanding as a process and as a revenue source for the company, it became pretty apparent that something had to be done with the current setup.
I conducted the research required and led the design for the new workflow to be put in place. While working through the project, I collaborated with the UX team (a Product Designer, a Senior Designer, and the Director of UX) for important feedback. I also conducted a large portion of my interviews with a member of the Site Ops team who would primarily be using the new workflow design.
In all, the project consisted of:
interviews with the Site Ops team to gain a clearer understanding of the flow currently in place
constructing a job map to dissect each step of the current flow
deciphering the MVP's that would have primary focus in the redesign
wireframes to build the skeleton of the new layout
a clickable prototype to see the potential workflow in real time
a usability test on Site Ops to gain final feedback
Once I began prepping for my initial research, I approached my potential findings with the Jobs To Be Done (JTBD) theory. The theory essentially states that any job can be deconstructed into specific process steps in order to gain clearer insight on the overall job. It also provides a structure to capture all of a user's needs and to identify opportunities of growth. I constructed my first interview with Miriam on this basis and formulated my findings into a job map.
As you can see, there was a lot of information gained from the JTBD approach, and a lot of pain points found along the way. When I sat down with Site Ops, she voiced a lot of issues with her process in general, but there were some key pain points that kept resurfacing.
In its current state, any template on the back end site can't be deleted, and the only edits you can make on templates are on the display name and the engraving method. If there's some template with a mistake on it, the only work around she has is to write DO NOT USE in the name field, which clutters the page with a bunch of useless templates mixed in with working ones.
To sum up my findings from Site Ops’s current workflow with engravables, I found that the actual creation and assigning of templates is fairly straight forward. The problems more so lie in what happens after those two jobs are completed. From this I was able to establish my 3 top MVP's:
editing all template fields post creation
deleting templates with a follow up confirmation message
adding multiple swatch images to selectively choose a swatch to represent on the product level
Since engravables exists within a larger backend site for admin, the framework in itself lives in an already cohesive layout, so it wasn't an enormous change.
To address the limited editing capabilities, I changed the format to show that all sections have the same coloring, illustrating that all text fields can be changed.
To revise the deleting aspect, I added a delete button under the action column in order to hopefully clean up the template library of "DNU" templates. Earlier on in my user interviews, Site Ops also requested a confirmation that a template would be deleted in the case that the team accidentally deleted a template they meant to keep. I employed a second popup to ascertain whether a Site Ops user truly would want to delete the template in order to resolve that issue.
The Site Ops team also made it a point that some templates they would want to delete might possibly be already tied to existing products and therefore should be examined prior to deleting them. The reasoning behind this was that if a customer had an engravable item in their cart with a template that Site Ops was trying to delete, deleting the template without investigating its usage on the B2C site could mess up a customer's order. To alleviate this, I added an error popup modal to indicate that a particular template might already be attached to a product, and in order to fully delete the template, Site Ops must first detach the template on the product level and then return to the template to delete it permanently.
To fix how swatch images were chosen, I decided to allow multiple images to be picked at the same time when searching for template images. Once they were added to a template, I created a way to toggle through the images to let a user verify how many images they chose and to see visually that they chose the right images as well.
In addition, once a user saves their swatch images on a template, they can go into the product page and selectively choose which image to be represented by that template on the B2C site.
While I integrated all of my MVP's into my final work, there was a bit more I would've liked to finish given if I had no time constraint. There were two major issues I would've liked to tackle:
Implement a way to detach templates from products on the error modal - there will be times when the Site Ops team is aware a template they're trying to delete is tied to an existing product. Rather than having them click on a link to open a new tab to the product specific page, I think it would make more sense to add a detach button from the modal itself to decrease bouncing back and forth between pages and to save time.
Add different languages to store specific columns - you might've noticed a blueish green header on some of the columns within the redesign, which is an indicator of copy that should show up in a particular language given what country the site is catering to. Stella & Dot reaches a number of countries, most notably Canada who has an obviously high population of French speaking customers. While the Canada site is in French, the backend site currently only has store specific language in English. This means that if a Canadian were trying to buy an engravable product on the French site, they would also see the display name of templates and their respective hint fields in English. While this remains a pretty big oversight, it was only discovered by the team and I after my 2nd round of wireframes had been created and ultimately didn't fit in with the timeline I had.