Artboard 1 Copy 5.png

Inventory Quality



Etsy: Inventory Quality

For the past few months, I've been a part of a pilot program that is redesigning the way that product teams work at Etsy. We're transitioning into cross-functional teams of designers, copywriters, engineers, and project managers using agile methods of working (my sincerest apologies for all that jargon). In this pilot, our task is to improve Etsy’s inventory quality through outcome-based work. 

Our two outcomes are to provide the right information for the product and to make sure Etsy surfaces products buyers are excited to purchase.

We broke this outcome into separate, testable hypotheses.


Hypothesis 1: Because the research tells us adding attributes is a lot of work we believe that helping sellers add attributes more quickly and easily will result in sellers adding more attributes.

One proposal of value that we've been testing has been focused on increasing attribute adoption. On Etsy's platform, Etsy Sellers have the option to select category-specific attributes (if you're in the rug category, you can select the type of rug—runner, area, accent, etc.). Currently, attribute adoption is low, and because of this, Etsy Buyers are unable to filter to find the perfect item. 

We kicked off by having a brainstorm of different tools we could build to help sellers add attributes quickly. From there, the product designer and I separately worked on a few dozen mocks. I opted for a tool that added one attribute at a time since our PM kept encouraging us to distill the experiment down into the smallest slice of value.

Artboard 1 Copy 7.png

We went forward with this idea of quick-adding a single attribute and I collaborated with the product designer to work through the final flow. 

Artboard 1 Copy 8.png

Engineers are currently building out this experiment and if it performs well, we can expand this feature into other attributes and into a bulk edit tool.


Hypothesis 2: Because the research tells us important item information often lives in the description and is often repeated throughout the listing we believe that pulling unstructured data out of the description and asking sellers to move it to the attribute fields will result in higher attribute adoption.

To kick off this hypothesis, a product designer and I separately worked through a dozen quick sketches in the morning, ranging from low engineering lift (such as a reminder) to a more intensive tool (such as a smart adding feature). In the afternoon, we met with the PM, who wanted to go forward with a more intensive option (my right-hand sketch below). 

Artboard 1 Copy 9.png
Artboard 5.png

We collaborated to refine the functional and visual aspects of pulling the attributes from descriptions before I handed it over to the engineers. 

Artboard 1 Copy 10.png

This experiment is currently being built and will launch in the next couple weeks to English-speaking sellers. If we see a meaningful increase in attribute adoption we may build out a similar tool for other types of attributes. 


Hypothesis 3: We believe that improving where and how we display attributes on the listing page will result in buyers finding the information they need to make a purchase decision and sellers being more likely to add attributes.

I kicked off this project with a product designer by focusing on the overview section of listing pages. Here's what the current page looks like:

Artboard 1 Copy 13.png

 I worked with the product designer to refine the type and add a call-out to what makes a particular item on Etsy 'special' with an icon. They're drawn in a new brand refresh style that has slowly been making its way into the product.


We worked with our brilliant copywriter and project manager to hone the language accompanying these icons. We tested the iconography and text with buyers in lo-fi research sessions.

Artboard 1 Copy 11.png

With a few valuable insights from research (everyone thought 'made to order' meant personalization) we made final copy decisions and handed the work over to engineering.

Artboard 1 Copy 12.png

We had great feedback from Etsy Sellers during this experiment (a rarity on the Etsy forums). 

Screen Shot 2018-06-03 at 10.14.31 PM.png

Since the initial experiment was a success, engineers are making a few category-specific engineering adjustments before ramping the experiment up to more types of sellers.

This is just a snapshot of the dozens of ideas we've tested out with sellers in this pilot.