Background

The main job for retail buyers is to buy the products that will be sold in their retail stores/ecomm sites for the upcoming seasons. There is always some kind of planning and budgeting involved and it can be as simple as tracking your budget for the season on an excel sheet to working with planners and division merchandise managers (DMM) to figure out budgets, product categories and trends to buy. Enterprise retailers can spend months before market period (the buying period for the next season) calculate budgets, predicting trends and forecasting revenue. 

On the retail side of the JOOR platform, our main business goal is to increase more order engagement across the platform by providing buyers enough planning functionality so they can more accurately complete their orders and send it to the brand. It took us almost 4 years, multiple iterations and multiple people swapping in and out of the team, to finally figure out what the right product we should be building was. I was involved right from the beginning and grew a lot as a designer as I helped evolve this tool.

Industry and JOOR Terminology

- Buyer: person in charge of buying the wholesale products that will be sold in the retail stores

- Divisional Merchandise Manager (DMM): the manager of a division who supervises and works with the buyers on what to buy for the season

- Assortment (JOOR term): a JOOR document that has the group of products a buyer has selected to buy from a brand

- Styles (JOOR term): old term for products on JOOR

- Buys: products that buyer has selected

- Quantify: add the quantity of units to a product that is being bought

- Size curve/size range/size distribution: the range of sizes for a particular product category (ex. shirts might have size range A which consists of XS, S, M, L, XL)

- Size-mapping: Associating a brand size to a retailer size (ex. brand size might be called Small but retailer equivalent naming might be Petite)

- Door: a retail store location

- PO: purchase order

- Retailer hierarchy: enterprise businesses have different departments (womens, mens, kids) and within those departments are divisions (women's ready-to-wear, women's shoes, women's accessories); enterprise buyers can buy for a division or for a product category within a division (like jeans, swimwear, activewear, etc.)

State of the assortment tool when I joined JOOR

When I first joined JOOR, we already had an existing assortment tool and what it did was provide an area for buyers to see the products they selected and add quantities. What we saw was that most of our buyers, especially the enterprise buyers weren't completing their orders and sending it to their brands for confirmation on JOOR. Many of them were exporting the assortment and completing the rest of the planning and ordering process through their traditional ways of a buying excel sheet.

At that point in time when I joined, the product team (and company in general) did not fully understand all the user types we supported and there was no user research documented as well. Most of the functionality that were added at that time were built for specific clients and so the platform became really complex and hard to use.

We also had recently signed a deal with Neiman Marcus and promised them a lot of custom work so they can fit the assortment tool within their buying processes. Like the same for the rest of our platform, we ended up focusing too much on catering to their needs with the hopes that their processes would work for all the other retailers we support.

Neiman Marcus project

Being the only designer (and a junior one, at that), I was automatically assigned to design the functionality asked for by the client. It was quite a struggle for me because I did not have any onboarding about the platform and had to immediately jump into this project. I had to understand who our buyers were and what they needed to do to complete their work, then understand what problems Neiman Marcus was having because they just gave us the specific functionality they needed. After that, I had to understand why the assortment tool didn't currently fulfill these needs. All of the user information I had were proto-personas given to me from retail success managers and the product manager leading the project so there were a lot of assumptions made about our users and their processes. Any additional questions I had I ended up having to make assumptions for and trust the product manager's say in.

The specific asks Neiman Marcus wanted were:

- to have an easier sizing page to quantify units for their buys (similar to an excel sheet because traditionally enterprise buyers have been using that to complete their buys; though how the PO sheets are displayed and the content within vary by retailer)

- a way for buyers to select and apply their size ranges per product and then map those sizes to the brand sizes

- to see a financial rollup of cost, budget and other calculations by door

- to be able to group by certain criteria to help buyers more easily determine how their units compare with the rest of their order

How I approached the designs

I spent a lot of time in the beginning trying to understand more about our buyers and our platform. Since I was not allowed to talk to our users directly at the time, I had to use our internal stakeholders as proto-personas. Within my limitations and my limited knowledge about the users at the time, I tried to push the product manager I was working with to not only think about the sizing page but about the whole assortment planning experience together. However, since we had a tight deadline, I was forced to focus only specifically on Neiman Marcus' needs. Overall, the design process was rushed and I was not able to validate my designs with user testing and it went straight to build.

What I knew about the buyers at Neiman Marcus (and enterprise buyers in general) is that for quantifying they:

- needed a hierarchy data structure so buyers only have to view the information related to them

- needed to size map brand to retailer sizes before quantifying their units

- needed to reference their budgets as they quantify their units

- needed to be able to add in both targeted (projected number of units to buy) and actual units for the products

- needed to be able to quantify by door

- needed to be able to add bulk (targeted) vs. a sized (actual) units

What I didn't know was how quantifying differed between buyers from different departments and divisions or how our sizing page would fit or replace their current processes. I ended up having to take a lot of design risks. Some of the decisions I made were to:

- separate the bulk and sized units into separate tabs. This was because the information I got was that buyers would be the ones inputting bulk units then handing it off to the assistant buyer to quantify the sized units. What I didn't know was if every buyer had an assistant buyer and what the interaction was like between the two roles; was it a simple hand-off process or was there more collaboration involved?

- create a grid structure (similar to excel) that supports inputting units across many doors and many sizes. One of the internal stakeholder feedback was that Neiman Marcus had over 40 doors and some buyers (like shoe or jeans buyers) could have a lot of sizes. Buyers were use to seeing a lot of information at once and want to do as little scrolling as possible. What I didn't know was not all buyers bought for all doors, there were cases where buyers only have 1 or 2 doors and some buyers didn't have to buy different sizes (like sunglass or handbag buyers). 

- a big decision I pushed for was to have keyboard functionality, especially copy/paste since everyone kept mentioning the ease of excel. However, engineering said they did not have time to implement that and it was cut. I ended up having to design another not as user-friendly way to copy/paste.

Results and things I learned

Much of the initial release feedback was not relayed to me but some of the feedback I generally heard as we were onboarding new clients was that the size mapping and size distribution functionality was unusable because it was built specific to Neiman Marcus' use case. Toggling between bulk and sized tabs was a nuisance for buyers who had to quantify both bulk and sized units because there was a lot more tweaking and editing involved and was not just a simple handoff. For buyers that only have 1-2 doors with one size to quantify, there was a lot of unused space that could have been utilized more efficiently.

What I learned from this project was how important it was to do the necessary research to understand the entire user base we are trying to solve for and how proto-personas are a good start put it needs to be validated because there are a lot of assumptions that are made.

Assortment overview improvements

After the release of the sizing page, the product manager wanted me to focus on improving the user experience of the overall assortment tool. There were already plans to re-build the page because it was built with PHP at the time and the code was too messy to build any additional functionality on top of. Retail success managers also had a strong say in what they thought needed to get added/adjusted to the tool because of the feedback they got from clients. Again, I was not able to talk directly to buyers so I had to resort to talking with the retail success managers and the product manager in figuring out how to improve the tool.

The assortment tool is essentially a purchase order sheet that consists of:

- an area with general order information

- a financial rollup area to see how the products roll up to the budget

- a product card area displaying all the products a buyer has selected

- an area to add custom products and edit existing product information

- an area to quantify the products (sizing page previously mentioned)

The main complaints from our users were not being to see more of the product cards on the page, cluttered information on the product cards, functionalities were unintuitive and we did not support an area to add retail attributes associated to the products for data-enrichment.

I ended up doing a heuristic evaluation of the existing assortment tool to assess the interface experience and proposed some improvements. Some of the things I did were:

- assess, organize and group the similar content and functionality together

- establish a page and text style hierarchy

- introduce visible selections on the product cards with CTA buttons that enabled or disabled so users know which actions they can apply

I also added the ability to add attributes/tags on the products, which was mostly copied over from the same functionality already designed on the iPad app. Since I wasn't able to do user testing and the feedback of the iPad functionality was mostly positive, there would be less risk involved by copying the experience.

This project ended up getting scrapped because the product manager left the company shortly after I created these designs.

Holt Renfrew project

Our next big push to improve the assortment planning tool was when we got Holt Renfrew as a client. This time around, there was a new product manager that took over this project and a senior designer joined the team as well. 

The management of this project was very similar to how we approached Neiman Marcus as well. We had to design under a lot of assumptions and we agreed to build a lot of what Holt was asking for without questioning what problems they were really trying to address. What we did do differently this time was push to be able to speak and even shadow different types of buyers to understand their work processes. However, by the time it took us collect enough research to truly understand our buyers, this project was mostly built and released.

For the most part, Holt's asks were generally similar to Neiman Marcus' asks. They needed the assortments to have their retailer hierarchy, have size curves and size mapping when quantifying and being able to apply attributes to their products. However, their data model was a lot more complex than what we could support and the code for the existing assortment tool was built with no flexibility for change so we had to re-build it from scratch. Lastly, they also mentioned the ease of use of excel sheets for filling out orders.

As we started tackling the re-design, we ended up reusing a lot of from my previous design layout but tweaking some parts that we thought could be improved. What we assumed was that buyers thought of their tasks in a linear fashion, one after the other, which is why we separated the tasks of reviewing products, adding attributes, editing product information and quantifying all as different tabs. We also had to re-think how to apply attributes because we needed to support Holt's complex attribute nesting behaviors. We also had to re-design the sizing page completely to fit Holt's size mapping requirements.

By the end of this project, the senior designer and I knew Holt buyers and buyers in general would have many complaints about this experience because we had gathered a lot of user research and knew much more clearly what the different types of buyers' planning processes were like. 

Research gathered and mistakes learned

From our research, we learned:

Enterprise and mid-sized buyers

- work with the planning or merchandising teams before the market period starts to figure out what products sold well the last season and how their season's budget should be allocated by brand, product category, door and/or any other criteria

- work with the DMM and/or fashion office to determine themes, trends, types of products to be focused on buying this season

- once buyers have all the above information, they schedule market appointments with the brands to review what the brands are selling this season

- during a market appointment, the buyer will work with the brand sales rep to select the products they're interested in in the form of an assortment, make some notes and potentially add target units depending on the type of buyer

- after the appointment, the buyer will review the buys again, make some edits and either send the assortment to the assistant buyer to continue quantifying the actual units or do it themselves if they don't have an assistant

- there might also be review sessions with the DMM during market but it depends on the company

- there is also a lot of back and forth communication that happens with the brand sales rep after an appointment because buyers might need more product information to inform their buys, product information might change because brands are still finalizing product details and/or buyers might ask for exclusive products to be created for their retail store

- as the assortment buys and units get finalized on the buyer side, the buyer will then send it to the brand for final approval; as this happens they (or their assistant) will start filling out the retailer attributes for data-enrichment

Small business buyers

- are usually small boutiques with a few branches or less

- many buyers are also the boutique owners or work very closely with the owners

- have a budget for the season; do not need to strictly follow their budget; do not have as much budget/planning prep work as enterprise buyers

- schedule market appointments with brand sales reps

- during a market appointment, the buyer will work with the brand sales rep to select the products they're interested in in the form of an assortment; many buyers will fill out actual quantities during the market appointment or shortly after and immediately send to the brand sales rep for confirmation

- there is also back and forth communication that happens after market appointment because brands might be updating product details but it's not as involved as the enterprise buyers' processes

All buyers

- have a budget for the season which is then sliced by brands they buy from

- compare their budget against their actual buys

- have to schedule market appointments with brand sales reps

- wants to see visual images of their products within each brand they by from and/or across all the brands they buy from and potentially visually merchandise them

- constantly communicate to their sales reps during the market period

- constantly review and update their buys during the market period

- most buyers use excel or some sort of excel-like tool to track their buys and units

We presented these findings to the product manager and worked with him to concept an idea of a truly innovative planning tool that would have the foundation common to all buyer processes but be flexible and customizable so our buyers can tweak the tool to fit their specific needs. 

We envisioned a tool that would allow buyers to:

- add all their season's buys across brands in a space

- filter and view by any criteria they need

- visually merchandise

- tweak and compare products against budgets by brand, door and/or product category

- easily edit their products, details, attributes and units all in one experience

- share their assortment at any time with sales reps to collaborate in real-time

- direct comments on any part in their buying space to their team members or sales reps

After I created some happy path mocks, I did some initial user-testing with a range of different buyers to validate our concept and we got overwhelmingly positive feedback so we knew we were on the right path. I continued to user-test as I added additional functionality to make sure what I was designing was something that was needed and easy to use.

Finally innovating

To evolve our current assortment structure (1 assortment = 1 brand order) into the ideal structure we needed (1 workspace = multiple brand orders) and to make sure there is a smoother transition for our users, we had to phase out the work.

Phase 1a - Build new assortment UI, adding in foundational functionality

Phase 1b - Add sharing and commenting functionality

Phase 2a - Add multiple brands into one assortment (workspace) - currently here

Phase 2b - Add in workspace settings; add in workspace management

Phase 3 - Add in visual merchandising and budgeting

Phase 4 - Add in analytics

We have come a long way in evolving our assortment tool and as a collective whole, we now see the need to validate our decisions with user feedback to not waste time, money and user trust. We are now developing a tool that will not only fulfill the pain points our buyers currently have but actually cut down the areas of churn in their planning process and save them time so they can do what they were hired to do.