Background & Challenge: The indoor plant industry has exploded in recent years, growing over 50% from 2016 to 2019, then rising another 19% and 30% in 2020 and 2021. One of the stars of this new flood of popularity is the rare tropical industry. Exotic members of the Araceae family -- commonly referred to as aroids -- have flooded the United States and Canada from Southeast Asia and South America. A microeconomy with booms and busts, plants that were once seen as novel, interesting mutations in Thailand regularly fetch hundreds, thousands, sometimes tens of thousands of dollars depending on perceived rarity and size. As a result, the once esoteric hobby of rare plants has found a new, significantly larger audience. And with it, a new breed of online rare plant sellers has emerged to cater to it.

Founded in New Orleans by two rare aroid collectors in 2020, Canopy Plant Co. saw potential in the industry and used their own collections as a first round of inventory. Quickly selling out and expanding, the operation moved from their back yards to a 4000 sq. ft. warehouse/greenhouse within 3 months.

My primary challenge was clear from the outset: establishing design patterns for an industry too young to have them. While there were plenty of competitors to look to, none appeared to have put much (if any) thought into how best to categorize (I.A.), navigate (UX), or present (UI) a massive inventory largely consisting of product names in Latin.

Further, the project came with the major constraint that I would be working alone and without a budget for a developer, so all solutions must be compatible with existing Shopify apps and features.

Results:

Using conversion rate as the KPI, the project was wildly successful. While Canopy had very high traffic from the outset, helping users to browse in a more useful manner than simply scrolling through hundreds of options, then better presenting products both visually and informationally led to a doubling of Canopy’s conversion rate.

In addition, several competitors were noticed to have been implementing several of the changes made to Canopy; a pattern for the industry has taken shape.

Upon starting the project, I was given access to sales records/order history by the stakeholders in order to identify potential users to interview. Immediately, I was struck by the existence of two distinct ordering habits: Highly frequent ordering of multiple low to mid price plants (usually on the weekend), and less frequent ordering of extremely rare, high ticket items (usually on Thursdays). Operating under the assumption that this division between ordering habits would also trace division among other habits, I conducted initial interviews using the same script but with users separated into two distinct groups of five.

Insight: Personal, But Professional

All users interviewed shared similar frustrations when it came to new online plant sellers. On the one hand, users found larger startups and big-box retailers to be great at offering low-price products, but at the cost of feeling as though they were supporting outsiders who don’t really care about the plants themselves.

On the other, smaller sellers (on Etsy in particular) had negative associations of inferior products: be it fungal infestations, plants received dead on arrival due to poor shipping practices, or misidentified species, every user interviewed had standout negative experiences with homegrown sellers.

In short, users went to Canopy precisely because they met this middle ground, personal yet professional.

Insight: Two Ordering Habits, Two Mental Models

Upon starting the project, I was given access to sales records/order history by the stakeholders in order to identify potential users to interview. Immediately, I was struck by the existence of two distinct ordering habits: Highly frequent ordering of multiple low to mid price plants (usually on the weekend), and less frequent ordering of extremely rare, high ticket items (usually on Thursdays). Operating under the assumption that this division between ordering habits would also trace division among other habits, I conducted initial interviews using the same script but with users separated into two distinct groups of five.

Ranging from typical feedback questions and competitor impressions to abstract discussions on why blah blah blah… the first major insight came from an exercise in visualization:

Seasoned Plant Collectors Shop Spatially/Logistically

To clarify, experienced plant collectors think of their ideal shop as one organized in terms of simply where things will fit and survive, not taxonomically. From the point of view of botany, the only major factor in their mental organization model was growth pattern, i.e. whether it was a climbing, hanging, or so-called “creeping” plant, and this only mattered insofar as it described the plant’s shape and thus where it could potentially fit in a given space. This is unsurprising, given that a number of the more serious collectors mentioned a lack of space for more plants in their home as a major hurdle in their hobby.

New Plant Collectors Shop Visually

It is unsurprising that the younger, less experienced users tend to focus on individual, visually striking species. According to a report by the National Aroid Convention in 2020, 60% of all Millennials list “Colorful, Unique, and Variegated Foliage” as a primary driver toward buying a particular plant. This is, however, at odds with the organization of online plant sales: organization by Genus is expected as a primary sorter, but Millenial plant enthusiasts use purely aesthetic criteria to sort afterward.

Further Elaboration

Confirming and building upon this insight, The National Aroid Convention funded a survey in 2020 to better understand this new market that seemingly appeared from thin air. Among those surveyed under 35 listed “Unique, Colorful/Variegated Foliage” as their primary draw when selecting a new plant:

Insight: Users Need Reassurance, Regardless of Experience Level

Every user interviewed, regardless of experience level, dreaded the idea of purchasing a plant only to have it die because their home is not suited to it. Even those that had never had this experience, the level of emotional attachment users felt for every individual plant in their collections made the idea of it unbearable. To combat this, most users research plants extensively before purchasing; however, this was usually done outside the site in which they were planning on purchasing from. And further: users felt immense frustration with the lack of reliable information on any site that could prevent this.

Concept: Robust Facet Filtering

In order for users to find not only the plants they want, but those that indeed feel like they are personally “for them”. A nod to the fashion industry that the tropical plant scene takes its drop model from, facet-filtering in particular seemed conceptually useful for sorting a vast inventory according to a more humanized set of terminologies.

While competitor sites simply offered ‘Genus’ and ‘Pot Size’— if they offer filters at all — I categorized the filters according to characteristics gleaned from the user interviews.

1. Growth Habit: A spatial consideration, experienced users often are looking to fill a particular available spot in their home, then look for a plant to fill it. For example, a window with room for a hanging vine, or a corner for a climbing-type Philodendron.

2. Color/Foliage: A nod to the younger, individually plant-obsessed. While many plants are valuable based on pure rarity, the desire among newcomers for variegated versions and those with specific hues or leaf features is not to be underestimated.

COncept 2: Intuitive Categorization

An early audit of Canopy’s analytics revealed an alarming waste of space: almost none of the product categories were meaningful enough for anyone to click on them.

While “New Rare Plants” — those that followed a fashion-oriented drop model — took up the majority of traffic, only a few familiar Genus names registered as worthwhile categories otherwise. This problem was further exacerbated by a handful of plants that defied easy categorization, and when were included in “New Rare” were lost in the fold of hundreds of products. And further, the drop rate when browsing increased with every successive page, with the majority of users closing the tab after four pages of results — leaving less than half of available products ever seen.

A case in point: Geogenanthus Ciliatus. Exceptionally rare, moderately priced, extremely low maintenance and deeply hued with reds and purples. By all means, a desirable specimen, if only users could find it! As the only available species within its genus, the current industry IA framework of taxonomical categorization fails it, and the users who seek it. This specimen is not alone: for all intents and purposes, a third of Canopy’s inventory was inaccessible for all but the most fastidious.

While card-sorting would be my preferred technique for such a problem, this method would fail to shed much light in such an information-dense field. Outliers of the taxonomical system are commonplace when dealing with rarity, and sorting several hundred species available of a single genus would simply mirror the fatigue of browsing them. Instead: the user interviews would serve as a basis for qualitative insights, while the Google analytics info would serve as a reference point for which categories were simply taking space.

To solve this, unused Genus categories would be nested within a “See All” card. In addition, several categories would be added: “Easy & Low Maintenance” and “Staff Favorites”.

Further: during interviews it was noted that users of both experience levels are looking for their ideal plant to have three qualities on average. In new users, “colorful/variegated” was considered to be the most important quality. In experienced users, “easy/low maintenance” was primary. In both user types, the second and third qualities are variable depending on the individual. With the aforementioned facet-based filtering, categories become customizable to users, making the I.A. structure within a given top nav category more robust:

Concept 3: Revamped Product Page UI

Another nod to the plant rare tropical industry’s similarities to the fashion industry, as well as the hyper-visual inclinations of the inexperienced users, photography needed to be highly uniform and display the features users were seeking as well as create a seamless appearance within the UI. Firstly, this involved maintaining a slight top-down view: while this is standard in the industry with smaller plants, mimicking the perspective a collector has when watering a plant was considered paramount to the aesthetic. Second, as users romanticized their plants highly, I shot the photos (as well as wrote precise instructions) to mirror this romantic nature. Images would be shot with aperture wide-open using cheap, less than crisp 35mm lenses adapted to Canopy’s M43 camera on a seamless white backdrop:

While the changes made to Canopy were arguably subtle, the devil was in the details. Store conversion benefited greatly from the facet filters in particular. In addition, items that were desirable yet unfindable found their audience.

This project is ongoing yet stalled, due to hurricane Ida.