In the last three months we’ve been working on a review of categories on Folksy. We’ve read your feedback, reviewed the site analytics and tested our assumptions and ideas. Now, we’re confident that the updated navigation will be better for you and your customers.
We used a variety of different methods to analyse category performance:
- Google / web analytics – to see how people enter and exit (and move around) categories. Good for existing category performance.
- User testing the site – to see how people use categories and describe their browsing experience. Good to test out existing and planned categories but only in small groups as very time intensive.
- Card sorting – a method to determine how people group together concepts / words. Good for testing out existing and planned categories. Scales reasonably well.
Each of these methods has its own benefits and also its own weaknesses but when taken together they provide a good means through which to assess performance.
What categories ‘do’
Categories act as a pointers to the kinds of thing a site sells: they are one of the ways to navigate around.
In sites where the stock is well known then categories are relatively easy to determine. If you sell footwear for example you might have Men, Women, Children as your top categories followed by different styles for example, casual and formal, or type such as trainers, boots, shoes, loafers, slippers etc. In a marketplace like Folksy where people could feasibly make just about anything (lunar spacecraft might be a challenge) it becomes slightly harder. We have to prioritise and adapt according to what people search for as well as what people list. We mediate between the two parties of buyers and sellers and try to reduce the friction involved in discovering things people might want to buy.
Last year we implemented a new pattern for categories to Folksy. We followed the standard e-commerce model with the main section headings acting as a funnels to a much greater number of associated categories. This pattern widely used by department stores (for example see John Lewis).
The new categories were driven by sellers explicitly deciding where their item should go and putting their work ‘into’ a category. This replaced a hybrid category system we used to have that worked on keywords to determine categories.
We tested the ecommerce navigation pattern before releasing it and it worked well. It was familiar to people and it made them feel comfortable. However, there were also some issues. The most significant issue was that it forced jewellery makers and people making bags, sleeves and accessories to choose between a male and female audience when their products were often unisex.
These are the ‘headline’ things we have discovered through our analysis:
Many item and shop pages perform better as landing pages than category pages do. This is not entirely surprising as some items get shared widely and may become ‘viral’, shared amongst friends of friends on sites like Facebook and Stumbleupon and services like Twitter.
“Men” and “Women” as labels do not act as good ‘wayfinders’ into the kind of content Folksy has. Approximately one third of everything for sale on Folksy is a jewellery or accessory item. Yet neither are mentioned in the core navigation. For people coming to Folksy for the first time (wherever they land) this isn’t particularly helpful – more useful pointers would, we believe, provide better jumping off points into the content.
A formal taxonomy / category system is only part of the solution to browsing a marketplace. A common browsing behaviour on Folksy is often referred to”wilf-ing” or “what was I looking for?”. Having landed on Folksy they scan the content then follow interesting links, creating random trails of exploration. They aren’t necessarily guided by a premeditated decision to search for something specific, rather they want to be inspired or to discover something new, something unusual. So whilst categories work to help people browse, other prompts and features are also important such as search, tags / facets, editorialised content (gift guides, featured sellers) and also best sellers.
More labels are helpful. Category labels often act as prompts. Many people don’t know what to look for so are looking for prompts. The benefit of seeing lots of prompts needs to be balanced against the ‘overhead’ of processing all the labels (we can only hold around 10 things in our heads at any one time) and the space a large site-wide navigation would take up.
Makers don’t tend to make many things for men. There are roughly eight times more items available in the Women’s section than in the Men’s.
Labels/categories don’t need to be ‘logical’. Due to the type of work available through Folksy (and the number – over 130,000 items as I write) people don’t always think ‘rationally’ about particular categories. Combining practices (sewing, silversmithing) with items seems to help rather than hinder browsing. Some people want to explore through a variety of different views on the data (the items available).
As a result of the review we have decided to change the category system to the following core section headers.
- Clothes and Accessories
- Home and Garden
- Weddings and Parties
- Art and Cards
- Craft Supplies
The main changes are obviously using Jewellery and Clothing and Accessories instead of Men and Women. The other significant changes are:
- Bringing Babies and Children under other relevant categories, so for example nursery will become a room in Home and Garden just like Kitchen and Dining, and babies’ clothes and shoes under Clothes and Accessories.
- Pulling out party and event items from Babies and Children (which wasn’t performing well enough) and putting these within a section “Weddings and Parties”.
For a full breakdown of the new categories, please refer to this spreadsheet.
How will it affect you?
Only those people who list in Jewellery will have to re-categorise their work. All other items will be merged or moved automatically.
People who list in Jewellery will have to select a sub-category for their work. This will help people find their work more effectively. When the revised navigation is launched all Jewellery items will be placed in the sub-category “all”. You can then move your work to the relevant category using the bulk listing editor.
We want your feedback on the proposed categories before we implement them.
Take a look at the proposed changes. The category changes are in yellow and we have included information on how the changes will affect different items.
If you have any feedback or comments please leave it via Suggested Features in Support. Only feedback provided through Suggested Features will be considered in the review.
The closing date for feedback is the 21st September.
When will this happen?
The navigation on Folksy will be changed on or before the 1st October after we have reviewed feedback. The bulk edit facility will enable people to move their work quickly and easily.
There will be no further public facing work released between October and December this year. After the major changes we released last year we appreciate that you’re busy focusing on making and selling and don’t want to be distracted with any new features or functionality. That doesn’t mean we’ll stop developing, on the contrary we’ll be continuing to improve the service but will not release any significant changes until the New Year.
Last year we had a specific category for Christmas items and people have asked for this again. It was a temporary category and meant for things that were part of Christmas, such as Christmas Stockings, rather than potential Christmas gifts as *everything* on Folky is a potential gift.
The Christmas decorations category is now live and listed, as last year, under “Home and Garden” in the top navigation.
The category will be live until the end of January. Anything still for sale after this time will be able to be bought but will only be able to be found through search or via a direct link.
We’ll be monitoring the new categories during the busy Christmas period and will provide a review of performance in the New Year.
Other findings from this review will be tested again and it is our intention to continue to find ways to improve the browsing experience with different and interesting ways into the variety of things available to buy on Folksy.