Don’t laugh. I’m sure you’ve done this before. At the office, there’s a refrigerator cleanup every two weeks. At least I think it happens every two weeks. The office administrator sends out an email or posts a note on the fridge, warning you that things will be dumped if they’re not labeled. You’ve seen these long-forgotten food containers of who-knows-when science experiments pushed up against the back of the fridge. Same with those things that start growing in your pantry…. Don’t ask. I won’t continue. Please don’t tell my mother I had so many potatoes left.
When it comes to explaining governance, the one in the kitchen is the best example to illustrate exactly what happens when you take a taxonomy for granted. Not only do you see it, you smell it. You’ll feel it if you consume the foods way past its best by or expiration date. You’ll taste the food quality deteriorate if the ingredients used are not as fresh as they could be. What better way to illustrate ROT analysis than the five senses? This kitchen analogy doesn’t stop at organization.
Previous articles in this kitchen taxonomy series went through outlining the business case for building a taxonomy, card-sorting to generate labels, and tree-testing to assess findability. At this point, it’s an overhead project at most companies.
However, this is an important reminder: Once you’ve developed and applied that taxonomy to your content, the project is far from done. Establishing a taxonomy governance is a crucial endeavor, one that makes sure that your content or application continues to be well-maintained and performing as well as the day it launched.
We’ve asked these questions before in part 2, Planning a Taxonomy Project. Why and what will the taxonomy be used for? Who is using it? How will it be built? How will it be maintained? How will we ensure that it is properly maintained? And—of course—who will do all of this?
In this part 6, we’ll be revisiting those questions and think about how to account for taxonomy management and quality control. Remember, taxonomy governance maximizes the ROI of the taxonomy project and prevents the moldy science experiments in the pantry in the first place.
For documenting these discussions and officiating them, you should consider drafting a document like a charter to keep these decisions in line. In this charter, include the following sections:
- Purpose. What value does this taxonomy bring?
- People. Who makes the decisions and who manages the taxonomy?
- Process. How often and how does the taxonomy get updated?
Our kitchen taxonomy came about because we have many cooks in my household. We need to be able to:
- Know the name of the ingredient in English and Chinese. Specifically, we needed to know the regional differences in American English, Northern Chinese (Putonghua), Hong Kong Chinese, and even Chinglish!
- Know how to find unusual ingredients. Referring to the 80-20% rule, this is the 20% of things that we don’t use often.
- Know how ingredients are organized. This considers streamlining workflows so that we can find things easily when we need them.
Over time, this project has evolved. Talk about scope creep! From the initial physical classification of the items in my pantry, it has evolved to organize the printed recipes in my recipe binder and the digital copies of the recipes saved in a note-taking app.
The taxonomy scope itself has remained the same: spices and food items. Along the way, I learned that storage requirements should not be a limiting factor. There are cooking ingredients that one keeps refrigerated. Not all spices are kept at the same temperature. So now sauces, cooking oils, and other condiments used either before, during, or after cooking are also included in this scope.
Purpose: What value is your taxonomy bringing you?
Stating the value of a taxonomy in an elevator pitch is important to get everyone on the same page.
Here are some ways to consider the value of a taxonomy1:
- Search. How would a taxonomy make search better?
- Navigation. How would a taxonomy support site navigation?
- Standardization. How would a taxonomy standardize terminology being used to categorize content? How would a taxonomy help create a common language?
- Discovery. How would a taxonomy help users discover new terms or relationships?
A statement of purpose for my kitchen taxonomy could be:
- We are creating a taxonomy to enhance findability for cooking ingredients at Grace’s house.
- We are creating a taxonomy to standarize the terminology being used that comes from regional differences to describe the same ingredients.
Findability and standardization here are two different goals. Determine which goal has the higher importance for its success. At the same time, don’t forget to follow through on achieving secondary goals. Break the assumption that the interim solution is the permanent solution!
After a few false starts, I decided that findability is the primary purpose of this taxonomy.
A taxonomy usually has three phases of development, notes Mark Doane2:
- Set up. What you should do short term to get the taxonomy ready for use
- Launch. What you should to get the taxonomy up and running
- Maintenance. What you should do to keep the taxonomy relevant and useful
Once you’ve determined the primary goal, you can consider the taxonomy’s secondary goals by prioritizing users of your taxonomy and addressing their needs in a structured manner. Although most taxonomy projects tend to end at setup and launch phases of development, you should do your due diligence and keep the taxonomy as relevant and valuable with maintenance. Testing the taxonomy with each release helps validate and confirm the user’s expectation and search behavior. For more information about taxonomy validation, check out Alberta Soranzo and Dave Cooksey’s work.3
People: Who should manage the taxonomy?
When thinking about taxonomy stakeholders, consider the RACI matrix.
- Who should be responsible (for the work)?
- Who needs to be accountable (for ownership)?
- Who needs to be consulted (provide input)?
- Who needs to be informed (told after the fact)?
As people who live in the house all year long, the husband and I are responsible for the daily maintenance of the kitchen. We also participate in the daily upkeep of the pantry and kitchen activities. We do the shopping and the cooking.
We usually decide which type of soy sauce or fish sauce is purchased. Personally, I tend to pick up the brand my parents use. My in-laws don’t seem to prefer a certain brand of soy sauce over another since the brands are not what they are used in China, and they definitely don’t use fish sauce in their dishes … but they determine when it is time to buy another bag of rice. My father-in-law doesn’t go a day without rice.
As frequent users of my kitchen, my parents (who visit every once in awhile) and my in-laws (who live with us half the year) are consulted in the taxonomy. They aren’t expected to make taxonomy updates; they are consulted as subject-matter experts.
If my mother-in-law has a preference for a certain brand of rice, we’ll take it into consideration, test it, and determine whether it’s worth a long-term investment. A 25-lb of rice won’t last very long while they’re in residence, but we will need to decide whether to continue purchasing that brand when the in-laws return to China. It would not be a good investment to waste a bag of rice to feed the rice weevils.
For a small kitchen taxonomy, you wouldn’t need a full-time taxonomist, a team, or a committee to manage the taxonomy. However, you should consider the following for an enterprise taxonomy:
- Editor/Taxonomist. Ideally an information architect, taxonomist, or business analyst who is familiar with the content and can manage updates, solicit feedback from end users, and integrate changes.
- Team. Ideally 2-3 people trained in information architecture (from the user’s perspective) and search (from a technical perspective).
- Committee. A small committee of 3-5 people to meet a couple times of year to discuss taxonomy changes and approvals.
Process: How often should the taxonomy be updated?
A kitchen taxonomy should be updated as often as necessary. That means it could change as often as every day while putting groceries away or during meal preparation.
In a company, however, this frequency could potentially cause chaos. An enterprise taxonomy should be updated on a regular schedule, according to defined rules set forth by a governance team or committee.
Part of this process is to set policies and procedures so that taxonomy updates are made in a consistent manner. This is important to prevent an arbitrary decision to move all the coffee to a new location.
I’d start with a few guidelines from Heather Hedden’s “Accidental Taxonomist”4 (pg. 317) and build from there:
- Rules for adding, changing, moving, or deleting terms or relationships such as hierarchical relationships, alternative terms, associative (or semantic) relationships
- Examples of types of changes to expect and the processes for handling such changes
- Specific guidelines for handling feedback and change schedules
Then, using Mike Doane’s top ten guidelines5, you’d be able to build a solid starter document for taxonomy governance. These top ten are important for starting out slowly and simply—keys for successful change management. If an organization doesn’t have a taxonomy in place already, having fully-decked out guidelines at the start is a sign of the taxonomy falling flat on its side.
When I consider my own kitchen taxonomy, the rules for adding terms are pretty straightforward. It occurs whenever the in-laws are here or whenever someone decides to try out a new recipe.
In the past year, I’ve tried and experimented with making a Chinese recipe for 8-Treasure Congee (八寶粥 bābǎozhōu). The ingredients are interchangeable and easy to put together, but it’s eight items. And you know what? There is a version of this with 18 ingredients that’s touted as an extremely healthy breakfast. Eighteen! The pantry is bursting at the edges, just thinking about it.
When it comes to indicating relationships with ingredients, it gets a little complicated. Imagine having to hunt through different places in the pantry for glutinous white rice, red beans, raw peanuts, and barley. These are all used in 8-Treasure Congee, but they are also used in other recipes, including soups, desserts, and rice dishes.
What’s the best way to add these new ingredients? Should they be grouped by their ingredient type? Or should I group them together as a special functional group as I currently do for baking ingredients? What would be the most optimal way to do this, considering workflow? If there is no good answer, is this another case where there should be two homes for an ingredient? Here, it’s time to consult the subject matter experts.
Next, I’ll talk about some best practices for enterprise taxonomies. But right now, I need to schedule another quarterly pantry cleanup session before the in-laws return from China. Somehow, our collection of ramen and spam has grown out of control while they were away…
Footnotes and further reading
- 1. Doane, Mike. “Taxonomy Governance: Why You Need It, How It’s Done.” CMSWire (May 29, 2012): http://www.cmswire.com/cms/information-management/taxonomy-governance-why-you-need-it-how-its-done-015813.php
- 2. Doane, Mike. “What to do now: Immediate needs,” in Enterprise Taxonomy Governance: Practical Advice for Building and Maintaining Your Enterprise Taxonomy (Volume 1). CreateSpace Independent Publishing Platform, 2017.
- 3. Soranzo, Alberta and Dave Cooksey. “Testing Taxonomies.” Association for Information Science and Technology Bulletin (June 2015): https://www.asist.org/publications/bulletin/jun-15/testing-taxonomies/
- 4. Hedden, Heather. The accidental taxonomist. Medford, N.J.: Information Today, 2010.
- 5. Doane, Mike. “Process,” in Enterprise Taxonomy Governance: Practical Advice for Building and Maintaining Your Enterprise Taxonomy (Volume 1). CreateSpace Independent Publishing Platform, 2017.