Taxonomy Implementation Analysis
Overview
In my last post I shared some taxonomies that, if implemented purposely, should provide value through:
Ease of use and findability for users
Gains in operational efficiency through common understanding
Readily available data insights
But, right now they’re just sitting in a spreadsheet! I need to think through:
User needs - what functionality contributes toward ease of use and findability?
Operational efficiency - what systems and components would benefit from systematic understanding?
Data insights - where and how will measurement be handled and is that information readily available?
As I previously mentioned, we want to avoid process for the sake of process. Structure should always have a purpose. But, if we can identify how activity can be labeled and where it will be measured, we’ll have a general idea of scalability. After a bit of ideation, I came up with some features that I would like to implement. For each, I want to define why it’s helpful to the user and then prioritize based on the RICE framework. My only revision to this framework is that I’m going to estimate reach similar to how you would estimate effort sizing - 1,2,3,5, or 8. This is simply because, well, no one is on my website! I have a good grasp of what would be needed for each of these features, otherwise I would likely dive into a separate analysis beforehand.
User Needs
Starting with the most important part, let’s identify what functionality can be unlocked through taxonomy that promotes ease of use and findability. If we go back to the value generation to users of the website, I classified the each content type as:
Blog - Education
Resources - Enablement, Education
Portfolio - Operational
There’s internal value for some content as well (Lernen durch lehren content is mostly for my own momentum), but if I’ll be standardizing some form of classification then it’s easier from a quality perspective at scale to always have classification instead of sometimes having classification. I don’t want to go through unclassified content just to confirm it’s all LDL - I’d rather have clear action items that value is being measured.
Footer Navigation
As a website visitor, I want to use the footer to see a deeper map of the website so that I can go to a category of things that interests me.
| Metric | Value |
|---|---|
| Reach | 5 |
| Impact | 0.5 |
| Confidence | 1 |
| Effort | 0.5 |
| Score | 5 |
Website Search
As a website visitor, I want to search for key phrases so that I can find what I want quickly and not have to manually navigate the website page by page.
| Metric | Value |
|---|---|
| Reach | 5 |
| Impact | 2 |
| Confidence | 0.8 |
| Effort | 1 |
| Score | 8 |
Virtual Assistant
As a website visitor, I want to quickly navigate and receive summarized information so that I don’t have to navigate and read a lot of unrelated information to satisfy my true intent of visiting.
| Metric | Value |
|---|---|
| Reach | 3 |
| Impact | 0.5 |
| Confidence | 0.5 |
| Effort | 2 |
| Score | 0.375 |
Content Linking
As a website visitor, I want to look and search for content only related to the website operations so that I can discover easily implementable solutions.
As a blog reader, I want to see related content so that I can continue reading about what I wanted to read without searching or navigating for it.
| Metric | Value |
|---|---|
| Reach | 8 |
| Impact | 2 |
| Confidence | 0.8 |
| Effort | 2 |
| Score | 6.4 |
Operational Efficiency
For operational efficiency, I want to understand what systems and components would benefit from systematic understanding. But, instead of trying to project out various use cases that are made up, I think it’s better to acknowledge that following sound database design standards achieve the same goal with more flexibility. In other words, this is something that will need additional analysis as I start standing up resource heavy projects. For now, I only have a few cloud resources and I don’t even have a test environment. I’ll add a separate task to my kanban board for this analysis as I plan my next project.
Data Insights
For data insights, I want to know where and how measurement will be handled. Again, we’re focused on user value. While it may make sense to monetize certain projects or even content itself, and I do need a way to measure costs, I’d like to avoid complex modeling for something that is supposed to just be for fun. Most of the cost and revenue data will be more tied to system design anyways.
Website Analytics Dashboard
As a content creator, I want to see trends in the types and categories of content over time so that I can understand what is primarily driving people to the website.
For this one, I see incremental improvements that can be made over time. I can prioritize enhancements made to the user experience and likely repurpose some of those pieces to granularize website analytics. But, additional work can be done if I implement other applications in the process. For example, if I utilize Google Tag Manager, I can set up custom events and I can build a custom dashboard through Looker Studio.
| Metric | Value |
|---|---|
| Reach | 1 |
| Impact | 3 |
| Confidence | 5 |
| Effort | 3 |
| Score | 5 |
Prioritization
After calculating the score of each feature, here’s where we landed -
| Feature | Priority |
|---|---|
| Website Search | 8 |
| Content Linking | 6.4 |
| Footer Navigation | 5 |
| Web Analytics Dashboard | 3 |
| Virtual Assistant | 0.375 |
An important acknowledgement here - objectively, some of these items aren’t worth doing. The Virtual Assistant, for example, shows very little value for the effort it would require. As exciting as the technology is, I don’t see value to the user at this point in time compared to other work. And, this is siloed from other work that I could be doing. The Web Analytics Dashboard, albeit little value here specifically, would be incremental toward other deliveries so I’ll probably reassess that one later on within the context of other projects.
What’s next?
The top priority item is to build Website Search! Squarespace has built-in Search, so that’s an option if it meets our needs. Spoiler alert: It doesn’t. I’ll be diving into the analysis that led to this decision and how to build a Programmable Search Engine next.
