I’m so excited to have an old friend, John McGuire, joining us to talk about his career as a Search Product Manager. John is one of those people I love to brainstorm with, and he always has a great story about search or about life to share. I’m happy he’s agreed to stop by and answer some questions about his great career in search.
- Can you tell us about your current role as Global Search Product Manager at GitLab? What does that mean. What do you do on a normal day?
Thank you Chris! I am a big fan of Searchnews and it’s a true honor to get an opportunity to participate.
A year ago I got the opportunity to join Gitlab. Gitlab is a one of a kind company. It is 100% remote work and hires in many countries. GitLab is a single product that offers everything a CI/CD organization need. My part is managing the roadmap and feature development for the Search Platform. We call it Global search because it searches the entire product even across multiple code repositories.
- 2. How did you get started in search and Product Management? It’s such a great career choice because so many companies are looking for Product Managers to own search.
While at Dell Technologies, I created my own position by proving the value that can be created by investing a very small amount in managing search. What I figured out was that 80% of our content is viewed less than once per year. At the same time we were constantly getting feedback from customers that content is difficult to find. Michael Dell had often discussed his frustration with finding things on our site. As I started looking into ways to improve this experience I connected with an IT team that was looking for business guidance for how to set up FAST search. We created many years of successful improvements together.
- 3. What are the challenges that a Search Product Manager faces working in e-commerce? I ask because the role is a bridge between engineering, product data, merchandisers, and just about everyone else.
The biggest challenge in e-commerce search is cleaning and structuring content to make meaningful facets. Facets are worth all the work, but content is not static so it’s always been an ongoing effort.
2nd would be getting the design/info architecture figured out. Finding the best information to display in a result while not adding so much it becomes confusing. If you have a facet or sort option you need to include the meta data referenced in the result. As you start adding more facets and sort options, you can start stretching the line of having a confusing result.
The hardest thing was searching by store availability.
- 4. We’ve spent many hours discussing search relevance, and I’d like to ask you how you define and measure search relevance?
Relevance is the judgment of the strength of a relationship to what was asked for and what was returned as a result.
I think it is much easier to judge the extremes. For measurements, I like to look at result clicks. I then split this down by keyword and look at total searches to the keyword. This builds a ratio. Searches/result clicks.
This doesn’t account for accuracy. Calibration for accuracy takes a more detailed approach to establish and test for known constants.
- 5. Lastly, what interesting innovations are you seeing these days in the world of e-commerce search, and search in general? As a follow on, do you think personalization is key to success or is it overrated?
We are going to see 10 years of innovation in 2 years. Everything is changing right now. Product customization is going to change how people shop and how comparing products happens.
Imagine shopping for a shirt and instead of having 8 different prints to choose from, there is a pallet wall like when you’re shopping for paint.
E-commerce search will need to become more visual, it will need to help choose options available for a product and not just between products.
Personalization will take on a completely different meaning.There will be challenges with the increasing privacy laws and security focuses.