giga
My Role
Team
Time
Overview
Search Results Experience
The search results page eases users into the details and embeds follow-up features as they scroll down, only feeding them granular information if desired / needed. A key point of the search UX was maximizing credibility by highlighting source citations and incorporating opportunities to deep dive into discussions.
1.0
Anatomy of basic search results page.
ANNOTATED IMAGE
Giga Chrome Extension
Recognizing that most users, by default, go to Google / Safari first to search something – the extension incorporates the Giga search experience into an existing user flow to mold around existing user behaviors.
1.1
Giga Chrome Extension Demo
VIDEO LOOP
Micro-Expert Interactions
Giga aims to create an online knowledge "marketplace" by bringing in experts of niche topics from platforms like Reddit. Users can seek 1:1 advice from these micro-experts and feel confident initiating these chats with Giga's profile analysis feature and screening flows.
1.2
Profile page of a micro-expert with credibility analysis.
IMAGE
1.3
Outgoing chat requests of various stages.
IMAGE
The internet is filled with noise!
Page ranking algorithms
of search engines are easily gamed.
This shorthand method of determining online reputation means people aren't actually getting the best, most up-to-date answers / results.
People are increasingly relying on platforms like Reddit, Tiktok, Beli, Yelp, or Twitter for more reliable information on subjective topics.
Or they just add "Reddit" to the end of Google searches.
1.0
User pattern example.
IMAGE
A need to centralize real answers from real people.
User lab findings & community feedback helped identify the overarching goal: to make people the atomic unit of Giga's knowledge ecosystem rather than website domains by scraping billions of discussions from online knowledge communities – starting with Reddit in the initial launch.
1.1
Press & community feedback.
IMAGE
Top priorities
to guide design decisions:
Arrive at the point quickly & accurately
Summary modes, TLDRs, intuitive search flow, etc.
Find credible micro-experts
Profile analysis, 1:1 screenings, bot-detectors, etc.
Adopting new online searching habits
Giga Chrome extension, follow-up opportunities, etc.
Turning internet "rabbit holes" into a productive thing.
Identifying pain points
in existing user journies.
It was important to recognize why & at what point people get lost in irrelevant information when searching online.
2.1
User Journey Map - Identifying existing pain points
IMAGE
Uncertainty around credibility of sources and the relevancy of search results
[1.3], [3.1]
Increased noise / distraction from main search / learning objective
[2.1], [3.3]
Lacking proper redirection and increased chance of falling down irrelevant rabbit hole
[2.3], [4.1]
Reworking the ideal search UX.
How might we help users feel confident in their findings and stay on track while encouraging deep dives into interesting / engrossing topics?
2.2
User Journey Map - Pinpoint in-line search opportunities
IMAGE
Concrete tools for evaluating credibility, including upvote counts, user profiles, and source-first result methods.
[1.2], [3.1]
Fine tuning of results and summarized opinions to arrive at the point sooner.
[2.2], [3.3]
Capture insights and guide the next follow-up query or learning opportunity.
[1.3], [2.1], [4.2]
Scoping in-line tools into the search flow to guide curiosity in the right direction.
By embedding tools seamlessly into the overall UX, the flow guides users down purposeful rabbit holes, steering them from distractions towards productive, focused research.
2.3
In-line follow up search to lead user down a insightful "rabbit hole".
2.4
Summary mode option for source threads so users can save time.
2.5
Suggested micro-experts for queries that require personalized advice.
2.6
Source previews to help users assess credibility right off the bat.
It's hard to change people's habits.
Users who were active at first
didn't seem to stick around for long.
User lab participants largely attributed this to a habit of pulling out Google first when searching a query, forgetting to use Giga instead (even if they prefer the search results on Giga).
Turning weaknesses into opportunities
How might we harness users' Google search habits to seamlessly guide them to Giga with out being intrusive?
3.1
Problem Framing / Brainstorm Chart
In the same way that Paypal Honey's Chrome extension allows users to decide when to apply & check for coupons…. The hypothesis was that presenting Giga as a compelling choice through a Chrome extension would give users the autonomy to choose when to use Giga search results over Google search results.
Extension popup testing: how much info peaks interest?
I focused my testing on varying levels of detail in the initial popup for the Chrome extension. My goal was to determine how much information – and in what format – effectively encourages users to click, rather than dismissing it as just another annoying popup.
3.2
Select iterations of more condensed versions of extension popup.
IMAGE
"Not enough info to know what I'm looking at"
This version doesn’t show compelling enough information to prompt further exploration. Users expressed they would likely swipe away and continue their Google search instead because they aren't sure what the popup is.
3.3
Select iterations of detailed extension popups I tested
IMAGE
"Information overload taking up too much screen real estate upon first popup"
When the initial popup has too much detail or is too large, users are overwhelmed. This felt too intrusive to the original search experience to the point where some said they would uninstall because it would get too annoying.
3.4
Final Iteration of initial extension popup
IMAGE
"Just enough to keep me interested"
Users expressed that if they saw brief source previews that seemed interesting and relevant, they would be more compelled to explore Giga and choose its search results over that of Googles.
Mapping out the progressive disclosure of the extension experience.
Ultimately, it was the most effective for the extension experience to be multi-step. From user testing, we found that sources are ultimately what drive user decision making. When a user searches, the popup displays just enough information to peak their interest, then the user can click in to see more, and finally they can click into the website for full immersion into Giga search results and features.
3.5
Full extension experience mapped out.
IMAGE
The extension boosted the sites 1 month user retention to a consistent ~35%.
We gained about 7000 new registered users over a couple months, with at least 200-300 coming onto the site daily. This project was a strategic one that helped us test PMF, boost retention, and adopt users who were willing to test new features and give us feedback. The extension remains a core piece of Giga's engagement today.
Increasing user adoption & confidence in a monetized, micro-expert chatting flow.
The chatting flow introduced Giga users to potential micro-experts matches from the search results page and allowed then to initiate chat requests. While there was a high click rate into the micro-expert profiles, indicating some level of interest… there was a low execution rate of 1:1 chats. My PM and I worked together closely to analyze and identify the falloff point of the user experience.
4.0
Identifying what part of the chatting experience users were falling off.
IMAGE
After processing Mixpanel sessions and click rate analytics, we found that most users decided to quit out at two obvious points:
Drop Off Point #1
Micro-expert profile page
Users don't feel comfortable initiating a chat with a random person purely based off of their activity & interests. Redditors specifically mentioned their fear of running into bots, marketers, and suspicious accounts.
Drop Off Point #2
Payment / tip stage
Users don't feel comfortable tipping prior to a chat, even if it is only a few bucks, and they want to feel reassured that their money will be worth it.
Competitive analysis helped familiarize me with various micro-expert systems that have worked.
After identifying the drop off points, I dove into competitive research of other platforms like Naver Expert, Stack Overflow, Chegg, Intro.Co, and many more to understand why and how their systems have worked and what methods they use to instill a sense of trust.
4.1
My "messy thinking" Figjam file showcasing some of my competitive research and UX analysis.
IMAGE
Synthesizing platforms
by expert-matching systems
By grouping and affinity mapping the platforms in the follow categories, I was able to narrow down what type of system aligns with Giga the most and identify tools to test in the micro-expert chatting flow:
Experts Offering Help
Experts advertise themselves and try to get people to book sessions to learn from them.
Ask the Audience
Learners seek public opinion, and micro-experts who are passionate about helping contribute to the discussion.
Learners Seeking Experts
Learners discover experts either through a feed or list of suggestions or through personal research paths.
Idea Takeaway #1:
Add a screening phase
Allow for users to screen one or multiple micro-experts to gauge credibility and match before committing to full 1:1 session.
Idea Takeaway #2:
Present payment as a tip
Frame the chats as "coffee chats" instead of payments. This is mainly copy-related, but will change to connotation of the feature.
Idea Takeaway #3:
Profile analysis
Automate profile credibility based on common heuristics to eliminate bots / marketers from start.
Adding a screening step and a more transparent system for analyzing the credibility of a potential micro-expert.
The option to "fan out" questions to multiple micro-experts and include a free screening period gave users more comparison points and decision-making power to confidently choose whether to start a 1:1 session or not.
4.2
The reworked user flow.
IMAGE
1.1
Screening steps, semi-prototyped
VIDEO LOOP
Reframing the monetized system with strategic copy changes.
We also strategically updated the copy to make the monetized process feel less transactional and more like a tip or a gesture of appreciation. In user lab sessions, participants shared that they’d be more willing to engage in a 1:1 session—even with the risk of losing $5–$20—if it felt more like a supportive gesture, similar to the experience on platforms like www.buymeacoffe.com).
4.2
Reframing the monetized system as a form of gratitude.
IMAGE
Adding these extra steps increased how many users completed an full micro-expert chat, from initiating to payment.
When we first released this feature, we observed several users show interest in initiating micro-expert chats, but most of them never fully went through with it. Once we added the extra screening and profile analysis steps, twice as many users were fulfilling micro-expert "coffee chats" and were satisfied with their interactions. It appears the feature has been taken down since, so I assume the engagement levels were not high enough to maintain the complex chat and payment systems.
Scaling a design system that helped me to iterate & test concepts quicker.
The (contract) designer that came before me started the design system by selecting core style tokens and branding assets, which I branched off of to create a more robust design system. Adding flexible variants helped me to design faster and mock concepts up in higher fidelity during our tight feature sprints.
5.0
Basic design system tokens and header elements.
IMAGE
5.1
Snapshot of some of they key components and variants.
IMAGE
Creating assets for landing page test & more.
For each feature launch / concept test, I was also responsible for creating landing pages, marketing assets, and more to advertise Giga's features, familiarize users with how things work, or boost SEO engagement. These were fun side projects for me to explore my visual + motion design skills!
6.0
Example of landing page post-extension install that shows users how to use Giga on Chrome.
IMAGE
Positive reactions from users with feedback for growth:
Some of the features we launched were successful, while others helped us identify and learn from incorrect hypotheses. Additionally, a few features I worked on didn’t make it to launch, offering valuable insights for future iterations. So far, we have received overwhelmingly positive responses on the core extension and site search experiences, but the supporting features require further testing.
7.0
Recent Google reviews snapshot
IMAGE
Learnings
and reflections:
Be comfortable with necessary pivots.
Especially in earlier product stages, it's important to have humility and acknowledgement that some of your core hypotheses may be wrong. I learned from incorrect assumptions to make the product better and embraced when we had to change scope.
Adapt to user patterns, don't try to change them.
It's easy to 'revamp' a user journey and make it better on paper, but it's usually not that simple to get people to adapt to a new flow. Stay realistic and take behavioral constraints seriously.
Stay focused on big-picture business goals.
It was helpful to frame my projects from both a user-impact perspective as well as from a business angle. This was crucial in Giga's initial growth of its user base and improving retention.