Tearing down a product with ChatGPT.

What is TradingView?

TradingView is a web-based platform for traders and investors to analyze financial markets and make trading decisions. It offers real-time market data, charting tools, technical analysis features, and social trading capabilities, making it a one-stop shop for market research and analysis.

Someone is using aggressive keyboard backlighting while looking at TradingView.

What is the goal of this product teardown?

Doing research into competitors is a huge component of product management. Although TradingView isn’t a competitor of mine or my company, I was interested in using the ChatGPT guides to walk me through a comprehensive breakdown.

As a start, I asked for some recommendations on how I can use ChatGPT to help me with this decomposition:

To use ChatGPT for a UX product tear down of TradingView, follow these steps:

• Familiarize yourself with TradingView: Spend some time using the platform to understand its features, functionalities, and user flow. Take note of the different sections, pages, and components of the platform.
• Identify UX aspects to review: Determine the aspects of the user experience you want to focus on, such as navigation, layout, content, visual design, accessibility, and performance.
• Develop a list of questions and prompts for ChatGPT: Create a list of specific questions or prompts related to the UX aspects you want to evaluate. For example:
• What improvements can be made to the navigation menu on TradingView?
• How can the readability of text and charts be improved in TradingView?
• Are there any accessibility issues on TradingView that need to be addressed?
• Use ChatGPT to generate insights: Input your questions or prompts into ChatGPT and analyze the responses. Remember to use clear and concise language to ensure the AI understands your request. Keep in mind that ChatGPT’s knowledge is limited to information available until September 2021, so it may not be familiar with the most recent updates to TradingView.
• Compile the findings: Organize the insights generated by ChatGPT into a structured format, such as a list or a table. This will help you easily identify patterns and prioritize recommendations.
• Perform a manual review: Compare ChatGPT’s insights with your own observations from using the platform. Confirm the validity of the recommendations, and add any additional insights you may have.
• Prioritize recommendations: Rank the recommendations based on their potential impact on the user experience and the feasibility of implementation. Consider factors such as cost, effort, and the platform’s overall goals.
• Prepare a UX review report: Create a comprehensive report outlining your findings, recommendations, and prioritization. Include supporting evidence from ChatGPT’s insights and your own observations.

What are the main features of TradingView?

I copy/pasted into ChatGPT everything on TradingView’s Features webpage and received a quick bulleted list.

ChatGPT breakdown of TradingView features.

What are some evaluation questions and answers based on my observations?

I then had ChatGPT generate a list of potential UI/UX questions to walk through with their website and provide feedback. I collected everything in OneNote:

Using OneNote to keep track of teardown notes.

Framing the teardown.

Next, I wanted to ensure I was feeding the right information into ChatGPT related to teardowns, so I found a random one on Facebook on the internet. I reviewed the high-level important points and documented them accordingly in my OneNote.

Notes on a Facebook Product Teardown.

The last step: Putting it all together.

I took a 3-step approach to compile everything.

1. I processed all my data and questions through ChatGPT.
2. I organized the output in the desired order I would want in slides.
3. I installed an auto-slide generator called SlidesAI.io to test going from blog to slide.

Generating slides with SlidesAI.io.

I’ve never used this tool before, but it was one of the ones that came well-reviewed on Google. Since I wasn’t paying for it and there was a 2,500-character cap per deck, I had to create three different decks to get here. Kind of a pain, but we made it. I was able to copy/paste all of my notes, have some slides auto-generated, and had a good base to work with from a presentation perspective:

My product teardown deck was compiled in Google Slides.

Here’s a link to the presentation.

Final Thoughts

Since TradingView has been around since the end of the last GPT4 data set, a fair amount of information was available natively. It was easy to combine what was there, generate ideas on what would be needed to do the teardown, and then have TradingView reformat my ideas differently.

The automated slide tool I used was fine, but I’m sure there are way better options available — especially if you’re willing to give them some cheddar ($$$).

All in all, I spent about 30 minutes playing around with all of this, and the output was the start of a nice little deck I could use to discuss TradingView in depth.

Mike @ Product Party