ChatGPT’s Deep Research tool makes compiling reports on complex topics easy. But did you know that Microsoft Copilot has a similar feature?
So, which deep research tool should you use?
1
Getting Overviews of Complex Topics
One of the main reasons I use AI tools is to get a better overview of difficult-to-understand topics. So, this was naturally the first place I started with Copilot and ChatGPT. I fed them both the same prompt:
Please help me explain the science behind protein synthesis, and whether it’s legitimate or not.
Microsoft Copilot provided a comprehensive overview of the process in bullet points that were easy to understand. I also received a diagram that explained the process in more detail.
After confirming with ChatGPT that I wanted an overview, the tool assessed dozens of resources. Once it had done that, it gave me a comprehensive topic overview. While reading through, I saw that I also received multiple sources to double-check everything.
Both took a long time to produce results (which I expected from deeper research). Copilot took around 30 seconds, while ChatGPT took multiple minutes. Although Copilot was quicker and had a better layout, ChatGPT gave me more resources.
Winner: ChatGPT
While I use Perplexity for shopping recommendations, I wanted to see if Copilot or ChatGPT could do a better job. To do this, I prompted:
Please help me find the best shoes from across the web for male CrossFit athletes.
I deliberately asked a niche question to see whether these resources could extract deeper information. Copilot gave me a comprehensive list and then provided more information about these shoes in a table. Once it had done this, Copilot showed me some options I could purchase. There were some strange coding issues in the table, but I saw the needed information, so it didn’t make a difference.
Like before, ChatGPT asked me to be more specific. Since I have very specific requirements for my shoes, I quite liked this feature. I clarified which shoes were acceptable before the app started its next research journey. This time, I received research progress updates.
When I received my search results, ChatGPT went into huge detail. For each shoe, it provided images and also showed me the pros and cons. For each of these, it provided the source of this information. I would feel very confident if I needed to make a purchasing decision, though I would still ensure that I cross-referenced elsewhere for a better-rounded decision.
Admittedly, I’d have liked a table where I could see a direct comparison—like Copilot offered. Nonetheless, ChatGPT gets another point here because it went into so much detail.
Winner: ChatGPT
3
Planning a Trip
I sometimes use AI travel-planning apps to get general insights on a destination before exploring it further. My next experiment involved asking Copilot and ChatGPT to plan a trip for me. I had already used ChatGPT for trip planning, so I wanted to see if the Deep Research tool was any better.
Here’s the prompt I fed both tools:
Please help me plan a 4-week trip in New Zealand. We have a campervan and would like to see both the North and South Islands. we’ll be visiting in the summer.
To be honest, I was disappointed with Copilot here. The response I received was what ChatGPT’s ordinary version would’ve given me. Since ChatGPT asked me for more information, I then asked both apps the same follow-on prompt to see if Copilot could redeem itself:
I am happy to mix free and paid campsites; they need to have showers and cooking facilities, though. I am particularly interested in hiking and photography. I’m not interested in alcohol, but some good food places would be good. We also want to travel more slowly.
Though Copilot’s itinerary was more customized to my needs, I was still disappointed. It also gave me a pretty pointless graph at the end; a table would’ve been much more valuable. Though Copilot competes with ChatGPT in many areas, I was starting to think that the Deep Research tool was not one of those.
Once ChatGPT had finished its research, I saw a four-week itinerary based on the information I fed it. This was quite comprehensive, though I still would not use it as my only point of research. It was good to get an overview, though, and I could then watch YouTube videos or read other travel blogs for more unique perspectives.
Since I wouldn’t fully trust either tool in this category, it’s the first tie.
Winner: Tie
4
Sourcing Statistics
Accurate and up-to-date statistics are always difficult to find online, so I tried seeing how Copilot and ChatGPT would work in this sense. I used statistics that I knew existed to improve accuracy. Here’s the prompt that I gave these two tools:
Please show me the latest B2B content marketing statistics, and use stats from 2023 at the latest.
I liked Copilot’s layout; it gave me tables and bulleted highlights. The app mentioned the sources used, but I wish it had specifically given me links to them. Without sources, I have to manually search for these in Google, which adds more time to my research.
ChatGPT, like Copilot, gave me a breakdown of statistics in different categories. However, it also provided me with links to the sources it used.
One annoyance from both tools is that they took my “2023” point to mean only 2023. I wanted stats published between 2023 and 2025, but I clearly needed to be more specific here. With little to separate the two tools, I think that ChatGPT linking to the sources used means it must get the point here.
Winner: ChatGPT
5
Text Analysis
I next wanted to see how ChatGPT and Copilot performed when analyzing text. I used a book I’d already read so that I could verify for accuracy. Here’s the prompt I gave:
Please summarize the book “The Magic of Thinking Big” by David Schwartz.
Copilot quickly gave me a summary, which was pretty decent. But again, it was very similar to what ChatGPT would normally give me. So, I asked the tool to go deeper and summarize each chapter.
The app summarized each chapter but did so in just a few points; I wanted more detail than this.
ChatGPT gave me a very comprehensive book summary, and while I would still read the book, this was more of what I was looking for. I can imagine using this tool when considering buying a book, but I want to decide if reading it is worthwhile. Once again, I got all the resources from which the information was gathered. I liked how ChatGPT gave me more context, too.
Winner: ChatGPT
6
What’s the Overall Winner?
I’ve given ChatGPT the point in most categories because I think the tool is better. While I generally like Copilot, I was disappointed with its Deep Research tool. It didn’t seem too different from the basic version of ChatGPT.
ChatGPT took a lot of time to find the information I wanted and asked follow-up questions to ensure I was as specific as possible. I appreciated this, especially for complex topics that would’ve taken me hours to research. However, I would only use this feature for truly comprehensive research because it took so long to find the required information. For things like shopping, the general version is more than good enough.
On the other hand, I think Copilot’s Deep Research feature was better for overviews like product recommendations. It performed well with more complex topics, but I always cross-reference these with ChatGPT and Google. So, I think it’s better as a starting point than going into truly deep research.
