This article is adapted from the AI in CI Playbook. Grab your complementary copy here.

So, what work can AI actually do for us, and with us, to make us more efficient and effective in our competitive intelligence work?

First, we’ll walk through some of the competitive intelligence tasks accessible AI tools can meaningfully help you with. In Part 5 and beyond, we’ll walk you through your options for improving outputs to make them even more useful.

What can AI do in competitive intelligence work?

Here are some examples of common competitive intelligence-related tasks generative AI can help you perform.

  • Automation.
  • Data processing, summarization, querying, and analysis.
  • Data visualization and reporting.
  • Market trend analysis.
  • Competitor monitoring.
  • Sentiment analysis.
  • Product and pricing analysis.

Let’s dig into each one of these in some more detail. We’ll offer sample prompts for each use case, though we’ll come to discuss optimizing output in later sections.

1) Automation

Some AI tools integrate with the apps you use every day. It’s worth squirreling out an exhaustive list of these integrations, because they could save you a ton of time.

But there’s more.

Gone are the days when you’d have to learn Python, Lua, or some other scripting language to automate time-consuming tasks. ChatGPT can write code. What’s more, it’s really good at it.

Sample prompt:

“Write me a python script that scrapes our website at <> and reads all the news articles from the previous month, then emails a bullet point summary of each of these posts to <>. Write an accompanying crontab entry, with instructions, that schedules this python script to run at midnight on the first of every month.”

Because developers are such sticklers for writing detailed documentation, all the big programming languages have extensive written instructions on how to use them, just sitting there on the internet. While you could go and peruse them yourself, your favorite LLM was trained on that information. It knows it. And, since there are usually accepted methods for solving coding problems, and these are also extensively documented online, your favorite LLM knows those too. So you can ask it for help with tasks like this:

Sample prompt:

“Adopt the role of a software developer specializing in automation with 20 years of experience. I am a competitive intelligence analyst specializing in producing reports summarizing monthly developments in our market and offering points of view on what those mean for our business strategically.

I spend a lot of time writing in Google Docs. Give me 5 ideas for Google Apps Scripts that would improve my workflow and make me more efficient in my everyday work. Before responding, ask me 5 questions to improve your understanding of my requirements and therefore your final response.


“Now write scripts to achieve all the use cases you’ve outlined above.”

2) Data processing, summarization, Q&A, and analysis

An AI can ingest large amounts of data, of all kinds, extremely quickly.

When you’re pushed for time, and you have a new PDF report on a new competitor product or feature to sift through, upload it to your AI tool of choice.

Ask the AI to summarize the data for a quick overview, or get even more targeted by asking it specific questions about the document.

Note: Response accuracy tends to decrease as the context window (length of input) increases. However, GPT-4 made huge progress versus GPT-3.5 with mitigating these performance dips with longer context windows, and we can expect future models to improve further. So, while output won’t always be perfect, it’s more than usable for now, and output will only improve as time goes on.

“It mostly comes down to text analysis. So we're doing it on news, and PDF analysis. Every day, there are dozens of news articles that come out. A lot of those are noise, but some are really important. There are tools now that can condense those articles down to the key takeaways. And you can get a sense really quickly of whether an article is worth digging into more deeply or not.
“If I'm looking at a 50-page PDF, I can get information out of the PDF that I don't have to dig for. And we use another tool that analyzes reviews from customer review websites like G2, TrustRadius, Gartner, Peer Insights, and is able to piece together trends based on the keywords using AI. So those are the use cases, and they're all things that are major time savers for us.” - Ben Hoffman, Senior Competitive Intelligence Manager, Adobe

Sample prompt:

“Assume the role of an industry analyst specializing in competitive trends. I am a product marketing manager who owns competitive intelligence at a small startup, tasked with putting together a presentation on our competitive landscape for the upcoming company meeting. Using <source x>, generate a report summarizing the latest trends in <specific technology or industry sector> for the last quarter, highlighting major innovations, consumer behaviors, and emerging competitors. Pay particular attention to our main competitors <list of competitors>.”

3) Data visualization and reporting

GPTs aren’t limited just to text data.

ChatGPT’s Data Analyst is a Custom GPT that specializes in data handling. Though they share the same training data, thanks to detailed (pre-configured) custom instructions, the Data Analyst’s outputs will often be better (for data-based inputs) than those of vanilla GPT-4.

Note: Depending on how your business chooses to handle data privacy, either deselect the Chat history & training option in your Data controls settings, or run your LLM on a private server, before uploading any proprietary data for the AI to analyze, like your sales data, for example.

Ask the AI to identify correlations and patterns in the data for you, or to summarize general findings. You’re only really limited here by your ability to ask the right questions, so think about what you’d like to learn from the dataset if you were analyzing it yourself, and ask the AI about that.

Another excellent use case here is to ask the AI to visualize the data for you. Most often, it’ll lean on its knowledge of the Python libraries Pandas and Matplotlib to plot your data. You can even ask it to combine various datasets, aggregate data from particular columns, and plot their outputs.

Sample prompt:

“From <source X>, plot media sales cycle duration for each rep against their enablement content access frequency.”
An AI will often use its knowledge of libraries like Pandas and Matplotlib to work with and plot numeric data.

4) Market and competitor monitoring

If you’re able to provide real-time, ongoing access to various data sources, like news articles, financial reports, social media, and industry publications, an AI will be able to parse this data and scrutinize it for emerging trends.

The same holds true for markets as it does for information from and about your competitors. AI tools can scan and analyze your competitors’ presence online across social media channels, websites, and customer feedback platforms like G2, to name a few examples.

It can alert you when there’s a new development, a shift in sentiment, or a new product launch, and provide automated summaries. Given AI’s chatbot capabilities, you can even engage it in a Q&A and ask it pointed questions about its findings.

Sample prompt:

“Assume the role of an expert in enterprise marketing and competitive intelligence. I’m a CI analyst working in healthcare, interested in learning more about a specific competitor. Perform a competitive analysis of <corporation name>, whose website is located at <url>. Include, as part of your competitive analysis, a SWOT analysis of this organization versus their top 5 competitors, only for <service A>, and none of their other services. Don’t include large companies with more than <$x million> in gross annual revenue in your choice of competitors.”

“AI can help with market research, collecting the information, providing value, summarization, classification, and really sorting the information for you to use. For positioning and messaging, the system can help you identify those key differences around which you may create positions or messages. And then, for internal communications, AI can save you hours when it comes to creating a particular deliverable. “ - Ed Allison, Founder, CompeteIQ

5) Sentiment Analysis

We don’t need to tell you competitive intelligence is as much about your customers and your target audience as it is about your competitors.

Thanks to its natural language processing (NLP) capabilities, GPTs are particularly useful for evaluating customer opinions and evaluating their general sentiment towards you or your competitors from various sources.

Combine findings from sentiment analysis on particular competitors with well-established (well documented online means well understood by the LLM) frameworks like SWOT analysis to identify weaknesses, opportunities, and threats posed by how your target audience is feeling about those competitors.

Sample prompt:

“I am a competitive intelligence professional looking to better understand how our products compare with those of our competitors. Judge the strength of customer sentiment towards <product X> based on all the text data available from <source Y>. Make each individual mention a separate data point, and assign a numeric value to its sentiment on a scale from negative 5 to positive 5. Use Python libraries to plot these values to visualize overall customer sentiment towards our product.”

“A transcript from a win/loss interview might be 24 pages. If I want to know what the customer said about the competitor’s scalability, the AI would be able to pull out a sampling of different mentions of that. That's a great use case, and a major time saver. And once you have the prompt for that, you can reuse it to save even more time.” - Ben Hoffman, Senior Competitive Intelligence Manager, Adobe

6) Product and Pricing Analysis

Like any competitor strategy, a pricing strategy is best understood framed through the context of various sources. Pricing data, promotional materials, even an increase in the frequency of promotional posts on social media, are all data points that are difficult to track or make sense of manually, that an AI can make short work of.

Your GPT of choice can aggregate such data, and analyze it for insights, much more quickly, efficiently, and cost effectively than any human could.

In short, AI can automate repetitive tasks, process and analyze data faster than you’d ever be able to, and run advanced analytical and statistical analyses on this data. Already, but especially when fine-tuned by a professional, AI can even offer strategic insights and recommendations that will only improve with time, and as more advanced LLMs become available.

Your copy of the AI in CI Playbook

If you found these sample prompts and insights useful, you have to check the AI in CI Playbook.

Bursting with no-nonsense info to bring you up to speed with AI for CI in no-time, the AI in CI Playbook sidesteps hype and speculation to bring you the most practical guide on using AI for your competitive intelligence work possible.

Stop feeling frustrated with AI. Start prompting like a pro and discover how to use context the right way so you can finally receive outputs that save you time and maximize your impact. 🦾

Click the banner below to grab your complementary copy. 👇