Alright, let’s say you’re staring down the barrel of a quarterly competitive review. Your boss wants a “summary of the landscape,” which really means “tell me what our competitors are saying so we can sound smarter without actually innovating.” You’ve decided, perhaps against your better judgment, to let ChatGPT do the heavy lifting. Having migrated data between more CRMs than I’ve had hot dinners, I can tell you this: garbage in, garbage out, but the AI era gives us a fascinating new flavor of garbage.
Here’s my contrarian take: don’t just ask it to “summarize competitor X.” That’s how you get a bland, possibly hallucinated marketing brochure. You need to structure the interrogation like you’re building a sales process in a new CRM—define the stages, the data points, and the exit criteria.
My methodology, born from years of watching shiny new platforms turn into legacy headaches:
* **Source Material Curation is Key:** Don’t feed it raw Google search results. Manually gather specific inputs: their latest three press releases, the “Features” page, two recent blog posts about their vision, and maybe a YouTube transcript of a founder interview. This is your data migration. If you migrate junk, your new CRM is useless from day one. Same principle applies.
* **The Comparative Grid Prompt:** Force it into a structure. Start with something like: “Act as a revenue operations analyst. Using only the text I provide from [Competitor A] and [Competitor B], populate the following table. Leave cells blank if the provided text does not explicitly support an answer.” Then define columns: Claimed Core Strength, Target Customer Profile, Recent Strategic Shift, Implied Weakness (from what they emphasize), Pricing Model Clues.
* **Interrogation for “What’s Not Said”:** This is where the skepticism pays off. Follow up with: “Reviewing the provided data, what specific features or market segments are conspicuously absent from Competitor A’s communications when contrasted with Competitor B? List three, with the source material gap noted.” This mimics checking integration gaps when you’re planning a platform switch.
* **Tone and Rhetorical Analysis:** Ask it: “Analyze the adjectives and verbs used in Competitor C’s press releases. How do they frame their market position: as a disruptor, a reliable partner, a luxury, etc.? Provide the top five recurring terms with examples.” This tells you what they’re trying to sell, which is often more revealing than the feature list.
My biggest pitfall warning? ChatGPT will default to consensus and generalization. It hates stating that something is *missing*. You must corner it into acting as a diff tool. I’ve seen it confidently attribute features from Salesforce to HubSpot in a summary because, at a high level, they’re both “CRMs.” The same will happen if you’re comparing SaaS tools. It will blend them into a homogeneous soup unless you enforce strict, source-bound constraints.
Treat it like a terribly literal, overeager intern. You wouldn’t let an intern write a competitive analysis without a brutally clear template and primary sources. The value isn’t in the AI’s “insight”; it’s in how you force the structure, which ultimately makes *you* think about what variables actually matter. The summary it produces is just a byproduct. The real output is the framework you build to get it.