AI Marketing Reverse Engineering
Turn public brand evidence into an actionable marketing playbook with AI
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 96%
The method uses an AI-enabled browser to analyze a company’s public marketing surfaces and turn observable evidence into a working strategic hypothesis. Begin with a website, product page, social account, or campaign, then ask the agent to identify the audience, positioning, offer, messaging, funnel, conversion architecture, distribution loops, creative process, and technology stack. Follow the initial diagnosis with increasingly tactical questions about tools, workflows, tracking, and implementation. Because some conclusions are inferred, compare evidence from the company’s pages, third-party sources, source code, tracking scripts, and vendor materials. Finally, translate the research into experiments suited to your own audience and save the complete prompt as a reusable browser shortcut for future analyses.
Origin
Extracted from Marketing Against the Grain after the host used Perplexity Comet to investigate why Gruns repeatedly served highly personalized Instagram ads and how its team appeared to produce them.
Core principles
- 01Study observable marketing rather than guessing internally
- 02Use multiple public sources to validate the initial diagnosis
- 03Move from strategic patterns to specific tools and tactics
- 04Translate findings into experiments instead of copying blindly
- 05Save repeatable prompts to make research faster
How to run it
- 1
Select the marketing surface
Choose a competitor or admired brand and open the most relevant public surface, such as its homepage, product page, Instagram account, landing page, or campaign.
Pro tip Start with a surface directly connected to the tactic you want to understand.
Watch out A broad homepage may hide details that are clearer on a product or campaign page.
- 2
Request a strategic diagnosis
Ask the browser agent to explain how the company’s marketing works. Have it identify the target audience, positioning, offer mechanics, messaging, creative approach, funnel, and growth loops.
Pro tip Tell the agent why the company caught your attention so it can focus its investigation.
Watch out Treat the diagnosis as a hypothesis rather than confirmed inside knowledge.
- 3
Drill into tools and execution
Follow up by asking which platforms, attribution systems, creative tools, influencer systems, and commerce technologies appear to support the strategy.
Pro tip Ask how each identified tool is likely used instead of accepting a simple software list.
Watch out Vendor pages may exaggerate their role in a customer’s operating process.
- 4
Expand and cross-check the evidence
Let the agent examine third-party resources, vendor materials, source code, tracking code, and related pages. Compare these sources to separate observable facts from plausible inferences.
Pro tip Repeat the analysis across several pages when one page provides limited evidence.
Watch out Do not present inferred tools or workflows as verified facts.
- 5
Translate findings into experiments
Identify what can be adapted to your audience, resources, and positioning. Create experiments that improve on competitors’ messaging, creative, or conversion architecture rather than merely reproducing it.
Pro tip Prioritize experiments with a clear metric and a short feedback cycle.
Watch out Direct imitation can erase differentiation and import tactics that do not fit your market.
- 6
Operationalize the research
Save the full analysis prompt as a shortcut and reuse it across competitors, aspirational brands, and individual campaigns. Synthesize recurring patterns into your own marketing point of view.
Pro tip Include your role and business context in the saved prompt so recommendations remain relevant.
Watch out A reusable prompt still needs periodic refinement as your strategy changes.
In the wild
After repeatedly seeing personalized Gruns ads, the host opened the company website in Perplexity Comet. The assistant identified persona-driven funnels, dedicated landing pages, coordinated creative testing, influencer content, and rapid elimination of unsuccessful creative directions. Follow-up questions surfaced likely tools for attribution, ad creation, influencer management, and commerce.
→ The investigation converted a vague impression about effective ads into a tactical hypothesis about the company’s creative system and technology stack.
A B2B founder analyzes three competitor product pages with the saved prompt. The research reveals that every competitor routes visitors through a long demo-request flow. The founder creates a shorter self-qualification path, measures completion and sales conversion, and retains the version that reduces abandonment without lowering lead quality.
→ Competitive analysis becomes a measurable conversion experiment rather than a collection of copied page ideas.
Common mistakes
Treating inference as certainty
The agent may make a best guess from public evidence. Distinguish observed facts from inferred workflows and verify important claims through multiple sources.
Stopping at the summary
A high-level strategic diagnosis is only the beginning. Ask follow-up questions about specific tools, workflows, implementation, and measurement.
Copying without adaptation
A tactic that works for an admired brand may not fit your audience, budget, positioning, or funnel. Convert findings into contextualized experiments instead of replicas.
Is it for you?
Best for
It is best for founders and marketers who need rapid competitive intelligence, positioning inputs, or campaign ideas.
Not ideal for
It is not ideal when a company has little public marketing evidence or when verified internal operational details are required.
From the transcript
“it uses the website as kind of the context to answer those questions.”
“you can quickly and easily reverse engineer anybody's marketing strategy”
“you can just go and do this for a bunch of relevant companies in your industry and get a deep understanding very quickly of how…”
From the episode
Using Perplexity AI to Reverse Engineer a $500M Marketing Strategy