We embraced generative AI to rapidly develop a comprehensive platform category section, showcasing the power CrowdStrike Falcon® Cloud Security, just in time for a major industry event.
We were tasked with an ambitious goal: relaunch all of our cloud security pages — 26 individual pages — in just two weeks, ahead of a fast-approaching cloud security event.
Our mission was to deliver a high-quality microsite showcasing Falcon Cloud Security's unique strengths while pioneering the use of generative AI in our content creation.
We aimed to engage IT professionals, cybersecurity experts, and potential customers actively seeking cutting-edge cloud security solutions.
Check out the live pages and witness firsthand how we leveraged AI and innovative content design to inform and inspire our target audience.
We mapped out a meticulous strategy, aligning key objectives, target audience needs, and core messaging with the capabilities of generative AI.
We developed and refined prompt templates, tailoring them to capture specific product details and generate compelling content for each of the 26 pages.
We ran the prompts, iterated on AI-generated responses, and carefully filled out templates for each page. The content underwent rigorous review by subject matter experts to ensure accuracy.
After thorough testing, we deployed the microsite and executed a coordinated launch strategy to maximize visibility — just in time for the event.
increase in visitor sign-up conversion
increase in overall organic traffic
increase in non-paid conversion
increase in all non-paid cloud security-related conversion
Strategic planning is paramount. Using ChatGPT effectively requires thoughtful planning. Lazy or mindless use results in poor answers. Throughout the project, we found we were able to dramatically improve our results by spending time articulating our goals and steps. It’s tempting to think that you can use ChatGPT without much effort, but this approach only leads to weak outputs.
Precision fuels AI. During the project, lazy word choice often led to flawed responses, so we had to carefully think through each piece of instruction. The best analogy is to think like a programmer, but in English. Clear, precise prompts increase the likelihood of strong results, even for someone unfamiliar with the domain.
We found that breaking down content creation into smaller, manageable chunks allowed for more focused and accurate AI output. Attempting to produce full-page content in one prompt often resulted in generalized content or hallucinations. We found that feeding it component by component produced stronger, more specific results and allowed for precise iteration.
Our attempts to get ChatGPT to write page content from scratch typically failed, resulting in general, non-specific writing and inaccuracies. However, providing the AI with strong messaging points and specific differentiators from our corporate resources significantly enhanced the output. The more expertise we infused into the prompts, the more precise and unique the generated content became.