Why Your AI Strategy Needs Your Own Data—Not Just Better Prompts

There’s a pattern playing out in marketing strategy meetings everywhere: someone opens an AI tool, types a prompt, and gets back an answer that sounds polished and confident—and nearly identical to what every competitor is getting from the same tool that morning.

Forbes contributor and AI practitioner Lisa Peyton calls this “AI flattening strategy.” The fix, she argues, isn’t better prompts. It’s better inputs.

Key Takeaways

• Generic prompts produce consensus, not strategy. Recent Harvard Business Review research found that leading LLMs recommend similar strategic directions regardless of context, and adding rich context shifts the pattern by only about 11%. The implication: the AI isn’t the differentiator—your inputs are.
• Named examples orient AI more effectively than abstract instructions. Asking AI to “be more inspiring” gives it a mood. Asking it to “apply Nike’s empowerment framework to a risk-averse B2B buyer justifying budget internally” gives it a pattern, an audience, and a constraint. Specificity is the competitive asset.
• Your primary research is irreplaceable input. Customer interview themes, sales call objections, brand audit findings, and campaign performance data are things no competitor has access to. Fed into AI, these inputs transform output from “statistically familiar” to “genuinely strategic.”

The article makes a clean argument: AI amplifies whatever thinking you feed it. The organizations that will win are those that bring their proprietary knowledge into the system—not those chasing the best off-the-shelf prompt.

Read the full article on Forbes