Spraying GenAI Everywhere? Try This First

Everyone’s asking: How can we use GenAI to grow faster? Sell more? Keep customers longer?

Here’s the thing—rolling out GenAI across your business isn’t just about flipping a switch. It’s messy. There’s uncertainty. You’ll try things that don’t work. You’ll learn. You’ll adjust.

In other words​​, it’s precisely the kind of challenge that benefits from an agile, experimental mindset—not just for building products, but for building how your company operates.

I see too many companies spraying GenAI around like fertilizer, hoping something sprouts. A better approach? Slow down.

First, take a hard look at the bottlenecks in your business—the critical processes that waste time, drain energy, or block growth.

Those are the areas where improvements through AI can move the needle.

“An hour lost at a bottleneck is an hour out of the entire system__. An hour saved at a non-bottleneck is worthless__.” Eliyahu Goldratt, the creator of the Theory of Constraints (TOC)

But even when focusing on the bottleneck, don’t rush to plug in a shiny AI tool.

First, implement some basic process improvements. Tighten things up. Clarify how the work should flow. Once you’ve got something stable and sensible, then you can ask, “How might GenAI help us do this faster or better?”

For example, maybe you notice your sales team spends hours each week copying data between systems—CRM, spreadsheets, proposals. It’s tempting to jump straight to “Can GenAI automate this?”

But first, ask: Do we really need all these handoffs? Maybe you can rework your process so one system is the single source of truth. Maybe you can delegate approvals.

Try that first. Let people live with it for a few weeks. See what changes. Once you’ve simplified, then bring in GenAI—maybe to auto-draft proposals or summarize meeting notes.

Or take customer support. Perhaps your team wants GenAI to respond to all incoming emails. But if your help center articles are outdated and your processes are inconsistent, the AI will just give faster wrong answers.

GenAI can be powerful. But it’s smarter to aim it at the right places, get ready for it, and steer it using old-fashioned human effort,