The art of influence: The single most important skill left that AI can't replace | Jessica Fain

Jessica Fain spent years watching product managers pitch ideas at Slack. Some proposals sparked funding and momentum. Others — even brilliant ones — died on the vine. The difference wasn't the quality of the idea. It was the PM's understanding of how executives actually make decisions. Most PMs center themselves: their user research, their roadmap, their domain expertise. They forget that the executive sitting across from them just came from a finance meeting, a legal fire, and a people problem. The exec hasn't thought about your project since your last review. They may not have gone to the bathroom today. You have to help them get into the right headset — fast. Fain's breakthrough came when she became chief of staff to April Underwood and later Tamar Yehoshua. She saw firsthand that execs want to be successful, too. They're optimizing for a global maximum, not your local one. If you can align your work with their incentives, speak their language, and make them feel like co-creators rather than gatekeepers, you unlock momentum.

Fain's team once brought Rachel Woolen, Webflow's CPO, a single solution to a strategic problem. The review didn't go well. Woolen wanted to see the permutations: what else did you consider, and why did you rule it out? Within two days, the team returned with a doc showing all the options, the trade-offs, and the reasoning. Woolen immediately saw why their chosen path made sense. The lesson: don't bury your exec in process, but do show that you're not missing something obvious. Fain recommends a «Goldilocks» approach — present three options, with the middle one being your recommendation. Put the full analysis in an appendix. If they glaze over, you haven't lost them. If they dig in, you're ready. At Slack, the design team would return from a review with Stewart Butterfield with «Stewart plus two more» — the exact design he asked for, plus two alternatives the team felt good about. It created space for debate without looking like they ignored his feedback.

One of the fastest ways to build trust with a leader is to kill your own work. Fain calls this a «very senior way of thinking.» If you can say «I know I spent months on this, but the data says we should stop» — and then actually stop — you signal that you're optimizing for the company, not your ego. It also frees up the exec's bandwidth. They don't have to be the bad guy. You've done the hard work of deprioritization for them. This is especially powerful in early-stage pitches. Use the «shrink the change» principle from the book Switch: make the ask so small that it feels like an experiment, not a bet. A one-week proof of concept is easier to approve than a six-month roadmap. Once you show traction, you can expand. At Slack, Fain's team ran a two-week «customer love sprint» — engineers could work on anything that improved the user experience, as long as they shipped. The result: 65 improvements and a renewed culture of craft. It aligned with leadership's belief that polish was a differentiation, and it built momentum for more ambitious bets later.

Fain argues that we're all «directors of work» now. If you have an army of agents or a hundred new AI colleagues, how do you onboard them? What do they need to know about your product philosophy, your definition of success, your metrics, your taste? This is where codifying your beliefs becomes critical. Feed your agents transcripts of past product reviews. Train a GPT on your PRD history. Use tools like Claude Projects to capture your framework for what good looks like. But also set guard rails. Where do you need to be involved? Where do you uniquely have judgment or taste that an agent can't replicate? Fain's advice: use AI as a «really smart colleague that never gets irritated by your questions.» Ask it to poke holes in your ideas, simulate objections, and help you prep for the room. But don't abdicate the decisions that define your leadership.