I joined G2 during the pandemic while working completely remotely, which made me realize just how important it is to feel connected, supported, and engaged, even from a distance. While I wasn’t in an office setting, the tools and strategies G2’s team used played a big role in helping me feel part of the company culture right from the start.

If you’ve sat through even one SaaS demo in the last year, you’ve heard the pitch.AI sales assistants will automate outreach, personalize every interaction, prioritize your best leads, and unlock 30%+ productivity gains. And to be fair, that promise isn’t entirely invented. McKinsey estimates that AI could automate up to a fifth of sales activities and significantly reshape how revenue teams operate. The subtext is even bolder: less grunt work, more pipeline, and maybe quietly fewer people needed to do it.It’s a compelling story. And for a while, it sounded almost too good to question. But as a market research analyst, you learn quickly that the truth of any category doesn’t live in the pitch; it lives in the patterns. In what hundreds or thousands of buyers say when no one is selling to them. So I went there.

Most machine learning buying decisions today rely on demos, vendor narratives, and analyst perspectives. To ground this in real-world experience, we analyzed 500 verified user reviews from teams that have implemented and operated ML software over time. This approach reveals where ML delivers value, where it falls short, and how it impacts measurable business outcomes. Here’s what the data shows.