Jason M. Lemkin argues companies charging extra for AI features will lose to competitors who bundle it into the base product. Zendesk's $50 AI surcharge failed while Atlassian stopped charging for AI entirely. The winning strategy is to give AI away now and raise prices later if the market supports it.
Chapters
I have the strongest thoughts in all of planet Earth on this topic. And we're in a transition phase. Okay, so I'll tell you some thoughts and tie it to some of the speakers today. I think 6 months ago everyone was like, wow, well, AI is so expensive. I have to charge $2 every time someone clicks. Okay. Yes, bleeding edge models are expensive, but the rest of it's so cheap. It's so cheap you don't have to charge $2 per click. Okay, so there was a, there was a certain fear of costs maybe a year ago, right? And then a year ago that started to bleed into greed maybe 6 months ago. Hey, maybe I can triple my price for AI, right? So I talked about how impressed I am with Zendesk, right? I'm also unimpressed with Zendesk. You know why? They want $50 more a month for AI. Zendesk used to be $15 a month. Now it's more and they want $50. What did Atlassian say last week? We're not charging for AI anymore. We're not charging anymore. Salesforce went out. I just did a great deep dive with Marc Benioff. The best, right? They went out with Agentforce. It was $2, not per resolution, just $2 per action to click in Agentforce. Agentforce is working. It's working for the same reason all support and field is working. It's working as near as any Wall Street analysts can tell. They can't really charge for it. It's being bundled in. Okay. Thank you. So that's the tactic, the strategic answer. If you listen to every, a lot of folks want to charge for it, but it is, it is the core. AI is the core. So stop charging for it. If you're not sure today, don't charge for it. Give as much of it away as you can. If the market then lets you charge more for it, great. But like, if you're charging $50 a month for your AI add-on and your competitor builds it in because it's core, you will lose. You will lose, you will lose. So that's gonna hap— no matter what works today, you're gonna lose when the competitor includes it as core to their product. So those are my thoughts.
