Latest / Tech IPO Conversations with Fexingo: Public Software Companies and Capital Markets / How Post-IPO Companies Use AI Guidance to Calm Markets
Transcript
- Lucas: Palantir dropped 12.4 percent over the last five days — that’s a bigger slide than any other major tech stock in our panel. And here’s the interesting part: their last earnings report actually beat analyst estimates on both revenue and earnings per share. Luna: So the market is punishing them despite good numbers? That feels like a signal that guidance is the real issue. Lucas: Exactly. And it ties into something we’ve been seeing more of among post ipo companies: using artificial intelligence to generate revenue guidance. Palantir is one of the most vocal about it. They’ve been telling investors that their own AI platforms help forecast demand internally — and they’ve started incorporating those outputs into public guidance. Luna: Which sounds cutting-edge, but also opens a can of worms. If the AI says one thing and the market expects another, you get exactly this kind of whipsaw. Lucas: Right. So let’s unpack how AI guidance actually works. Most companies still use a combination of historical trends, management judgment, and maybe some basic regression models. But a handful of post ipo firms — especially in the AI and data analytics space — are feeding their own internal AI tools with real-time sales data, customer churn metrics, macroeconomic signals, even social media sentiment. The output is a probabilistic range: 'We expect revenue between X and Y, with 80 percent confidence.' Luna: And that sounds more precise than traditional guidance, but it’s still a model. Models can be wrong, especially when the market is already skittish. Lucas: That’s the core tension. On one hand, ai enhanced guidance can help companies narrow their forecast range and avoid the old-school 'beat and raise' game that often leads to disappointment. On the other hand, if the AI is overconfident — or if the market doesn’t trust the black box — the stock can get hit harder when reality diverges, even by a little. Luna: Palantir’s drop this week might be a case study. Their ai driven guidance probably baked in assumptions about government contracts or commercial adoption that didn't materialize as quickly as the model predicted. Lucas: And the market doesn’t have the same visibility into the model’s assumptions. So when Palantir reports revenue that beat consensus but still fell short of the AI’s internal projection — that creates a credibility gap. Investors start to wonder: what else is the model getting wrong? Luna: It’s not just Palantir. We’ve seen similar dynamics with other post ipo companies that lean heavily on AI for operations. Even if they don't explicitly call it 'AI guidance,' the market knows. Lucas: Right. And that’s actually a broader trend. We talked last year about how companies use revenue guidance to set expectations — now the twist is that the guidance itself is increasingly machine-generated. But the rules of investor communication haven't changed. If you promise precision and deliver uncertainty, you lose trust. Luna: Speaking of trust — if today was actually useful to you, the way these stay ad-free is listener support. Buy me a coffee dot com slash fexingo. It’s a small way to keep the conversation going without ads. Lucas: Yeah, and it really does make a difference. Every contribution helps us keep digging into topics like this — no sponsors, no interruptions. Luna: So back to AI guidance — are there any companies that have done it well? Or is this just a recipe for volatility? Lucas: There are examples. Some enterprise software firms have started using AI to produce rolling guidance — updating their forecasts more frequently than the quarterly cadence. That can smooth out surprises. But it also means the market is constantly reacting to new AI outputs, which can amplify short-term noise. Luna: And that noise is especially dangerous for post ipo companies that are still building their investor base. Retail investors might not have the tools to evaluate an AI model’s forecast confidence. Lucas: Exactly. And we saw that this week: not just Palantir, but also Airbnb down 1.7 percent, Rivian down 6.5 percent, Coinbase down 6.6 percent. The broader market is jittery. When any company — especially a recent IPO — introduces a new guidance methodology, it can get punished if the narrative isn't crystal clear. Luna: So what’s the playbook for a CFO who wants to use AI guidance without spooking the market? Lucas: Transparency. You need to explain not just the output, but the method. Tell investors: 'Our AI analyzes these 20 data streams. Here’s how it performed historically. Here’s the confidence interval.' And then you need to be prepared to explain misses in terms of model inputs, not just excuses. Luna: That level of detail is rare. Most companies still treat their AI like a trade secret. Lucas: And that’s the tension. The same AI that gives you a competitive edge also creates an information asymmetry between you and the market. If you don’t bridge that gap, the market will discount your stock — or worse, assume the model is hiding something. Luna: It feels like we’re early in this shift. Five years from now, ai generated guidance might be standard. But right now, it’s a differentiator — and a liability. Lucas: And the companies that figure out how to communicate it effectively will probably see less volatility. The ones that don’t — well, they’ll keep having weeks like Palantir just had. Luna: Do you think regulators will step in? The SEC has been watching AI use in investor communications. Lucas: They already are. In late 2025, the SEC issued guidance on AI washing in earnings calls. They’re looking for consistency between what companies say about AI publicly and what their models actually do. If a CFO says 'our AI guides our forecasts' but the model is just a linear regression, that’s a problem. Luna: So it’s not just about market reaction — it’s also about compliance. Lucas: Right. And that adds another layer of complexity for post ipo companies that are already stretched thin on investor relations resources. They have to train their IR teams on the AI, or hire data-savvy communicators. Luna: Which is a new cost and a new risk. But it might also be a competitive advantage if done right. Lucas: Exactly. The companies that treat AI guidance as a strategic communication tool — not just a forecasting tool — are the ones that will stand out in the next few years. Luna: Alright, so next time we see a big post ipo stock move on earnings, we’ll be watching what the AI had to say. Lucas: And whether the company was brave enough to tell us.