Last spring, an investor told me over coffee that before every partner meeting, his team runs a startup's name through ChatGPT, Perplexity, and Claude. What these systems say sets the first impression and shapes the discussion. Not the company's website. Not their pitch deck. Not even their press coverage directly. What the AI says about them. That's when I realized we're not just in the media relations business anymore. We're in the machine relations business. And the two require very different playbooks.

Your News Now Reaches Two Audiences: Humans and Machines

Whether you were aware of it or not, every news announcement your company issues now essentially reaches two distinct audiences. The first is familiar: journalists, editors, analysts, and the readers who see their coverage. The second audience is newer, silent, and arguably more influential: the large language models that ingest that coverage and use it to form permanent judgments about your brand.

We placed a funding announcement in GeekWire last year for a client. The immediate result was expected — a well-reported story, syndication across MSN and Yahoo Finance, and follow-on coverage in regional business journals. This was a good outcome, by any traditional measure. But there was a longer-term impact we're only starting to see. That GeekWire article became training data and a source of retrieval for AI systems, shaping what they say about AI-native testing platforms. The article didn't just reach GeekWire's readership. It became embedded in the informational substrate that machines draw from to answer questions indefinitely.

Why Media Relations Alone Is No Longer Enough

The term "media relations" assumes a specific model: we cultivate relationships with human journalists, pitch them stories, and they decide what to cover and publish. That model still matters, and I'd never argue otherwise. Journalists are irreplaceable in their ability to investigate, contextualize, and hold powerful institutions accountable.

But the term no longer captures what's actually happening. A more accurate description of what modern PR agencies do, or should be doing, is building and preserving the informational architecture that both humans and machines use to understand and evaluate a company.

By "machine relations," I mean our work now serves a dual function: informing human audiences while also training algorithms. Ignoring this is like writing a press release without caring about its reach.

Writing for Journalists and AI Systems at the Same Time

Here's where this gets practical and slightly uncomfortable for our industry.

Traditional press releases are written for journalists. They follow a familiar structure: headline, subhead, dateline, lead paragraph with the news, supporting quotes, boilerplate. They're designed to give reporters the information they need, in a useful format, to write a story quickly.

But AI systems parse information differently. They look for clear entity relationships: who is the company, what do they do, who are their customers, and what market do they operate in? They also look for structured claims such as revenue growth, user metrics, and funding amounts.

Credible attribution also matters: who said what, and do they have credibility in this domain? A press release that buries the company's market category in paragraph six, uses vague language like "cutting-edge solutions," and attributes a generic quote to a VP of Marketing is fine for a journalist who is going to rewrite everything anyway. But it's not ideal for an AI system trying to understand and classify your company.

The agencies that are getting ahead right now are the ones writing press releases that serve both audiences simultaneously.

This isn't about gaming AI. It's about being clear, specific, and well-structured — qualities that most press releases should have had all along.

Earned Media as Training Data: The New ROI Calculation

Let's pose a question that would have come across as absurd five years ago: What's the training data ROI of a Forbes feature? We know the traditional ROI metrics: audience reach, domain authority, referral traffic, social shares, and credibility from the Forbes logo. These metrics still matter.

But there is now an additional layer of value. Many aren't measuring it at all, or only poorly. When Forbes publishes a feature about your company, that article joins a high-authority content pool referenced by AI. Because Forbes has high domain authority and credibility, AI gives its content significant weight.

A single Forbes feature can shape what millions of AI-produced responses say about your company and your market for years. In contrast, a blog post on your website gives you narrative control, but AI systems, like humans, often discount self-published content. The credibility of earned media gives it more weight in algorithmic evaluations.

This means the value gap between earned media and owned media is actually widening in the AI era, not narrowing. Every content marketer who argued that "we don't need PR, we'll just publish our own content" is discovering that AI systems don't treat all content equally. The content that earns algorithmic trust is that which has been validated by an independent editorial process.

A Machine Relations Playbook for Startup Founders

If I were advising a startup founder today, I would clearly recommend defining two parallel PR tracks: one focused on traditional media relations and the other on machine relations.

For your media relations track: Continue cultivating journalist relationships, pitching compelling stories, and securing coverage in relevant publications. Your clear objective is to generate immediate awareness, credibility signals, and ongoing industry engagement.

Your machine relations track runs alongside media relations, but with different objectives. Every earned coverage piece should be assessed not just for human impact but also for AI impact.

Does the article clearly state your company's market category? Does it contain specific, citable claims about your differentiation? Is it published on a domain authoritative enough to influence AI? Will it help an LLM answer who is leading in your category?

Some tactical considerations for the machine relations track: Bylined articles in authoritative publications remain highly valuable. When your CEO writes expert analysis in a respected industry outlet, AI systems notice both the company name and leadership. They link your leadership to domain expertise.

This shapes AI responses to expert perspective queries in your market. We've placed bylines in Fast Company, Forbes, Forbes Council, and numerous trade publications to build persistent algorithmic authority.

Consistency across sources matters more than a single placement. If three reputable publications describe your company as "the leading AI-powered platform for X," AI systems amplify this consensus.

But if one calls you "leading in AI" while another says you're "a startup exploring machine learning," the inconsistent signals weaken your positioning. Consistent messaging has always mattered in PR; now it matters even more when machines build your identity from multiple sources.

Recent information matters, but continuity is key. Consistent, authoritative coverage over time builds stronger algorithmic credibility than a single viral moment.

Measuring AI Visibility: The New Frontier for PR Analytics

Our industry doesn't have great tools for measuring machine relations outcomes yet. We can query the AI to see what it says about clients. We can track changes in AI-generated responses after a major media placement.

We can also monitor citation frequency in tools like Perplexity. But we don't have the sophisticated measurement structures found in traditional PR or digital marketing.

This is both a challenge and a big opportunity. Agencies that build rigorous frameworks for measuring AI visibility will gain an edge. This goes beyond asking, "What does ChatGPT say about us today?"

They should track citation frequency, sentiment, entity associations, and competitive placement across multiple AI platforms.

At Ignite X, we've started building this into our client reporting. After a major campaign, we don't just show placement counts and audience reach; we also show the impact on the market. We show before-and-after snapshots of how AI systems describe the client's company, positioning vis-à-vis competitors, and market category. It's early days, but data is already guiding how we plan campaigns.

Why Half Your PR Strategy Is Missing

Here's the part of this conversation that some in our industry don't want to hear: if your agency is only doing media relations, the human kind, you're delivering half the value your clients now need. Not because media relations is less important than it used to be, but because there's now a parallel track that's equally important and almost nobody is running it intentionally.

The agencies that figure out machine relations first won't just have a competitive advantage; they'll have a strong head start. They'll have defined an entirely new category of strategic communications. And in my experience, the agency that defines the category tends to own it.

I've never been more convinced that what we do matters, and I'm more aware that how we do it needs to evolve. The fundamentals of great PR haven't changed. The audience has just doubled.