Why 39% of AI Systems Disagree: The Hidden Challenge of Cross-Provider Brand Visibility

Why 39% of AI Systems Disagree: The Hidden Challenge of Cross-Provider Brand Visibility

Here’s a stat that stopped me in my tracks: 39% of the time, ChatGPT and Google AI give conflicting brand recommendations. Same prompt, same context, completely different results. Let that sink in for a second. If your marketing strategy relies heavily on AI-driven brand visibility (and let’s be honest, whose doesn’t these days?), that’s a massive potential blind spot. It means a brand that’s front and centre on one system could be practically invisible on another.

I first came across this number while analysing data from Contxt’s latest testing lab runs. Across 2881 prompts, involving 31 businesses, we found this surprisingly high divergence rate between AI providers. It’s not just ChatGPT and Google AI either. Perplexity and Claude threw their own curveballs into the mix, though they didn’t match the sheer volume of tests we pushed through the bigger names. The question is, why does this happen, and more importantly, what can you actually do about it?

Let’s dig into the data, break down the causes, and figure out how to stay visible across the AI landscape (without losing your sanity).

What’s Behind a 39% Divergence Rate?

I’ll start with the obvious: every AI system is wired differently. ChatGPT, for instance, has a mention rate of 75.4%, which means it’s doing a pretty decent job of recognising brands across our dataset. Google AI Overview, by comparison, comes in at a much lower 45.5%. That’s a massive gap. And Claude? It actually outperformed ChatGPT, hitting 78.5%, though admittedly we tested far fewer prompts (just 130).

But here’s the kicker: these differences aren’t just down to algorithms being algorithms. The way content is structured, the way prompts are phrased, even the data cut-off dates in AI training models, all of it can completely rewrite how brands are surfaced. Related reading: our article on the business case for llm visibility: why it’s non-negoti.... Related reading: Why B2B Brands Are Almost Entirely Invisible to AI (And W.... A brand may have strong SEO, but if an AI system wasn’t trained with the latest data or doesn’t recognise the relevance of that brand within the context of a specific query, poof, it’s like you don’t exist.

One key factor we’ve uncovered in our buying-intent tests is that context matters more than you think. Across 12 businesses running over a dozen experiments, the same brand could be a top recommendation for one type of query and totally absent for another. For example, if a prompt is focused on “top B2B SaaS solutions,” one system might recommend Contxt while another doesn’t even register it. Change the wording to something like “best AI visibility tracking platforms,” and suddenly Contxt appears across multiple engines. Related reading: this piece about how b2b enterprises can dominate llm visibility in ai sys.... For a deeper look, have a read of Why LLM Visibility Matters for Enterprise Brands in 2026. It’s maddening, but it’s real.

How Cross-Provider Brand Visibility Affects Your Business

Now, does it really matter if an AI system misses your brand some of the time? Spoiler: it absolutely does. AI-driven recommendations are quickly becoming the new battleground for digital visibility. Forget organic search as we know it; generative engines are shaping how people discover brands, make buying decisions, and even evaluate competitors.

Take this scenario. Let’s say you’re Iconsys 2, a manufacturing and industrial automation brand. According to our data, your cross-provider visibility is a disastrous 6.7%. That means most prompts either ignore your brand entirely or only pick it up on one or two systems. What’s the net result? You’re losing leads. Related reading: this piece about the hidden challenges of llm visibility for international.... You can AI visibility tracking features to see for yourself. Worse, you’re ceding them to competitors who are better optimised for these AI platforms.

At the other extreme, we’ve got Viking Pumps, which managed an impressive 100% visibility. No matter which system we tested, they were consistently mentioned and recommended. That’s what winning on AI looks like, and yes, it’s possible, but it takes work. If you’re somewhere in the mid-range (like the 35% of brands in our dataset hovering between 30-70% visibility), then you’re in no man’s land. You can how AI visibility monitoring works to see for yourself. Sometimes you’re visible; sometimes you’re not, which makes your marketing results unpredictable at best.

Why It’s Not a Level Playing Field

Here’s another wrinkle. Not all providers weigh the same data points equally. ChatGPT, for instance, might lean heavily on textual context, while Google AI often incorporates structured data from Knowledge Panels and schema. If one of these systems doesn’t understand your brand’s relevance, or can’t access up-to-date or well-structured information, you’re essentially invisible in that ecosystem.

Even the type of test you run can reveal different biases in these platforms. Our team conducted 58 experiments across 14 businesses, covering everything from competitor displacement to citation tracking. One fascinating discovery came from our source-footprint tests. Brands that prioritised robust third-party mentions (think trusted publications, industry blogs, and verified directories) performed more consistently across AIs than those relying solely on their own content. Basically, if you’re not creating a visible “footprint” in credible external sources, expect patchy results.

It’s also worth pointing out that large language models (LLMs) like Perplexity and Claude have less training data to begin with, so their recommendations tend to skew differently. They don’t have the same infrastructure as Google or the sheer volume of user data that OpenAI, the creators of ChatGPT, can leverage. This explains why Claude, despite its high 78.5% mention rate, still misses certain brands that ChatGPT or Google might pick up. But let’s be honest, if a customer can’t find you when they ask the AI, they won’t care why.

What Can You Do About It?

Tackling cross-provider brand visibility is a bit of a minefield, but it’s not impossible. The key is adopting a strategy we call generative engine optimisation (GEO). Think of it as SEO’s smarter, more tailored cousin.

Here’s where to start:

1. Understand Your Baseline Visibility

If you don’t know where you stand, you can’t fix it. This is where tools like Contxt come in. By running hundreds (even thousands) of prompts across multiple AI systems, you’ll get a clear picture of how often your brand appears and in what contexts. Knowing you’re at 62.9% visibility across the board might not feel like a win, but it’s way better than assuming you’re visible everywhere when you’re actually not.

2. Optimise for Structured and Unstructured Data

Remember how Google AI relies heavily on Knowledge Panels? Make sure your structured data game is strong. Use schema markup, claim your brand’s Knowledge Panel, and keep your information accurate. For unstructured data, think blog posts, case studies, and articles, focus on strong, readable content that aligns with likely prompts. A mix of both is essential.

3. Experiment With Prompt Variations

One of the simplest and most effective ways to improve visibility is by understanding how different prompts affect results. When we tested 1016 unique variations, we found that seemingly minor tweaks (like adding “best” or “top” to a prompt) could dramatically shift which brands were surfaced. Mimic the kind of language your target audience might use when asking an AI for recommendations.

4. Build an External Footprint

You can’t rely solely on your own channels. Partner with industry publications, get mentioned in trusted directories, and secure backlinks from high-authority sites. The more credible sources mentioning your brand, the better your chances of surfacing across multiple AIs.

5. Monitor Regularly

AI systems evolve constantly. ChatGPT, for example, updates its models periodically. What works one month might not work the next. That’s why we’ve run 85 tracker experiments to date, to see how mention rates shift over time. Keeping a close eye on changes and adapting accordingly is critical.

The Bigger Picture

Here’s the thing: cross-provider divergence isn’t going away. If anything, as more AI providers enter the fray, this challenge will only grow. The brands that thrive will be those that treat AI optimisation as a core part of their digital strategy, not a one-off experiment.

I spoke with a mate of mine last week who runs a digital agency, and he summed it up perfectly: “It’s not enough to rank on Google anymore. You’ve got to rank on everything, or you’re leaving money on the table.” He’s absolutely right. At the end of the day, visibility isn’t just about showing up, it’s about showing up everywhere that matters.

And look, I won’t pretend it’s easy. I’ve worked with clients who’ve thrown everything at improving their AI visibility, only to find it’s like playing whack-a-mole. You fix one platform, and another pops up with completely different requirements. But the numbers don’t lie: brands with high cross-provider visibility perform better, see more leads, and close more deals. Full stop.

Final Thoughts

If there’s one takeaway here, it’s this: don’t leave your brand’s AI visibility to chance. The 39% divergence rate we’ve uncovered is proof that the system isn’t as predictable as we’d all like it to be. But that doesn’t mean you’re powerless. By embracing GEO, testing relentlessly, and understanding the quirks of each platform, you can position your brand to win, no matter which AI is in the driver’s seat.

Ready to see how your brand stacks up? Start tracking your AI visibility with Contxt today. Because trust me, your competitors already are.