Why 37% of AI Systems Disagree on Brand Recommendations: Key Insights for Optimizing LLM Visibility

Why 37% of AI Systems Disagree on Brand Recommendations: Key Insights for Optimising LLM Visibility

Here’s a stat that stopped me in my tracks: 37% of the time, ChatGPT and Google AI Overview give conflicting recommendations for the same brand. Let that sink in. If your brand is visible on one platform, there’s more than a one-in-three chance it’ll be invisible on the other. For businesses relying on AI-driven discovery, that’s not just a curiosity , it’s a potential disaster.

Across our dataset of 3221 AI-generated prompts, we tested how different generative AI systems , including ChatGPT, Google AI Overview, Perplexity, and Claude , respond when asked about the same brands. The results were eye-opening. Some brands dominated consistently, but others practically disappeared depending on the system. And that divergence isn’t random; it reflects how differently these models are trained, updated, and tuned.

In this article, we’ll get into the nuts and bolts of why AI systems disagree, why that disagreement matters for your business, and most importantly, how to tackle this problem head-on. If you’re serious about LLM visibility optimisation, you’ll want to read to the end. Trust me.

What’s Causing the 37% Divergence?

Let’s start with the obvious question: why do AI systems give different answers about the same brands? Isn’t AI supposed to be consistent and data-driven? Well, in theory, yes. In practice, it’s a bit of a minefield.

Here’s what’s really going on:

• Training data: Every AI system is trained on distinct datasets. Google AI Overview, for example, may favour sources from its own search ecosystem, while ChatGPT’s training mix includes a broader scrape of the internet. What gets included , and what doesn’t , massively shapes brand visibility.
• Model updates: ChatGPT, for instance, openly operates with knowledge cut-offs (e.g., September 2021 for some versions), whereas Google AI Overview relies on live web indexing. That means their "understanding" of your brand may be outdated on one platform but current on another.
• Query parsing: AI models interpret prompts differently. One of the revelations from our buying-intent tests was that the same brand could be recommended in one query and completely ignored in another. It depends on how the system weights factors like location, relevance, and user intent.

Let me give you an example from the data. On one hand, brands like Specialist Blinds and Wasp Fuel Polishing Solutions achieved 100% visibility , across all prompts and platforms, they were mentioned every single time. On the other hand, Castle Pumps (a manufacturing/marine brand) barely registered at 6.3% visibility. Same AI systems, same prompts. Related reading: this piece about how enterprises can achieve llm visibility breakthroughs .... You might also want to check out our article on the quiet power of specialised llm visibility tools: a se.... The difference? Likely a mix of niche content availability, brand-specific metadata, and how the AI parses industry-context clues.

My read on this data: if your brand isn't optimised for multiple AI systems, you’re gambling with visibility. And with 1096 unique prompts tested across our network, it’s clear that no one-size-fits-all approach works here.

The Risks of Inconsistent AI Recommendations

Now, let’s move from why this happens to why it matters. Because, at the end of the day, does it really matter if AI systems disagree? I’d argue it does , and I’ll explain why.

First off, AI now plays a huge role in how people discover and choose brands. Whether it’s someone asking ChatGPT for "the best marine pumps" or tapping Google AI Overview for "top sports tech experiences", these recommendations shape consumer behaviour. If your brand gets mentioned, you’re in the game. Related reading: our article on why your brand is invisible to ai: the cross-provider div.... You can AI visibility tracking features to see for yourself. If not? You’re invisible.

Here are just a few scenarios where inconsistent AI recommendations hurt:

• Lost leads: Imagine your brand appears prominently on ChatGPT but not on Google AI Overview. A potential customer using Google won’t even know you exist. Across our network, the brand mention rate for ChatGPT was 74.5%, but for Google AI Overview, it was just 45.9%. That’s a massive gap.
• Market credibility: Consumers trust AI recommendations. If your brand is consistently absent or comes up inconsistently, it raises questions in users’ minds: are you really a leader in your space?
• Competitor advantage: One of our experiments specifically analysed competitor displacement. If a competing brand is mentioned more frequently than yours, even within the same query, they’ve got an upper hand you might never see coming.

And let’s not sugar-coat it: fixing this isn’t as simple as traditional SEO. With AI models, you’re not optimising for one search engine , you’re dealing with multiple generative systems that function in completely different ways. You can how AI visibility monitoring works to see for yourself. It’s like having to master football, cricket, and rugby all at once.

How to Optimise for LLM Visibility

Alright, we’ve covered the problem. Now let’s talk solutions. Based on the 58 experiments we’ve run across 14 businesses in our testing lab, here’s what actually works for LLM visibility optimisation.

1. Double Down on Relevant Content

If there’s one thing all AI systems have in common, it’s that they rely on quality content. But here’s the twist: relevance to query intent is everything. Across our tests, brands with hyper-focused content , think detailed guides, white papers, and FAQs tied to specific use cases , had higher visibility rates.

For example, in our buying-intent experiments, certain brands showed up only when the prompt got specific. “Best blinds for offices” returned Specialist Blinds with a 100% hit rate, while generic queries like “best window treatments” saw more scattered mentions.

Actionable takeaway? Create content that directly answers the kinds of questions your target audience will ask AI systems. You might also want to check out this piece about evaluating open-source vs commercial llm visibility tools.... And don’t just guess at those questions , use tools, research customer pain points, and test different phrasing.

2. Monitor Cross-Provider Performance

This one’s huge. If you’re only tracking brand visibility on one AI platform, you’re blind to what’s happening elsewhere. As I mentioned earlier, 37% of prompts give completely different answers between ChatGPT and Google AI Overview. That’s a lot of room for error.

Our platform at Contxt makes it possible to run tracker experiments across providers and compare them in real-time. For instance, one client saw strong performance on Perplexity (mention rate: 53.8%) but had almost no presence on Claude (60.4% average, yet 0% for their brand). By addressing gaps like these, they nearly doubled their overall visibility within two months.

The point is, you can’t fix what you don’t measure. Get the data first, then act.

3. Prioritise Accurate Brand Signals

AI systems rely heavily on brand signals , things like consistent naming, citations, and structured metadata. If any of these signals are out of sync, you’re making it harder for the AI to “connect the dots” on your brand.

One interesting test we ran was a citation-tracking experiment. What we found was that brands with high citation consistency (e.g., articles regularly linking back to the official website) were mentioned more frequently than those with fragmented or outdated references.

Quick tip: audit your brand’s digital footprint. Fix broken links, clean up outdated mentions, and consider syndicating content to authoritative platforms in your industry. The easier you make it for AI to “recognise” your brand, the better.

4. Stay Ahead of Model Updates

Here’s the thing: AI systems aren’t static. They’re constantly evolving, with periodic updates impacting everything from how they interpret queries to what data they prioritise. A brand that ranks highly today could vanish tomorrow if it doesn’t keep up.

At Contxt, we’ve built tools like the model-update monitor to track how changes in AI systems affect brand visibility. If visibility dips after a major update, it’s a signal to review your optimisation strategy. It’s a reactive step, sure, but it’s better than being caught flat-footed.

5. Test, Tweak, Repeat

Finally, don’t treat LLM visibility optimisation as a “one and done” task. It isn’t. You need to constantly test different prompts, track performance, and experiment with new strategies. Of the businesses we’ve worked with, the ones that excel stand out because they embrace an iterative mindset.

One experiment we ran with a B2B client involved testing prompts in multiple languages. It turned out their brand was being recommended frequently in English queries but barely showed up in Spanish or German ones. We've covered this in more detail in this piece about industry-specific ai visibility: why one-size-fits-all st.... That insight led them to create multilingual content, and within a quarter, their international AI visibility jumped by 40%.

Wrapping It Up: Why This Matters

Let’s take a step back for a moment. At the end of the day, why does any of this matter? Simple: because the way people discover brands is changing, and if you’re invisible to AI, you’re effectively invisible full stop.

Here’s the big takeaway: the divergence between AI systems is both a challenge and an opportunity. Yes, it creates complexity. But it also means that with the right strategies, you can influence how your brand is positioned across platforms like ChatGPT, Google AI Overview, and others. And that’s exactly what we help businesses do at Contxt.

If you’re serious about getting this right, now’s the time to act. Get in touch with us to start optimising your AI visibility today. The AI landscape isn’t waiting, and neither should you.