First: make sure this is actually your problem
Before reading further, it is worth confirming you have a commodity result rather than a ghost one. The distinction matters because the fixes point in different directions.
- A ghost company does not appear in AI results at all. Either that, or it appears so faintly that a buyer would not register it. If that describes your prompt test results, the ghost fix article is the right starting point.
- A commodity company appears. It is mentioned in category queries and sometimes in comparison queries. But the description AI produces is generic: “a cloud software platform that helps businesses streamline operations”.
The commodity result is even more insidious than a ghost result in one important way: it is easy to mistake for a visibility win. Your company appears, so you assume AI models know you exist. But the description it produces would not move a buyer to choose you over anyone else.
Why am I getting the commodity result?
Good news first: a commodity type response is almost never due to a lack of content.
Most companies in this situation have invested in marketing. They have a website, a blog, case studies, and a LinkedIn presence. So the problem is not volume.
The actual problem is a little harder to solve: register.
Too much “capability” language blurs important lines
AI models synthesize what they find, and they reflect the kind language you publish back at you.
A commodity result often comes from content full of capability language that describes what your product does in general terms. What it lacks is language about who it serves and what specific outcomes it produces.
- Useless as an AI signal: “We help businesses work smarter”
- Strong AI signal: “We help mid-market logistics companies reduce ERP implementation time by 40%”
Inconsistency is a bi-product of growth, but it’s fatal to AI search results
There is also a consistency problem. Many B2B tech companies describe themselves differently depending on the channel and the audience.
The website says one thing, the G2 profile says another, the CEO’s LinkedIn bio says a third. This is inevitable to some degree, and it’s hard to avoid at scale. The problem is that AI models weight consistency as a trust signal. When language is inconsistent across sources, AI models blend and average it all out. Output like that is almost always generic.
Say it, then prove it
Finally, there is the proof point problem. Claims without evidence are invisible to AI. “We are the leading provider of X” is a claim every competitor makes and AI cannot differentiate on it.
But named customers and concrete use cases are different. These are the signals AI can extract and verify against other sources. Signals like that are far morley likely to feature in model outputs.
How can I fix this?
As we saw earlier, the commodity result is less about quantity than it is about quality. Specifically, it’s about whether AI models can extract anything from your content that distinguishes you from the fifteen other vendors in your category.
The five adjustments below address that at the level of your content and third-party presence.
1. Shift from claims to proof
This is the highest-leverage change a commodity company can make. It also requires the least additional content investment. The work is rewriting what you already have, but with specificity.
- Go through your homepage, your core service or product pages, and your most prominent blog content.
- Identify every sentence that makes a claim without evidence. “We help businesses grow faster.” “Our platform is trusted by leading companies.” “We deliver results.”
- Replace each one with a specific, attributable statement like a named customer or quantified outcome.
The test is simple: could any competitor in your category make the same claim without changing a word? If yes, the claim is doing no differentiating work and is contributing to your commodity result. Rewrite it until the answer is no.
2. Prioritize vertical depth over horizontal breadth
Commodity content tends to be horizontal. It’s written to appeal to as many potential buyers as possible, which means it speaks directly to none of them. The fix is to go deep on the verticals, use cases, or buyer profiles where you actually win.
Emphasize the verticals where you’re strongest
A B2B tech company that serves financial services, logistics, and healthcare clients will almost always have a stronger presence in one of those verticals than the others.
That is the vertical to write for first. Depth in one area builds the kind of authoritative signal AI models can extract and use. A single well-researched article about a specific problem facing mid-market logistics companies, backed by a case study and a named outcome, will do more for your AI visibility than five generic articles about digital transformation.
Vertical depth also helps with the specificity queries that reveal where your content is actually working. As discussed in the prompt testing guide, the point at which you start appearing in narrowed queries like “best ERP implementation partner for logistics companies” is a direct readout of where your content has built a signal. Commodity companies typically only appear when the query is very broad. Building vertical depth moves that threshold.
3. Make your ICP explicit everywhere
One of the most common commodity patterns is a company that serves a clearly defined ideal customer profile in practice but describes itself in generalist language in public. The sales team knows exactly who they sell to. But the marketing copy tries not to exclude anyone.
The problem: AI models cannot infer your ICP from your win rate
If your content does not state clearly and consistently who you serve (company size, industry, geography, stage, buying trigger) the model will describe you in the broadest terms available. That’s usually the category definition.
The fix is to make your ICP explicit in the places that carry the most weight:
- Homepage hero copy
- About page
- Clutch, G2 and Capterra profiles
- Introductory paragraphs of your most-visited blog posts
It’s important that this information features in the first place a reader (or AI model) will look. Don’t bury it in a disclaimer or a footer.
4. Invest in third-party proof you don’t own
Case studies published on your own website are useful, without a doubt. But they carry limited weight with AI models because they are self-reported.
The proof that sends the strongest signal is third-party:
- A customer who mentions you by name in a review
- Trade publication articles that cover a deployment you led
- The category report that lists specific capabilities and highlights you in the analysis
The goal is to create a paper trail of specific, attributable outcomes that exist outside your own domain. Even a handful of detailed external reviews that describe concrete use cases and named results will produce a more differentiated AI summary than a polished case study PDF on your website.
5. Give AI something to quote
This is a subtle but important point. AI models tend to surface responses that include specific, quotable claims. That includes statistics, named outcomes, or defined positions. A company that publishes content with clear, citable assertions gives AI something concrete to work with.
The challenge here, for many B2B companies, is that it means being willing to take a position.
State your ICP precisely. Quantify an outcome from a real engagement. Define the category problem in a way that is specific to your point of view. Publish a benchmark or a finding from your own client work.
These are the assets AI models can cite, and citing them is how you move from a generic description to a differentiated one.
| Am I a “commodity” or just partly fixed?Many companies sit between being invisible and being fully differentiated. They appear in broad category queries but disappear when a vertical or use case qualifier is added. If that describes your results, the commodity fix and the ghost fix are both relevant, but start here. The specificity problem is usually upstream of the volume problem. Companies that add more generic content to a Commodity foundation tend to entrench the Commodity result rather than escape it. Fix the register first, then scale the volume. |
Signs your AI visibility is improving
A company that has fixed its commodity result appears with a description that a buyer in a specific segment would recognise as written for them. The AI response should:
- Names an industry (with some level of specificity)
- References a use case that’s relevant to your work
- Cites an outcome, ideally with hard facts or figures behind it
- Places the company in a defined part of the market rather than the entire category.
Positioning first, publishing second: Improve AI search visibility with Inspired Marketing
Most B2B tech companies that get a generic AI result keep doing the same thing: publishing more content and running more more-of-the-same campaigns. The volume goes up, but the AI description stays generic.
If you have read this far and recognise your company in it, the next step is not another blog post. It is an audit of what you are actually signalling, and a clear brief for what needs to change.
That is exactly what we do. If you want to know what your AI visibility audit would look like and what we would prioritise first, get in touch.
