Does Unique Content Still Matter for AI Search?
If you have ever looked at the flood of content online and thought "I am one person, I cannot possibly keep up with the volume," I have news that should make your week. The thing you saw as your disadvantage is turning into your biggest asset, and Google is the one making it happen.
Here is what is actually happening. Google has a patent, and a set of internal systems, that measure how much new information a page adds compared to what a reader has already seen on the same topic. Content that just repeats what is already out there scores low and gets pushed down. Content that adds something genuinely new gets surfaced. And the one thing a content farm or an AI text generator cannot fake is your real, first-hand experience.
Key takeaway: Google now appears to score how much new information a page adds beyond what a searcher has already read, and to deprioritize pages that add nothing. For a solo maker, low volume is not the weakness it looks like. Your real bench experience and your own data are exactly what these systems reward, and they are the one thing the flood of generic content cannot copy.
Why this matters right now
More than half of the internet is now what the industry politely calls "slop," meaning mass-produced, low-effort content that says nothing a dozen other pages have not already said. Google knows this, and it is responding by thinning its index and culling the pages that add no value. I am using the blunt word on purpose: culled, not tidied away.
You have probably felt the shift already. The old way was simple: someone searched Google, clicked a result, landed on your site. Today Google answers many questions directly, and ChatGPT recommends stores by name with no click at all. In that world, being present is not enough. You have to be worth surfacing over the ten near-identical pages next to you.
Here is where it gets uncomfortable for those of us in jewelry. Think about the content we were all told to make. The gemstone guide. The birthstone chart. The types-of-silver explainer. The how-to-clean-sterling-silver post. That was the playbook every SEO course sold us: pick the keyword, write the same guide everyone else is writing, and rank. It worked for a while, in the old game where ranking was the whole prize.
It is exactly the content the new scoring pushes down, because a hundred versions of it already exist and yours adds nothing they do not. This does not mean burn your gemstone guide. It means the generic version, the one anyone could have written, is the part that stopped working. Put your own bench into it and it changes category entirely.
What "information gain" actually means
Strip away the jargon and it is a simple idea. Imagine a customer is learning about, say, caring for opal jewelry. They read one page. Google's systems ask a question about the next page they might show: does this one add anything the reader did not already get? If the answer is yes, it is worth showing. If the answer is no, showing it is a waste of the reader's time.
Google has a patent that describes exactly this, called Contextual estimation of link information gain. It gives a page a score for how much new information it adds beyond pages the reader has already seen. A page that scores near zero, meaning it adds nothing new, is unlikely to be shown to someone who already read a similar page. Two related systems, one called OriginalContentScore and one called contentEffort, appear to do the same job from different angles, with contentEffort using AI to estimate how much real effort went into a page.
An important note so you do not overcorrect: a patent proves Google owns an idea, not that the idea is running in the live system today. Treat this as a strong signal of where Google is heading, which lines up with everything else they reward, rather than a confirmed switch. The principle is the point, and the principle is sound.
Why this flips the script for a one-person studio
For years, the story independent makers told themselves was "I cannot compete on volume." And on volume, that was true. But Google is no longer rewarding volume. It is rewarding contribution.
A content mill can generate a thousand opal-care articles overnight. Every one of them rewrites the same facts. Under an information-gain model, the thousandth article adds nothing the first one did not, so it scores near zero. Meanwhile, you sat at your bench and watched a specific opal behave a specific way while you set it. You know something that is not written anywhere else on the internet. That is not a small thing. That is the exact thing these systems are built to reward.
Kevin Indig's research puts a number on it: pages with first-party research, meaning original data or first-hand findings, earn roughly 3.3 times more citations in AI answers than pages without it. Original data is the single strongest predictor of whether a page reads as genuinely original. Your real experience is not a nice-to-have. It is the single most powerful ingredient you have.
3.3x
Pages with first-party research (original data or first-hand findings) earn about 3.3x more citations in AI answers than pages without it.
Kevin Indig, first-party content research, 2026. Slop figure: Search Engine Journal, July 2026.
The proof this is not just theory
I do not ask you to take a patent on faith, so here is a real result from my own store, framed honestly.
In June 2026, I asked Google AI Mode how independent jewelers get found and recommended by AI. Google AI Mode cited Red Pin Geek three times in a single answer, including my Andrea Li Designs visibility case study. By July 2026, that citation had widened to a second engine, Perplexity, on the same question. And for a core shopping search, "pastel gemstone jewelry," the method I used on Andrea Li Designs now has the brand recommended, not just found, across Perplexity, Gemini, and the Google AI Overview.
The brand recurs. The engine varies. ChatGPT is still the holdout.
Same question, two dates. Citation widened from one engine to two over two weeks. New content earns citation on a lag.
Query tested: “How do independent jewelers get found and recommended by AI?” (advisory / Foundation Gap query)
| Engine | 2026-06-17 | 2026-07-01 |
|---|---|---|
| Google AI Mode | ✓ Cited3 RPG pages in one answer |
Established, not re-pulled |
| Perplexity | – Not cited |
✓ CitedThe SEO / AI-visibility guide |
| ChatGPT | – Not cited |
– Not cited |
| Gemini | Not tested |
– Not cited this run |
Andrea Li first-party multi-engine checks, June 17 and July 1, 2026. Point-in-time snapshots; AI citations drift.
Recommended, not just found — on 3 of 4 surfaces.
Door 2: the brand is the answer, not just a source. Absent on “engagement” (not bridal-positioned). The exact cited URL varies run to run.
Query tested: “pastel gemstone jewelry” (commercial / shopping) · brand: Andrea Li Designs (andreali.com) · 2026-07-01
3 of 4 surfaces recommend the brand. One holds out. Both are shown.
Andrea Li first-party check, July 1, 2026 (Perplexity, Gemini, Google AI Overview; ChatGPT via API). Snapshots drift.
The honest boundaries on that: those are specific engines, specific queries, on specific dates, and AI citations drift over time. ChatGPT did not cite me on that question. I am not claiming "AI cites Red Pin Geek" as a blanket fact. But the consultant whose whole subject is AI visibility is demonstrably visible to AI for that subject, and she got there with first-hand content, not volume. That is the method working on the person selling the method.
Length is not the lever
One trap to avoid: do not read "add more information" as "write longer." Google normalizes for length, which means a 3,000 word page does not win just for being long. Depth and rigor are good when they serve the reader, but answering the question well and getting the reader to their goal beats padding every time. A tight, specific, first-hand 700 word piece will beat a bloated 2,500 word rewrite of everyone else's page. Add information, not word count.
Commodity content vs information-gain content
Commodity content vs information-gain content
| Factor | Commodity content | Information-gain content |
|---|---|---|
| Where it comes from | Rewrites what already ranks | Your first-hand experience and original data |
| Who can make it | Anyone, including AI | Only you |
| How Google reads it | Scored low, pushed down | Scored high, surfaced next |
| AI citation odds | Low | About 3.3x higher with first-party data |
| Cost to produce | Cheap and fast | Costs your real time, which is the point |
| Shelf life | Culled as the index thins | Durable |
Framework: Red Pin Geek.
How to add information gain (without a big budget)
You do not need a research team. You need to put more of your actual self into the page. Seven ways, in rough order of power:
- Original data. Even a small number you tracked yourself counts. How many custom opal pieces you made last year, how long a repair actually takes, what a stone cost you at the show.
- First-hand experience. The bench moment. The setting that failed and how you fixed it. The thing the stone did that surprised you.
- Interviews. One conversation with a supplier, a customer, or a fellow maker adds a voice no rewrite has.
- Real reporting. What you actually saw at Tucson this year. What is changing on the show floor.
- Being first. Cover a shift as it happens, in your own words, before the rewrites catch up.
- Your own analysis. Your honest read on your own numbers or your own market. Nobody else has your vantage point.
- The kill test. Before you publish, ask: does this add anything to what already exists? If you cannot answer yes, do not publish it. A page that adds nothing costs you time to make and is less and less likely to earn anything back.
Stay in your lane. Just make sure the page has a piece of you in it that no one else could have written.
There are two ways to be invisible
Here is the trap on the other side, and I have watched makers fall into it the moment they hear "put more of yourself in." It is tempting to swing all the way to the opposite end and turn the blog into a personal diary: the trip that inspired a collection, the certification you earned, the season you are having at the bench.
That content is unique to you, which sounds like exactly what wins. But it is missing the other half of the equation. Nobody who has not already met your brand is searching for your trip or your certificate. There is no question out there that a diary answers. So it never gets found and never gets cited, for the opposite reason commodity content fails: it is one of a kind, but no one is looking for it.
So there are two ways to be invisible. Commodity content is what everyone searches for and no one needs another copy of. Pure lifestyle content is what only you could write and no one is searching for. The pages that get cited live in the overlap: a question real people ask, answered in a way only you can answer it. That overlap is exactly where most makers go missing when a buyer uses AI to research a purchase, the gap between being aware of you and buying from you that I have called the invisible middle. It is the content that sits between your lifestyle stories and your product pages, and it is usually the exact thing missing from a maker's site.
There are two ways to be invisible
The overlap is the invisible middle — the layer most makers are missing.
Red Pin Geek.
The fix is not to throw out the personal story. It is to use it as the ingredient, not the whole dish. Take the trip and turn it into "what real desert turquoise looks like up close." Take the certification and turn it into "how to choose crystals for intention," a question people actually type into a search bar. The searchable question goes in your title. Your first-hand experience fills the page. Now the story is working for you, because a stranger can finally find it.
Frequently asked questions
Does AI-written content still rank?
Generic AI content that rewrites what already exists is exactly what these systems are built to deprioritize. AI as a drafting assistant is fine. AI as the whole product, with no first-hand contribution, is the commodity content getting pushed down. The difference is whether anything on the page could only have come from you.
Do I have to publish a lot to compete?
No, and that is the good news. Volume is not the lever anymore. One page with your real experience in it can outperform a hundred pages of rewrites. Fewer, richer pages beat more, thinner ones.
Does my content need to be long?
No. Google normalizes for length, so long alone does not win. Answer the question well and add something new. A short, specific, first-hand piece beats a long generic one.
How do I know if a page is commodity content?
Read it and ask whether a competitor could have published the exact same thing. If yes, it is commodity content, and it is a candidate to either enrich with something first-hand or retire.
Should I just blog about my travels and my life instead?
That is the opposite trap. Personal stories are unique to you, but if no one outside your brand is searching for them, they do not get found or cited. Use the story as the ingredient inside a piece that answers a real question people type, not as the whole post.
Sources
- Harry Clarkson-Bennett, "How Google May Understand Unique Content," Search Engine Journal, July 2026.
- Google patent, "Contextual estimation of link information gain."
- Kevin Indig, research on first-party content and AI citations (first-party research earns roughly 3.3x more citations).
- Red Pin Geek first-party citation record, June to July 2026 (Google AI Mode, Perplexity, Gemini, Google AI Overview), documented per-engine and per-query.
Continue Learning
- The SEO and AI Visibility Guide for jewelry designers: the foundation this piece sits on, what makes a store findable and recommendable by AI in the first place.
- The Andrea Li Designs AI visibility case study: the real before-and-after behind the citation results above.
- Score your own store on the free AI Visibility Score: see where your pages actually stand on the signals that decide whether AI surfaces you, including whether your content adds anything AI would bother citing.
- Subscribe to the Red Pin Geek Substack (free): weekly inside-view content for jewelry designers building AI visibility on their own. The Premium tier unlocks the worked tools, including the scorecards and tier worksheets.
- Work with me on your specific store (Snapshot $97 / Full Audit $597 / Full Audit + 1:1 call $997): the private read. I run the same commodity-vs-first-hand check across your pages and tell you which ones are adding nothing, and how to fix the ones worth saving.
About the author: Andrea Li is the founder of Red Pin Geek and has spent 18 years in the jewelry industry as the designer behind Andrea Li Designs. She tests every AI-visibility strategy on her own store before she recommends it to anyone. redpingeek.com | andreali.com
Want Red Pin Geek in Your AI Answers?
Google now lets you choose preferred sources that get labeled in AI Overviews and AI Mode. If this content helps you navigate AI visibility for your jewelry business, add us and you'll spot our links faster next time Google answers your questions.
Add Red Pin Geek as a Preferred SourceTakes one click. Works in Google Search, AI Overviews, and AI Mode.
