When ten articles answer the same question, the one that gets read is the one that tells the reader something the others do not. That something comes from three places: what you did yourself, what a knowledgeable person told you, and what you counted or compared on your own. In SEO circles this idea is labelled "information gain". The label is community vocabulary. It does not appear in Google's documentation and it is not an official ranking concept, so do not present it to clients or readers as one. What Google's pages do ask is plainer: does the content offer original information, reporting, research or analysis, does it show first-hand expertise, and does it add real value beyond summarising other sources?
If you edit a blog, write for a publication or run content for clients who all compete on the same topics, the working rule is short. Name one concrete addition in every brief, make sure it exists before drafting starts, and give it a visible place on the page.
Starting from what Google actually asks
The Google page on creating helpful, reliable, people-first content (Opens in a new tab) poses a list of questions about content and quality. The ones relevant here: does the content provide original information, reporting, research or analysis; does it provide insightful analysis or interesting information that goes beyond the obvious; if it draws on other sources, does it avoid copying or rewriting them and instead add substantial additional value and originality; and does it provide substantial value when compared to other pages in search results?
The same page asks about expertise: does the content show first-hand expertise and depth of knowledge, such as expertise that comes from having used a product or service, or visiting a place? Is it written or reviewed by an expert or enthusiast who knows the topic well? Does it contain easily verified factual errors? And in the section on E-E-A-T, the page states that trust is the most important element and that E-E-A-T itself is not a specific ranking factor.
Two consequences follow. First, nobody can promise that adding an expert quote lifts a ranking. What you can say is narrower: a page that gives readers something unavailable elsewhere answers the question the guidance keeps asking. Second, the phrase "information gain" is shorthand you may use inside your team. In public, describe the actual addition: a test, an interview, a small survey, a comparison.
Three sources of difference
Table: The three sources of an addition, what each one gives the reader and what it costs the writer.
| Source | What the reader gets | What it costs you | Main risk |
|---|---|---|---|
| Your own experience | Specific detail, exceptions, mistakes | Time, careful notes | One case presented as a rule |
| Expert input | Knowledge you cannot supply yourself | Outreach, scheduling, verification | Weak credentials or altered quotes |
| Your own data | A count, a comparison or a pattern | Collection and tidy method notes | Small samples read as general truth |
You do not need all three in every article. One strong, honestly bounded addition beats three thin ones. The decision table later on helps you choose.
Reading the competition to find the gap
An addition is only an addition relative to what already exists. The SEO Starter Guide (Opens in a new tab) tells you to create content from what you know about a topic rather than rehash what others have published. To avoid rehashing, you first need to see what is published. The purpose of reading competing pages is to find the gap, not to borrow wording.
Practical suggestion: read the first handful of results for your topic and keep four notes per page.
| Note | Question |
|---|---|
| Claim | What does the page tell the reader to do or believe? |
| Basis | Does it rest on a source, an experience, an expert or data? |
| Missing piece | Which condition, exception or example is left out? |
| Shared line | Which sentence appears on nearly every page? |
The fourth note pays off most. A sentence repeated on every page may be well founded, or it may be an assumption nobody tested. If you can test it, you have your addition.
Questions to ask of a shared line
- Is there a source behind it, and does the source actually say that?
- Under which conditions does it hold, and under which does it fail?
- Can you check it with a small test of your own?
If you cannot test it, you can still repeat it, but say plainly that it is widely repeated and not sourced. Presenting an untested line as settled repeats the weakness you were trying to avoid.
Turning experience into evidence
Google's page counts first-hand expertise, such as having used a product or visited a place, among the signs of helpful content. Experience works as evidence only when it is specific and recorded close to the event. Memory smooths details out.
- Trial notes: try a method in a defined situation and write down what happened the same day.
- Before and after records: capture the state before you change something, then after.
- Mistake log: note what went wrong and how you fixed it. This record is the hardest part for a competitor to reproduce.
- Decision notes: record which criteria you weighed when you chose one option over another.
Placing experience inside the argument
Do not park experience in a separate "my story" box. Put it where it supports a claim: "Guides say X; in my case, under these conditions, I saw Y." The reader can then tell which sentences are general knowledge and which are your observation. State the limits of the observation too: one family, one season, one household, one trial.
When you have no experience to offer
Not every topic is something you can try. In that case do not claim experience. Name the role you do have: compiler, comparer, interviewer. Talk to people who have the experience, read what they have written, credit them and mark the limits of what you know. A writer who implies a visit that never happened is closer to the fabricated creator profiles Google's page calls deceptive than to a careful compiler.
Bringing in expert input
The Starter Guide says that providing expert or experienced sources can help people understand your article's expertise. That sentence makes expert input a trust signal, not a decoration. It only works when the expert is real. Google's page describes fabricated creator profiles, for example AI-generated headshots, made-up names or false credentials, as deceptive.
| Step | Do | Avoid |
|---|---|---|
| Choosing | Someone who works directly with the subject and whose background you can show | People whose credentials you cannot verify |
| Asking | Three or four concrete questions tied to the gap you found | Open questions like "what do you think about the topic?" |
| Recording | Written answers or notes taken with permission | Quotes reconstructed from memory |
| Crediting | Name, role and the reason you asked this person | "An expert says" with no name |
| Approving | Show the quote to the expert before publication | Trimming that changes the meaning |
The first message to an expert
Keep it short and concrete. Say who you are, what the article is about, the two or three questions you want answered, how the answer will be used and that you will send the quote for approval. Disclose any commercial relationship at the start. If the expert does not reply, do not fill the hole with an invented quote. Write the article with sources and your own stated limits instead.
Making the quote do work
A quote on its own does not make a page original. Say why the quote matters and how it connects to your observations. If the expert contradicts something you tested, explain the difference between the two. If two experts disagree, show both, name the conditions each assumes and give the reader a way to choose. Dropping one view quietly to make the article sound unanimous costs credibility.
For crediting sources and quotes, see citing sources in articles, and for showing the writer and the expert on the page, how to build an author page.
Producing your own data on a small budget
Original data does not require a research department. A small, clearly described collection counts as your own data as long as you are open about what it covers.
| Method | What it gives | Limit |
|---|---|---|
| Reader survey | Preferences of your audience | Represents only the people who answered |
| Trial of a product or method | Recorded results | One setting, one set of conditions |
| Count or compilation | A summary of a defined list | Only as current as its sources |
| Your own records | Patterns in your site's questions | Needs care with personal data |
| Side by side comparison | Options judged on the same criteria | Criteria are partly subjective |
The first rule is no invented data. Writing a result you did not measure is the opposite of the "easily verified factual errors" question on Google's page. If you cannot produce a measured result, describe the observation without a figure.
Writing survey questions
Ask open questions that do not steer the answer. "Do you use this feature, and what for?" beats "Don't you agree this feature is great?". Keep answer options mutually exclusive. Collect only the personal data you need, and if you need none, collect none. State plainly how few people answered and avoid building a sweeping claim on percentages from a small group.
Dating and bounding the data
Data belongs to a period and a condition. Next to every table, write when it was collected, by whom and under what conditions. Undated data invites readers to apply it to situations it never covered.
Gaining originality through sourcing
Using sources does not make an article unoriginal. Google's page lists clear sourcing and evidence of expertise among the things that build trust. What counts is what you do with the source. Practical suggestion: state the finding in your own words, then add why it matters for your reader and how it fits the case you are discussing.
The easiest original contribution appears where two sources disagree. Present the disagreement fairly, explain under which conditions each holds and add your own test if you have one. A long summary of a source with no interpretation does little to answer the question about adding value beyond summarising others.
Making the addition visible on the page
An addition the reader cannot see does not exist. Choose a form that fits the material.
| Form | Fits | Care point |
|---|---|---|
| Photo | Experience and before and after | Write a descriptive alt text for informative images |
| Table | Measured or compared values | Put the method note next to it |
| Quote block | Expert input | Show name and role, keep context |
| Method section | Trials and surveys | State conditions and limits |
| Mistake log section | Failed attempts | Include the fix |
Do not save the evidence for the end. A reader should see it while reading the claim it supports. Mention the addition in the opening paragraph with a sentence that states what you did and its boundary. The title should also reflect the addition, but without exaggeration, because Google's page asks whether a title describes the content accurately and avoids exaggerating or shocking.
How big should the addition be
Size the addition to the piece. A short how-to may need one tested detail and a sentence on its limits. A long comparison may deserve a method section, a table and an expert comment. There is no required quantity. A tiny but real and well-bounded observation helps a reader more than a large body of loosely sourced material.
Writing the addition into the brief
The brief decides whether the addition happens. Method details are in how to write an SEO content brief. Add one field called "addition" and require four answers.
- What is the addition: an experience, an expert, data or an interpreted comparison?
- Why do competing pages lack it?
- Where in the article does it sit?
- Who produces it, by when, before drafting begins?
The fourth answer prevents invention at draft time. If the addition is not ready, hold the article or plan it as a plain guide, and do not label it original.
Worked example: a home renovation blog
Example scenario: A blog written by a homeowner who renovates in stages wants to publish a post on preparing a bathroom wall for tiling. The editor, a friend who reads drafts, reads the first handful of results and finds the same line on every page: make sure the surface is clean and dry. Nobody says how to test dryness, what fails when it is skipped, or how the advice differs between old plaster and new boards. That is the gap.
The brief's addition field has three parts. The first is experience: the blogger keeps a short log for several weeks of what she prepared, what still went wrong and what she changed. Mistakes go into the log, such as a primer coat applied before the plaster had dried. The second is expert input: she asks a professional tiler, someone who prepares walls on site, what actually delays or ruins a tiling day, and gets written answers with approval to quote. The third is a small reader survey with open questions about which preparation step caused the most rework; she records how many readers answered and who they were.
In the post, she places the tiler's observation inside the section on surface checks, next to her own log entry. She sets the "make sure the surface is clean and dry" line out with its condition: it held for some walls in her house and not for others. The survey results sit in a short table with a method note and a plain statement that the group is small and self-selected. A short note under the post gives the date of the log and says she will revisit the post when product guidance changes.
What makes the post different is not its length. It is that every claim is tied to a visible basis and a stated boundary. The blogger did not need a technical term for it; the guidance questions about original information, first-hand experience and value beyond the obvious cover it.
Choosing the addition: a decision table
| Situation | Best addition |
|---|---|
| You can try the thing yourself | First-hand experience with recorded notes |
| The topic needs professional knowledge | A real expert's input |
| Your readers ask the same question in the same way | A small survey or analysis of your own reader questions |
| Sources disagree | An interpreted comparison that explains the disagreement |
| None of these is possible | Hold the topic or publish a plain compiled guide without original claims |
The last row is an honest outcome. If you have no addition, you also have no basis for calling the piece original.
Where AI tools belong, and where they do not
Google's page on generative AI content says these tools can be useful when researching a topic and to add structure to original content, and that all generated content should be manually fact-checked and reviewed before publishing. Experience, expert quotes and data cannot be requested from a model and used as yours. An experience you did not have, written as if you did, is as misleading as a fabricated creator profile. A quote generated for an expert who never said it misleads the reader about who stands behind the claim.
A simple removal test
Take the addition out of the draft and compare what remains with the competing pages. If nothing distinguishes it, the addition is not visible enough or does not exist. If a difference remains, the addition is working. This is an editor's check, not a prediction of how Google will treat the page.
Keeping the addition current
An addition ages like any other part of a page. A trial tied to one season, one product version or one supplier range stops matching reality when that context changes. Put a review date next to the addition, not only on the article, so the person who maintains the page knows which part to recheck.
| Addition | What can go stale | Review action |
|---|---|---|
| Experience | The conditions of the trial | Repeat the trial or label the date clearly |
| Expert input | The expert's view or role | Ask whether the view still holds and note the date it was given |
| Data | The period the data covers | Recollect or mark it as historical |
| Interpreted comparison | The sources compared | Check whether sources changed or disagreement was resolved |
Change the update date only when you actually changed something. Google's helpful content page lists changing page dates to seem fresh, without substantially changing the content, among its warning signs.
Credit and correction
If a reader points out an error in the data or a quote, correct it promptly and add a short correction note at the end of the post. A visible correction is a trust signal. A silent edit leaves readers who cited the earlier version without an explanation.
Pre-publication checklist
- The brief has an addition field, and competing pages were checked for it.
- Experience was recorded before drafting, with its limits stated.
- Any expert is real and identifiable, and approved the quote.
- Data comes with method, date, conditions and limits.
- Sourced material is interpreted, not copied.
- "Information gain" is not presented as an official concept.
- Every number and quote matches its source.
Mistakes to avoid
Quoting an expert without saying what changed in the text
A quote that could be deleted without altering any claim is decoration. Show which sentence, number or caveat in the article exists because of the expert.
Burying the experience log in an appendix
Evidence placed at the end of the page is easy to miss. Put each observation beside the claim it supports.
Promising original research in the headline
A label such as "our research" sets a standard the article must then meet. If the work was a short trial, call it a short trial.
Treating one survey as settled fact
A small, self-selected group describes itself. Report the count, the method and the limit, and avoid turning the result into a rule for every reader.
Forgetting that a trial has a shelf life
A test tied to one product version or one season stops matching reality when the context changes. Record the date and schedule a review.
Frequently Asked Questions
Does Google use the phrase "information gain"?
No. The phrase is not in Google documentation; it is informal vocabulary used in the SEO community. Google pages ask whether content offers original information, reporting, research or analysis.
When is an expert quote worth seeking?
No, the documentation sets no such requirement. An expert quote makes sense when the topic needs professional knowledge and the quote adds concrete credibility.
Can a small reader survey be original data?
It does if you describe the method openly. State the sample size, the conditions and the limits, and do not present the result as a general rule.
May an AI tool supply the expert voice?
No. A generated quote does not belong to a real person and misleads readers. Google describes fabricated creator profiles as deceptive.
Is reading rival articles a form of copying?
No. The Starter Guide advises against rehashing what others have published, so reading competitors is for finding the gap, not for carrying their wording over.
What if the addition misses the deadline?
Hold the article or publish it as a plain compiled guide without labels like original research. If the topic matters, wait until the addition exists.
Conclusion
Standing out on a crowded topic is a production decision, not a wording trick. Pick one addition that competing pages lack: your own recorded experience, a real expert's input or a small body of data with an open method. Put it in the brief, produce it before drafting, give it a visible section and state its limits. Google's pages supply the questions on original information, first-hand experience and value beyond summarising others, while "information gain" is not an official term. The choice of addition and the production routine are yours to adapt.
For your next article, read the first handful of competing pages, note the shared line and pick one addition that tests or extends it. If you want a team to produce articles with that kind of addition, see our content writing service.
Post a comment
Please log in to leave a comment. Login