A searcher typing "best standing desk" or "desk A vs desk B" has narrowed the field but not chosen. A page that serves this person organizes a decision: stated criteria, evidence you gathered, strengths and weaknesses for each option, a note on who each option suits and a closing section that says what to pick under which circumstances. A list of ten desks with affiliate-style blurbs does none of that. After the opening answer, the page should earn trust one claim at a time.
What a "best" or comparison query asks of a page
Think of the searcher as someone with a budget, a room and a back that complains. They need to learn which differences matter and which option fits their situation. The questions they carry are simple: which criteria count, how do the options score on them, and which one is right for me?
Reading the wording
| Query wording | What the searcher wants | Page shape |
|---|---|---|
| best standing desk | A short list with reasons | Criteria-led list with selection notes |
| desk A vs desk B | The differences side by side | Two-column table and a verdict |
| standing desk alternatives | Other routes to the same goal | Option groups, each tied to a use case |
| standing desk review | What it is like to live with | First-hand review |
| how to choose a standing desk | The criteria themselves | Selection guide, products secondary |
Table: query wording and page shape. These pairings are a working hypothesis. Confirm them against the result types on page one before you commit, using the steps in the article on matching search intent in articles.
A word of caution about the label
"Commercial investigation" is a term used in the SEO trade to describe this stage. Google's documentation does not define it as a category. Treat it as shorthand for your own planning and let the results page, not the label, decide the format.
Evidence that carries a recommendation
Google's page on writing high quality product reviews asks for specific kinds of proof. It advises you to provide evidence such as visuals, audio or other links of your own experience with what you are reviewing. It also advises sharing quantitative measurements about how something measures up in various categories of performance, and discussing benefits and drawbacks based on your own original research. Those three habits carry the whole comparison page.
Matching claims to proof
| Claim | Proof that supports it | If you have no proof |
|---|---|---|
| The frame stays steady at full height | Short video of the desk at height with a loaded monitor arm, plus a note on the load | Do not claim it, or attribute the statement to the maker |
| The motor is quiet | A recording with the room and the distance described | Describe what you heard, not what the maker says |
| Assembly is simple | Time taken, tools used, a photo of the parts list | List the steps and leave out the verdict |
| The memory presets work well | A short clip of recalling a preset and a note on any drift | State only that the feature exists, citing the maker |
Measuring versus impressing
A measurement can be repeated by someone else. Say what you measured, with what, under which conditions and how often. An impression such as "feels sturdy" is a judgment, and it belongs in a sentence that says so. Example scenario: a reviewer places the same loaded monitor arm on three desks, raises each to its top setting and records a ten second clip of each. The clips become the proof for a row in the table, and the loading is stated in the note under it. If you have no such records, leave the row out.
When you cannot test everything
You may not get every desk in your hands. Narrow the page instead of bluffing. Say which desks you tested, label specifications for the others as the maker's, and do not present a desk you did not use as part of your test. Google's advice on recommending something as best is to include why you consider it the best, with first-hand supporting evidence. A "best" claim without that support does not meet the standard.
A testing protocol for standing desks
Write the protocol before the first test and keep it fixed. For a standing desk, a protocol can include the following steps.
- Assemble each desk yourself and record the time, tools and any step that needed a second person.
- Load the same monitor arm and the same display on every desk, and note the clamp position.
- Raise each desk to its top setting, type for a fixed period and note any motion in the surface.
- Raise and lower the desk through its full range and record the noise with the recorder at the same distance each time.
- Use the memory presets for a week and note any drift in height.
- Photograph the cable routing, the underside and the controller.
The protocol makes the comparison fair, because every desk meets the same conditions. It also makes your page repeatable: another reviewer can follow the same steps and compare results with yours.
Capturing photos and clips that carry evidence
A photo proves more when it shows something a reader can check. Include a ruler or a tape in the frame for height claims, show the clamp position for stability claims and keep the camera angle the same for every desk. Write one line under each image describing the setup. Practical suggestion: favor original photos taken under the protocol over stock-style hero shots, which show little about how a desk behaved under your test.
Picking the criteria
Google advises focusing on the key decision-making factors, based on your own experience or expertise, and describing key design choices and how they affect users in practice. The factors differ by product. For a standing desk, candidate factors include frame stability at height, motor noise, height range, weight capacity, collision handling, cable management and what the warranty covers. Select the ones your reader's decision turns on.
Keep the list short
Practical suggestion: choose five to seven criteria. A longer list makes the table unreadable. For each criterion state why it matters, how you assessed it and who should weigh it more heavily.
The spec sheet is not the practice
A maker lists a lifting speed or a range. Your job is to describe what that does in use. Describing a range of travel means little until you say whether a seated user of a certain height could reach a comfortable position at both extremes. Translate each design choice into a consequence the reader can picture.
Point to other useful resources
Google's page also suggests including links to other useful resources, yours or other people's, to help readers decide. A short note pointing readers to an ergonomics checklist or a sizing chart that you trust can serve them better than reproducing it badly.
When one product disappoints
If a desk you tested disappointed you, decide whether it belongs in the page. A product that fails a criterion your reader cares about can stay in the page as a clearly labelled "not recommended for" entry, with the proof. Dropping it silently hides information the reader may want, because they might already be considering that desk. Whichever way you go, write the reason in a sentence.
Writing for a reader who is choosing
The same page asks you to explain what sets something apart from its competitors, to cover comparable things to consider and to explain which might be best for certain uses or circumstances. In practice, every option needs a use case instead of one global score.
Four lines per option
- Who should consider this desk: room, height range, weight to support.
- What it did well in your tests, with the proof.
- Where it struggled, with the proof.
- Who should look elsewhere, in one sentence.
The last line is the one ranked lists skip, and it is the line that earns trust. Using the same four lines for every option also makes the table easy to fill.
Qualify weaknesses by scope
Do not soften a weakness and do not inflate it. Tie it to the test: "the desk wobbled when the monitor arm was clamped to the rear edge at top height". A scoped sentence is accurate. A sweeping one overreaches what a single test supports.
The weakness that depends on the reader
Noise matters little to someone who works in a shared studio with music and a lot to someone in a quiet study. Next to each weakness, add a short note on who is affected. The judgment then rests with the reader, and your page does not assume a single ideal buyer.
Who wrote it, how and why
Google's helpful content page frames three questions. For who, it asks whether it is self-evident to visitors who authored the content and whether pages carry a byline where one might be expected. For how, it asks whether the use of automation, including AI generation, is self-evident to visitors through disclosures or in other ways. For why, it asks why the content is being created in the first place.
A method note on the page
Add a short method note near the top or the end: which desks were used, where, for how long and judged by which criteria. A method note shows readers how reliable the evidence is. It does not have to be long, but it should allow a reader to judge whether your setup resembles theirs.
Show the author
Show the author's name and a line about relevant experience. The line should be true and specific, such as the desks the author actually used. Padding it with credentials that do not bear on the product gives readers nothing to rely on.
Ranked lists that stand on their own
The review page says that ranked lists should contain enough useful content to stand on their own, even if you also publish in-depth single reviews. It also suggests considering links to multiple sellers so that readers can buy from the merchant of their choice. Taken together, the list should be a page a reader can decide from, and not a bridge to somewhere else.
Explain the order
If you rank, say what you ranked by. "Number one" means nothing without a criterion. Practical suggestion: consider groups in place of an order: "for small rooms", "for heavy monitor setups", "for the lowest outlay". Groups make plain that no single winner exists for everyone. The article on writing best-of listicles with disclosure covers independence and notices in more depth.
Organize the seller links
If you link to more than one seller, collect the links in one consistent place for each desk. Let the reader see which seller offers which terms, and leave the choice to them. If a desk is sold by one seller only, say so plainly.
The opening paragraph
The opening of the page carries four facts: a short answer (which desk suits which reader), how many desks you tested and for how long, the criteria used and the date of the last update. These four facts give a reader the answers to the helpful content questions before scrolling. Example scenario: "If you work in a small room with a single monitor, Desk B fits; if you mount two heavy monitors, Desk A is the safer choice. We used three desks for two weeks in the same room and compared them on five criteria. Desks we did not test are not on the list."
Openings to avoid
Skip the history of the sit-stand movement, general statements about sitting and a list of reasons to buy any desk at all. The reader came with a question about choosing. Spend the opening on that question.
A comparison table and a decision section
A comparison table puts criteria in rows and options in columns. Fill the cells with evidence-based observations and not empty judgments like "good". The decision section that follows turns the table into plain advice.
| Criterion | Desk A | Desk B | Desk C |
|---|---|---|---|
| Stability at top height (your test) | Clip and note | Clip and note | Not tested |
| Noise during travel | Recording and distance | Recording and distance | Maker's statement |
| Height range for your reader | Measured travel | Measured travel | Maker's figure, labelled |
| Assembly | Time, tools | Time, tools | Time, tools |
| Who it suits | One sentence | One sentence | One sentence |
Decision section template
- If you share a small room and relocate the desk: Desk B, with the reason and the proof.
- If you mount heavy monitors: Desk A, with the reason and the proof.
- If keeping the outlay low is the priority: Desk C, with its trade-off stated.
- If none fits: the next question to ask before buying.
Keep the table readable
Practical suggestion: use no more than four columns. For more options, group them and give each group its own table. After every table, add a short summary of what it shows and who should look at which column.
From last year's model to this one
The review guidance also asks you to describe how a product has evolved from previous models or releases. If you own the older desk, write a short difference table: what changed, what stayed and whether the change can be felt in daily use. If you do not own it, report the maker's announcement as the maker's claim and keep it out of your own test results.
A rule for what you may claim
Decision rule: put a claim on the page only when you could show its proof today. Show it if you can. If you cannot, attribute the claim to the maker or cut it. Calling something best means naming the use and the reason, with first-hand evidence behind both. When the evidence is thin, group the options and drop the ranking.
How far a claim may reach
You do not have to test everything. The rule forbids presenting what you did not test as if you had. If you tested three desks, title the page after those three and limit "best" to that set. The title tells the reader where the claim applies, and the rest of the page earns trust from there.
Standing desks, step by step
Example scenario: an editor prepares a page for "best standing desk". Page one, read with the steps from the guide on matching intent, shows short criteria-led lists. The editor picks five criteria, tests three desks for two weeks in one room, and records each desk with photos and short clips. Two other desks stay out of the list because the editor could not test them, and a short note says so.
| Step | Action | Output |
|---|---|---|
| 1 | Read page one | Short criteria-led lists as the format |
| 2 | Choose criteria | Five criteria, each with a reason |
| 3 | Run the tests | Photos, clips, notes with conditions |
| 4 | Write four lines per desk | Use case, strength, weakness, who should look elsewhere |
| 5 | Build the table | Evidence cells, untested cells labelled |
| 6 | Write the decision section | Three advice lines for three buyer types |
| 7 | Add method and author notes | Trust information for the reader |
The photos the editor leaves out
The editor could have used the makers' product photos. Those photos would not provide the kind of first-hand evidence Google asks for, and they would add nothing a reader could not see on the maker's own page. The editor uses original photos and writes the conditions under each one.
Where comparison pages lose trust
Passing off a spec sheet as testing
Echoing the maker's feature list is not a review. Write what each feature did when you used it.
Showing only upsides
A page with no weaknesses invites doubt. Give each weakness its scope.
Numbering options with no criterion
If the order does not follow a stated criterion, swap it for groups.
Repeating one verdict across every desk
Five desks with the same sentence make a tidy table and an empty one. Give each desk its own observation.
Letting seller buttons crowd the reading
A buy link after every paragraph turns a decision aid into an advertisement. Keep links beside the table and in the decision section.
Stretching one room to cover them all
A test in one apartment does not describe every home. Name the setting, such as a small room with a hard floor, and let readers judge how close it is to theirs.
Padding for length
Google's page points to the quality and originality of a review, not its length. Extra words do not stand in for missing proof.
Keeping the page honest over time
A comparison page does not stand alone. Link each option to its own review if one exists, to the selection guide and to the relevant product category. Show a visible update date and review the page when models change. Remove desks that have left the market or mark them as discontinued. When you retest a desk, write the new result next to the old one with both dates, so readers see that a changed verdict has a reason. If a maker releases a revised version with the same name, treat it as a new product in your notes and test the parts that changed.
A short change log
Add a short log at the end: date, change, reason. It shows readers the page is maintained and helps you track what changed. Change the date only when the content changes.
Where the guidance comes from
Google's page on writing high quality product reviews (Opens in a new tab) and its guidance on helpful, reliable, people-first content (Opens in a new tab) are the primary references. Google documentation can change, so check the current page before you act.
Final read-through
Checklist: work through these in order before the page goes live.
- Do the author, method and update date show on the page?
- Does the layout match what page one shows for the query?
- Are the test conditions written down for every desk?
- Do five to seven criteria each have a stated reason?
- Does each strength and weakness carry proof, or a label naming the maker as the source?
- Does every option end with a line on who should look elsewhere?
- Are untested table cells labelled rather than left empty?
- Is your own evidence kept apart from the maker's claims, including any older-model comparison?
- Do the seller links leave the buying choice with the reader?
- Could someone decide from this page alone?
Frequently Asked Questions
Is "commercial investigation" a Google category?
No, it is a label from the SEO trade. Check the format against page one for your query rather than relying on the name.
What backs up calling a product the best?
The review guidance asks you to say why you consider it the best, with first-hand supporting evidence, whether you recommend it overall or for a certain purpose.
Can I list desks I did not test?
You can, if you label what you say about them as the maker's statement. The alternative is to leave them off and say so in a short note.
Do seller links have to be plural?
Google suggests considering links to multiple sellers so readers can buy from their merchant of choice. It reads as a suggestion, not a requirement.
How many criteria fit in a comparison table?
Practical suggestion: five to seven, so the table stays readable.
When is grouping better than numbering?
Numbering works when one stated criterion drives the order. When several criteria matter, grouping by use case reads more clearly.
Conclusion
A "best" or comparison query asks the page to help someone choose. Confirm the format on page one, set five to seven criteria with reasons, back every claim with proof or label it as the maker's, describe each option by who it suits and who should look elsewhere, and close with a decision section. Keep the author, method and update date in view.
If you want your comparison or buying guide turned into a finished article, have a look at our content writing service.
Post a comment
Please log in to leave a comment. Login