Discussion starter: a content SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- Content SEO
Briefs built from SERP analysis tend to produce copycat content. Have you tested briefs built from support tickets or sales calls against them?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
Short answer: no controlled public test that I can point to compares SERP-built briefs with briefs built from support tickets or sales calls, so treat any claim that one method “wins” as unproven. The mechanism is still worth acting on. A brief built from the current top results samples what competitors have already published, so drafts converge on it. A brief built from tickets and calls samples what customers actually ask, including the questions those results don’t answer.
My working position is a hybrid: use the SERP to decide intent and format, and customer evidence to decide what the page says. Then test that hybrid against your current process on your own topics, because the result is likely to differ between a help-centre topic, a comparison page and an informational head term.
Why SERP-built briefs converge
A SERP-built brief lists the headings, questions, entities and word counts of the top ten results and asks the writer to cover them. Content scoring tools then reward overlap with those pages. Each step is reasonable on its own; together they produce a remix of the current results. Google’s guidance on helpful content pushes the other way. It asks whether content provides “original information, reporting, research, or analysis”, whether it offers “insightful analysis or interesting information that is beyond the obvious”, and whether, when it draws on other sources, it avoids “simply copying or rewriting those sources” and adds “substantial additional value and originality”. It also lists mainly summarising what others say, without adding much value, as a warning sign of search-engine-first content.
Building a brief from the top ten results is like designing a restaurant menu by reading the menus of the ten busiest restaurants on the street. You learn what diners expect to see, and you end up serving the same dishes. Reading your own complaint cards, and the questions your waiters get asked, tells you what diners wanted and couldn’t find. In research terms, SERP analysis is supply-side: it samples pages that already exist. Tickets and calls are demand-side: they sample problems as customers state them, before anyone has written the answer. That difference matters more now that AI answers can assemble the consensus version of a topic from the same pages your brief was built on.
What tickets and calls add, and what they get wrong
Source What it’s good for Blind spots Pages it suits Support tickets Real problems, exact error messages, edge cases, the steps that confuse people, customer vocabulary Existing customers only; your product’s own bugs; phrasing that doesn’t match searches; personal data Help-centre articles, troubleshooting guides, integration how-tos, “why does X happen” pages Sales calls Objections, competitor comparisons, pricing and implementation worries, security and procurement questions Late funnel; only prospects who reached a call; low search volume; confidential deal details Comparison and alternatives pages, pricing explainers, implementation and migration guides, security pages SERP analysis Intent, format, what searchers expect to see, table stakes Converges on competitors; blind to anything nobody has written yet Informational head topics, category and definition pages A practical gap that is easy to miss is language. People describe a problem to support differently from how they search for it. A ticket says “the export button does nothing since the update”; the search is closer to “csv export not working [product]”. Map each ticket or call cluster to search data before committing: Search Console queries for your help centre and blog, your site-search logs and a keyword tool. A cluster with many tickets and no search demand may still deserve a help article (for ticket deflection, and so AI answers about your product have accurate material to draw on), but it is not an acquisition topic.
The closest public evidence is a practitioner case study, not a controlled test. The agency Grow and Convert documents a “pain point SEO” approach that starts “with the intent of a buyer (the pain point of a customer)” and then finds keywords. Using one SaaS client’s analytics, it reports that the post with the most conversions in that report targeted a keyword SEO tools showed as having zero search volume, and that while the post’s traffic didn’t compare with others on the list, it “dwarfs most others from a conversion standpoint”. That is one agency’s selected examples, and case studies select for strong results, so read it as evidence that customer-led topics can convert well, not as a comparison of brief methods.
A hybrid brief that avoids both failure modes
- Intent and format, from the SERP. What type of page ranks, what searchers expect to see, and the few points every good result covers. Keep this section short so the writer doesn’t reproduce it at length.
- Customer evidence. Verbatim questions and objections from tickets and calls, anonymised, with counts and the customer segment they came from.
- Gaps. What customers ask that the current top results don’t answer. This is the page’s reason to exist; if the section is empty, question whether the page should be written.
- Proof. Data, screenshots, anonymised examples, and a named person from support, sales engineering or product to interview.
- What to leave out. The consensus material the page will cover in a sentence or link to.
- Next step. What the reader should do after reading, based on what the calls show happens next in real deals.
Add one check before publishing: compare the draft’s subheadings with those of the top five results. If most of them map one-to-one, the brief produced a copy, whatever its sources were. Tracking that overlap share by brief type gives you a process metric long before rankings arrive.
Watch for this:
Tickets and call recordings contain personal data and sometimes confidential commercial details. Anonymise extracts before they go to writers, freelancers or AI tools, check what your privacy policy and call-recording consent allow, and get sign-off before publishing anything that describes a specific customer’s situation or an unresolved product issue.
How to test brief types without fooling yourself
The unit of the test is the topic, not the page, because publishing two pages on one topic creates the cannibalisation you’re trying to avoid. A workable design:
Build a list of 30 to 60 topics you would write anyway↓Pair topics by expected demand, difficulty and funnel stage↓Within each pair, randomly assign one topic to the SERP brief and the other to the hybrid brief↓Same writers across both arms, and the same editing, internal linking and promotion rules↓Read results at 30, 90 and 180 days; decide on the 180-day readIf you would rather test on existing content, take SERP-built pages that underperform and rewrite half of them with customer evidence. Give the other half an equal-effort refresh from a new SERP brief rather than leaving them alone. A refresh can move a page by itself, so “customer rewrite versus no change” would credit the brief type with the effect of simply updating the page.
Metric Why it matters Where to get it Non-branded clicks, and the number of distinct queries with impressions Traffic, and whether customer language reaches long-tail queries the SERP brief missed Search Console Sign-ups, demo requests and assisted conversions The business case, especially for commercial topics GA4 or CRM Subheading overlap with the top five results Whether the brief changed the content at all Manual or scripted check Sales and support use Whether reps send the page, and whether tickets on the topic fall CRM and help desk AI visibility by page Citations or AI impressions, with no click data Bing’s AI Performance report (page-level citations); Google’s “Search generative AI performance” report (impressions only) The long window has two reasons. New pages have to be discovered and indexed, and can take months to reach stable positions, so a 30-day read mostly measures indexing speed. And commercial conversions are sparse: with 20 or 30 topics per arm, a few extra demo requests can swing the result. Decide the primary metric before publishing, compare the differences within each pair rather than only the averages, and expect noise. For the retrofit version, start the clock when URL Inspection shows the updated page has been recrawled. If a core update lands mid-test, both arms are exposed to it, which is the reason paired randomisation is worth the effort.
What would change the decision:
- Hybrid wins on conversions and roughly matches on clicks: make the customer-evidence section mandatory in every brief.
- Hybrid wins on conversions but loses clicks on informational head topics: split by funnel stage. Keep SERP-led structure with a required customer-evidence section for head topics, and lead with customer evidence for commercial ones.
- No difference on anything: the brief isn’t your bottleneck. Look at writer expertise, access to subject experts, or distribution.
- Hybrid loses on both: check whether the ticket-sourced topics had search demand at all. The mapping step is the first place I would look.
Hypothetical example:
A B2B software company plans 40 topics, pairs them by difficulty and funnel stage, and publishes 20 per arm over a quarter. At 180 days, the hybrid pages have similar median clicks, appear for more distinct queries and generate more demo requests. Two hybrid topics built on ticket phrasing with no search demand get almost no traffic, but one of them cuts tickets on that issue. The team keeps the hybrid brief, adds a demand check to the ticket-mining step, and moves the two low-demand pages into the help centre.
Where I would start
If this were my team, I would not frame it as SERP briefs versus customer briefs. I would shrink the SERP section of every brief to intent, format and table stakes, make the customer-evidence and gaps sections mandatory, add the subheading-overlap check before publishing, and run the paired test for two quarters. That tells you whether the extra research time pays for itself on your topics, which no public test can tell you yet.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.