In March 2026 we started publishing technical articles about pipe couplings. Not marketing pages — reference material: dimensional charts, standards explainers, material selection guides, failure-mode write-ups.
The premise was simple. David Phee Enterprise has traded industrial pipe couplings from Singapore for three decades, and most of what accumulates in a business like that never gets written down. Meanwhile the questions that arrive by email are remarkably consistent: what is the OD for this DN, will this coupling work on that pipe, which gasket for this fluid, what certificate do I need.
Those are answerable questions. So we started answering them in public.
This pipe coupling SEO case study is an honest review of what happened — what worked, what did not, and what we would do differently. It is written for anyone in a technical trade considering the same thing.
What We Actually Published
The numbers, as at 1 October 2026:
| Metric | Figure |
|---|---|
| First article published | 23 March 2026 |
| Articles live | 59 |
| Articles drafted and scheduled beyond that | 40+ |
| Publishing cadence, current | 3 per week (Mon / Wed / Fri) |
By category, the live articles break down as:
| Category | Articles |
|---|---|
| Pipe Couplings | 23 |
| Standards & Specs | 16 |
| Maritime Supply | 7 |
| Water Infrastructure | 4 |
| Oil & Gas | 4 |
| Expansion Joints | 2 |
| Product Guides | 2 |
| Others (Repair Clamps, Buyer’s Guides, JWC, Romacon) | 5 |
A few articles sit in two categories, so the rows add up to slightly more than the total.
That distribution was not planned in advance. It is the result of following which questions actually recur, and it tells you something: more than a quarter of everything we wrote turned out to be about standards and specifications, which is not where we expected the centre of gravity to sit when we started.
Five Things That Worked
1. Answering the dimensional questions first
Our earliest and most durable articles were the least glamorous: pipe OD versus nominal bore, size charts, flange dimensions. These are questions with exact answers, asked by people who need the answer now, and they age well — a DN50 pipe will still be 60.3 mm OD in five years.
If you are starting from zero in a technical trade, start here. Not with “why choose us.”
2. Writing the honest answer, including when it does not favour us
The article that best represents this was on HDPE couplings. The correct answer to “can I use a standard mechanical coupling on HDPE?” is no, not the one on your shelf — and, further, that on a long straight buried PE run fusion is the better engineering answer and a mechanical coupling is not.
That is a real limitation of the category we sell into, stated plainly. Buyers in this industry have functioning instincts for being sold to, and an article that concedes the boundaries of its own product is more credible than one that does not.
3. Grounding everything in an internal knowledge base
Every technical article is written against an internal reference vault — standards research, supplier documentation, material data, failure modes — rather than from memory or from whatever a search turns up.
This has two effects. Accuracy, obviously. But more importantly it makes the writing fast, because the research is already done and cited. The bottleneck in technical content is almost never the writing; it is verifying that what you are about to publish is true.
4. Refusing to publish certain things
Two rules we set and kept:
- No public pricing. We drafted a pipe coupling price guide and dropped it before publishing. It would have attracted traffic. It would also have handed our pricing position to every competitor in the region. The traffic was not worth it.
- Real product photography, not generated imagery, for product articles. Composites from our actual catalogue rather than plausible-looking renders of couplings that do not exist.
Both rules cost us something. Both were right.
5. Three per week, sustainably
We experimented with daily publishing to clear a backlog. It worked as a catch-up mechanism and was not sustainable as a steady state. Three per week — Monday, Wednesday, Friday — is the cadence we settled on and have held.
Four Things That Did Not Work
1. Batching drafts without batching images
Our single biggest self-inflicted delay. In early July we had ten articles fully drafted and none published, because the publishing script requires a featured image and none of the ten had one. The writing was done. The pipeline was blocked on a step nobody had scheduled.
Seven days of publishing were lost to this. Treat the image as part of the draft, not as a separate later task.
2. No automation on a session-driven pipeline
Related, and the root cause. The publishing pipeline is run by hand. When a working session does not happen, nothing publishes, and there is no alert to say so — the gap is only visible when somebody looks at the calendar. A scheduled process would have caught it on day one instead of day seven.
3. Topics that turned out not to exist
We had an article scheduled about a specific product series that, on investigation while drafting, does not appear in the manufacturer’s public catalogue at all. It was dropped. Another was dropped on commercial grounds after drafting had started.
Two slots out of roughly fifty-seven, lost to insufficient validation at the planning stage. Not disastrous, but both were avoidable with a ten-minute check before a title went into the calendar.
4. Optimisation as a trailing task
Content-scoring checks kept sliding to “before publish date” and then past it. A backlog of articles live but never scored accumulated. The lesson is structural rather than motivational: if a step is not in the definition of done, it does not happen.
What We Would Do Differently
- Definition of done includes the image and the optimisation check. Not drafted-then-images-then-scoring. One unit of work.
- Validate the topic before it enters the calendar. Confirm the product, standard or series actually exists and is public.
- Automate the publish trigger, or at minimum alert on a missed slot.
- Start with the dimensional and standards questions. They have exact answers and they recur constantly, and we got to them by accident rather than design.
- Decide the no-publish rules up front. Pricing, competitive intelligence, customer names, supplier commercial terms. It is much easier to hold a line drawn before there is traffic at stake.
The Honest Summary
Six months of technical publishing has produced a reference library that answers the questions we were already answering one email at a time. That is the actual return, and it would have been worth doing even if search had delivered nothing.
The thing we underestimated is how much of the work is verification rather than writing. Getting a dimension, a velocity limit or a certification requirement right takes far longer than putting it into a paragraph — and in a technical trade, getting it wrong is worse than not publishing at all. A single incorrect OD in a table would do more damage to the site’s usefulness than a month of missed posts.
If you are considering the same thing in a technical field: write the boring exact answers, ground them in something you can cite, and be willing to say when your own product is the wrong choice.
David Phee Enterprise supplies industrial pipe couplings, repair clamps and expansion joints across Singapore, Malaysia and Indonesia from our Kaki Bukit HQ. For technical enquiries and quotations, visit www.davidphee.com.