Writing Product Descriptions That Rank and Still Convert
Everyone wants to know how to write SEO product descriptions, and almost everyone gets bad advice about it. They're told to pick a keyword, stuff it into a paragraph, and call it done. Or the opposite: write pure sales copy and hope Google figures out what the page is about. Both camps are half right, which is exactly why the confusion sticks around. I scaled a DTC brand from $100K to $3M+ largely on the back of product pages that did both jobs at once. It's not magic. It's a template.
"Keyword density is what makes a product description rank"
This one comes from old SEO advice that technically worked in 2012. Back then, repeating a phrase 3-4% of the time could move a page up. Google's language models don't work that way anymore. They understand synonyms, intent, and context. A product description that says "running shoes" six times in 150 words reads like spam to a human and increasingly to the algorithm too. The kernel of truth: your primary keyword still needs to appear, clearly, early, and naturally. Once, in the first sentence. Maybe once more in a bullet or spec line. That's it.
"SEO copy and sales copy are different documents"
I hear this from both SEO people and copywriters, which tells you something. SEO folks write dry, keyword-first descriptions that read like a spec sheet. Copywriters write punchy, benefit-driven copy that never mentions what a human would actually type into Google. Both are leaving money on the table. On the infrastructure project I marketed, a $2.2B build, we had technical spec pages that nobody could find because they were written for compliance reviewers, not search intent. We rewrote them around how actual procurement teams searched, and organic traffic to those pages tripled in four months. The copy didn't get worse. It got found.
"Long descriptions always outrank short ones"
Length correlates with ranking in aggregate studies, sure. But correlation isn't a writing instruction. A 400-word product description for a $15 phone case is not going to outrank a tight 120-word one that matches intent better. What matters is whether the page answers the actual questions a buyer has: what is this, who is it for, how is it different, does it work with what I already own. If you can answer that in 120 words, 400 words is just friction before the add-to-cart button.
What actually matters instead
Write for the person searching, then check that the words they'd search for are actually on the page. Here's the template I've used across dozens of product lines:
- Line 1 (the hook): One sentence with the primary keyword and the core benefit. Example: "This merino wool travel blanket keeps you warm on flights without the bulk of a regular throw."
- Line 2-3 (the differentiator): What makes this version different from the ten others like it. Specific, not vague. "Packs down to the size of a water bottle" beats "compact and convenient."
- Bullet list (the specs and secondary keywords): 3-5 bullets covering material, size, use case, compatibility. This is where secondary keywords live naturally, because specs are exactly what people search for: "machine washable," "fits carry-on," "100% merino wool."
- Closing line (the objection killer): One sentence addressing the thing that stops people from buying. Shipping time, return policy, fit concerns.
That's four parts, usually 100-180 words total. It covers search intent in the first line, differentiation in the second, keyword coverage in the bullets, and conversion psychology in the close. None of it feels like it was written for a machine, because it wasn't. It was written for a person, with enough discipline to make sure the words that person types into Google actually show up on the page.
The most common mistake
People write the sales copy first, then try to "SEO it" by inserting keywords afterward. It almost always reads clunky, because you're forcing phrases into sentences that weren't built for them. Do it in the other order. Know your primary and secondary keywords before you write line one. Let the hook sentence be built around the keyword from the start, not patched later. It takes the same amount of time and the result doesn't sound like two different writers fought over the page.
One more thing worth checking: pull up Google's autocomplete and the "People also ask" box for your product category before you write anything. Those are free keyword research tools nobody pays for. If "waterproof" keeps showing up next to your product type and your bullets don't mention it, you're missing search volume for no reason.
Takeaway: write the hook for the human, build the bullets for the search engine, and never let either job replace the other on the same page. Test one rewritten product page against your current best performer for 30 days. Compare both organic sessions and conversion rate, not just one or the other. If you're only tracking rankings, you'll optimize yourself into pages that get traffic and never sell anything.