Project · Labor-law compliance posters, e-commerce
The pages were selling posters. The searches were asking questions.
A US compliance catalogue that had been losing ground for a year. The demand sat one step earlier than the product — an employer looks up what their state’s minimum wage is long before they look for a poster — and the pages that answered that did not exist.
The problem
A compliance catalogue is a hard thing to grow. The product terms are few, expensive and already owned — everyone selling a federal poster is selling the same federal poster, and the sites that got there first have the links. Competing head-on for labor law posters is a spending contest, not an SEO problem.
The site had also been losing ground for about a year before I took it over in March 2026. Ranking keywords had fallen roughly a third from their 2025 peak, and organic traffic with them. That is rarely one dramatic cause. It is a catalogue that has stopped being the best answer to anything in particular.
The impressions were already there. Search Console showed the site being served for wage and compliance questions, then sitting at positions nobody clicks, on pages that were never written to answer them. That is a different job from ranking. It is the job of putting the answer on the page.
What I found
- The impressions and the intent were on different pages. The URLs collecting question traffic were not the URLs that could sell anything, and nothing linked the two.
- The largest informational demand had nothing against it. State minimum-wage queries are the top of this funnel, and the catalogue had no page built to answer them.
- An entire commercial cluster was missing. Nothing addressed digital and electronic posters. Three of those terms now hold 12.5% visibility each, from a standing start.
- The template was doing the damage at scale. Defects on the state-page template repeated across every state, so a single fix was worth more than any individual page rewrite.
What I did
- Read the Search Console query data first and worked the list by where the site already sat. Terms with impressions and positions in the teens get fixed before terms with nothing, because that is where the same effort pays soonest.
- Rewrote the priority pages so the answer came before the pitch, and pointed the internal links at them.
- Built the landing pages the catalogue did not have: the state minimum-wage guides, and the digital and electronic poster cluster.
- Fixed the state-page template rather than the state pages, so the correction propagated across the whole set at once.
- Tracked 328 commercial keywords in Semrush so movement was measurable rather than anecdotal, and read it against Search Console rather than instead of it.
The result
Across a tracked set of 328 commercial terms, between 20 August and 12 September 2026:
1–34–1011–2021–100Outside top 100
| Tracked set, 328 keywords | 20 Aug 2026 | 12 Sep 2026 | Change |
|---|---|---|---|
| Positions 1 to 3 | 18 | 49 | +31 |
| Positions 4 to 10 | 31 | 30 | −1 |
| Positions 11 to 20 | 49 | 53 | +4 |
| Positions 21 to 100 | 91 | 81 | −10 |
| Outside the top 100 | 139 | 115 | −24 |
| Ranking anywhere in the top 100 | 189 | 213 | +24 |
The two bands that shrank did not lose anything. Positions 21 to 100 drained upward into 11 to 20 and above, and 4 to 10 drained into the top three faster than it refilled. Over the slightly wider window Semrush counts 158 keywords improving against 43 declining, and the two reports reconcile: 37 new in the top ten minus 7 lost is the same +30 the distribution shows.
| Tracked set, 13 Aug to 12 Sep 2026 | Start | End | Change |
|---|---|---|---|
| Visibility | 1.62% | 15.06% | +13.44 pts |
| Average position | 97.39 | 47.06 | ↑50.33 |
| Keywords improved | — | 158 | vs 43 declined |
| New in the top 3 | — | 35 | 6 lost |
| New in the top 10 | — | 37 | 7 lost |
The pages that did it
The pages carrying the result are the ones the diagnosis predicted, which is the part that matters. The minimum-wage guide went from an estimated 648 visits to 1,163. A second wage guide went from nothing at all to 147. The digital-poster cluster went from no visibility to three terms at 12.5% each, and two terms — alabama minimum wage and digital labor law poster — arrived from outside the top 100 altogether, a 99-place gain on Semrush’s scale.
The ranking figures come from a Semrush Position Tracking campaign on a keyword set I chose, measured the same way on both dates — those are the result. The traffic curve is a different kind of number: Semrush’s modelled estimate of organic visits, not analytics. Google Search Console for this domain records roughly a third of what Semrush estimates across the same months — the ordinary gap between a model and a measurement, not a contradiction. The shape is the same in both, and the shape is what this curve is here for. Where an estimate and a measurement disagree, the measurement wins, and neither belongs on a portfolio page without being told apart.
Where it stands
Two-thirds of the tracked set now ranks somewhere in the top 100, and a seventh of it ranks in the top three. 115 keywords still sit outside the top 100, which is where the next quarter goes.
It is not a victory lap. Over the same thirty days two competitors grew visibility faster than this site did, and more significantly the US Department of Labor has started moving into these queries, growing visibility by 35.19%. A government source is a hard thing to outrank on a question about what the law requires, and the guides that produced this growth are aimed squarely at that intent. Holding the position will take a different argument than winning it did: the state guides have to be more useful than the statute, not merely present.
| Competitor, same thirty days | Movement |
|---|---|
| laborlawcenter.com | Visibility +29.54% |
| workwisecompliance.com | 51 new keywords in the top 3, 18 lost |
| dol.gov | Emerging competitor · visibility +35.19% |
What I would do differently
Set up position tracking in March, not August. The traffic curve on this page covers six months and the ranking table covers three weeks, because the tracker was only running for the final month of the engagement. I can show what changed and I can show which terms carry it, but I cannot join those two things as tightly as I would like, and the strongest single week of movement happened before anything was recording. The keyword set was the right one. It should have been running from the first week.
Google Search Console, Semrush Organic Research and Position Tracking, GA4, Microsoft Clarity, Screaming Frog, Google Sheets.
Losing ground and nobody can tell you why?
Send me the domain and I will pull the query data before we speak. Most of the time the impressions are already there and the pages are answering the wrong question.
Typical reply time: within one working day