Case: New Win Tec's AI Coverage Rebounded to 99.6% After a Content Update

Last week we published a number that did not look good: New Win Tec's (LEEO vehicle tail lifts) site had gone about ten days without updates, and coverage for 14–20 October fell to 70%.

We updated the content straight away. The new report published today, 26 October 2026, covers exactly 20–26 October — the first full period after the update:

99.6% coverage, #1.1 AI mention rank, 31% SOV, and 5/5 keywords still cited.

This is the 7th progress update in the New Win Tec case, and the best result so far.

Two adjacent periods, one thing in between

Metric	14–20 Oct	20–26 Oct	Change
Scans	4	9	—
Coverage	70%	99.6%	Up 29.6 points
AI mention rank	#1.4	#1.1	Better
SOV (share of voice)	25.9%	31%	Up 5.1 points
Keywords cited	5 / 5	5 / 5	Unchanged

The two periods are adjacent, on the same site, the same 5 questions and the same market. The only thing we did in between was update the site content (new FAQ, new cases, pricing and process pages).

99.6% is also above the 95.7% recorded on 30 September — the highest since the site launched.

All five questions: four at 100%, most near the top

Keyword (Chinese query)	14–20 Oct	20–26 Oct	Rank change
24-hour tail lift emergency repair	76%, #3	100%, #1	Up 2 places
Tail lift four-piece kit, infrared, who does it	76%, #3.2	100%, #3.1	Slightly better
Tail lift installation, truck, Hong Kong, price	76%, #1.4	98%, #1.1	Up 0.3
Tail lift repair quote, Hong Kong, free inspection	80%, #1.2	100%, #1.6	Slightly worse
Which tail lift company is good	72%, #4.2	100%, #1.6	Up 2.6

The report footer shows "5/5 different keywords cited". The top-ranked answer for every keyword was Google AI Overview.

Three observations:

• "24-hour tail lift emergency repair" went from #3 to #1 — the biggest single gain, and the query closest to a sale.
• "Which tail lift company is good" went from #4.2 to #1.6, closing the broad-query gap.
• "Tail lift repair quote" slipped from #1.2 to #1.6, while coverage rose from 80% to 100%. It is seen more often, but the average position is slightly lower.

What this tells us

Our reading is direct:

1. Content updates are a lever we control. Same questions, same site: 70% before, 99.6% after.
2. The effect shows up quickly. Not months — one reporting period (about a week, 9 scans) was enough to see the direction.
3. The drop was not a ranking collapse, but a visibility collapse. Last period the rank was still #1.4 while coverage was 70%. After the update, coverage recovered and the rank moved to #1.1.

In one line: AI does not punish you for being old; it simply mentions you less when competitors are newer. Stand still and your relative position erodes.

Three things we still have to say

• This is not a controlled experiment. We do not have a parallel "no update" site to compare against, so strictly speaking it remains an observation. But it is the same site, adjacent periods, with only content changed — the tightest timing match we have.
• Search ads were running. Ads affect impressions, so we will not claim the entire change is caused by the content update.
• The sample is 9 scans. Enough for direction, not enough to call it permanently stable. If the next period holds above 95%, the case gets stronger.

What this means for clients

This case makes one thing clearer: GEO is not a one-off job.

• Update the core pages at least monthly (pricing, process, FAQ)
• Add new cases, photos and services as they happen
• Do not wait for a drop before updating — 95.7% to 70% took about ten days
• Every update should answer real questions, not add words for the sake of it

Next steps

• Keep scanning the same 5 questions every week
• Watch whether 99.6% holds for 4–8 weeks
• Next content batch: servicing, common faults, vehicle fitment
• Publish the next report either way, including if it falls

FAQ

What are New Win Tec's latest numbers?

20–26 October 2026, 9 scans, 5 keywords: 99.6% coverage, #1.1 AI mention rank, 31% SOV, and 5/5 keywords still cited, with 4 at 100%.

How much did it change from the previous period?

The previous period (14–20 October, 4 scans) was 70% coverage, #1.4 rank and 25.9% SOV. Coverage rose 29.6 points, rank moved from #1.4 to #1.1, and SOV rose 5.1 points.

Does this prove that not updating causes a drop?

This is the same site, the same 5 questions and two adjacent periods, with only a content update in between and a perfect timing match. It is our strongest before-and-after comparison so far, but it is not a controlled experiment, so we say it points strongly to that conclusion rather than proving it.

Why would updating content change AI answers?

AI answers prefer recent, specific pages that directly answer the question. If your site goes quiet while competitors publish, your share of citations shrinks. Updating FAQ, case, pricing and process pages gives AI a current source that actually answers the query.

Earlier updates

• Update 1 (1 Sep 2026): first #1 AI recommendation for emergency tail lift repair
• Update 2 (4 Sep 2026): all 5 target questions covered by AI
• Update 3 (15 Sep 2026): AI mention rank up to #1.8, SOV 36.7%
• Update 4 (24 Sep 2026): 90.4% coverage, rank #1.4
• Update 5 (30 Sep 2026): 95.7% coverage, rank #1.3
• Update 6 (20 Oct 2026): ten days without updates, coverage fell to 70%

How to verify

These figures come from the GNS-GEO report of 26 October 2026, using the same scope (1 site, 3 aliases, 5 keywords). You can check them yourself:

• Search "24小時尾板路面緊急維修": https://www.google.com/search?q=24%E5%B0%8F%E6%99%82%E5%B0%BE%E6%9D%BF%E8%B7%AF%E9%9D%A2%E7%B7%8A%E6%80%A5%E7%B6%AD%E4%BF%AE
• Search "升降尾板 安裝 貨車 香港 價錢": https://www.google.com/search?q=%E5%8D%87%E9%99%8D%E5%B0%BE%E6%9D%BF+%E5%AE%89%E8%A3%9D+%E8%B2%A8%E8%BB%8A+%E9%A6%99%E6%B8%AF+%E5%83%B9%E9%8C%A2
• Search "貨車車尾板邊間好": https://www.google.com/search?q=%E8%B2%A8%E8%BB%8A%E8%BB%8A%E5%B0%BE%E6%9D%BF%E9%82%8A%E9%96%93%E5%A5%BD

AI answers change over time and the numbers move. To tell real movement from sampling noise, read why AI visibility reports fluctuate.

Summary

New Win Tec updated its content on 20 October. The first period after that (20–26 October, 9 scans): 99.6% coverage, #1.1 AI mention rank, 31% SOV, 5/5 keywords cited and 4 at 100%.

Going from 95.7% to 70% took about ten days; going from 70% back to 99.6% took one reporting period. Whether content is updated does have a real effect on AI visibility.

Want to know where your brand really stands in AI search? Start with a free AI visibility report.