Data-Driven SEO: Insights From the Experts

One example comes from the supplements space, where I used data to front-run a product trend before it hit the wider market.

I started with Exploding Topics. It’s a good tool to track fast-rising keywords.

One product concept was clearly breaking out: it had low absolute volume but a very steep growth curve and almost no established competition.

I validated this with Ahrefs by looking at the SERP history, competition, and traffic potential of the broader topic cluster.

The data showed two things:

1. We had a window of a few months before competitors woke up.

2. The opportunity was in the supporting content around the product, not just the product page itself.

Before the product launched, I built out a small content hub targeting all the informational and comparison queries I saw in Ahrefs.

Then I mapped an internal linking plan so that those pages would funnel authority to the future product page.

As soon as the product went live and search demand spiked, we pushed internal links from the hub to the product page and refreshed articles based on early Search Console data.

Because we were essentially the only site with a complete topical cluster, we took the #1 spot quickly and have held it for over a year.

That single cluster now drives consistent non-branded traffic of about 12k visitors per month (and rising).

It has attracted a few organic backlinks simply because we were ranking first when bloggers and journalists started researching the topic.

I run an SEO agency for private jet charter companies.

I analyse search data to identify what people actually book versus what they merely search for.

One example: Google Search Console showed 67 impressions for “empty leg flights London to Dubai” over 14 days. Zero clicks.

The client ranked in position 12.

The insight: People searching “empty leg [route]” want specific availability, not generic information. Our client had a catch-all “empty legs” page — wrong intent match.

What changed: I built a dedicated page for that route. I added weekly flight schedules, typical pricing (£28,000 vs £45,000 for a standard charter), the aircraft types flying the route, and departure patterns.

Result: The page reached position 3 within four weeks. It generated eight qualified enquiries in two months. Three converted to bookings.

The broader lesson: Search volume doesn’t matter if intent alignment is wrong. Forty impressions with high intent beat 4,000 impressions for research queries.

Also, sometimes you have to produce content to get data — start broad to identify the long-tail, then target high-intent queries.

When we combined Search Console, GA4, and ranking data from SEMrush, we found a pattern that changed the direction of the SEO work we were doing for a financial services client. At first, the numbers suggested that their visibility was stable, but SERP analysis of keywords related to their core services told a different story.

The data showed that almost all page-one rankings were tied to low-intent, informational keywords that generated impressions but no meaningful conversions. By contrast, every commercially valuable term, such as “financial advice for over-50s,” “retirement planning advice UK,” and “pension drawdown advisor,” was buried between positions 18-35. The affected pages weren’t underperforming because of strong competitors, but because the website had limited topical authority and barely covered relevant or related content outside of those pages.

The analytics from Search Console also uncovered page cannibalization across multiple pages, where near-identical content caused URLs to compete with each other for the same search queries.

Taking these insights into account, we rebuilt the SEO strategy around commercial intent rather than informational visibility. We created new pages that directly targeted the keywords stuck between positions 18-35, redesigned the internal linking structure to push authority towards bottom-funnel advice pages, and consolidated overlapping pages.

This shift improved the client’s visibility for the keywords that matter, led to a 42% increase in clicks to mid- and bottom-funnel pages, improved the average ranking for commercial keywords from 24.7 to 13.2, increased organic quiz entries by 31%, reduced keyword cannibalization by 68%, and improved on-page engagement on core advice pages by 27%.

For years, the SEO industry was obsessed with the idea that more is better. More blog posts, more pages, more keywords. But about two years ago, we hit a wall with a major B2B client. They had thousands of pages, but their traffic had plateaued, and leads were drying up.

We decided to stop creating new content and look strictly at the data we already had. We performed a Content Audit using Google Search Console and GA4, and that was where we found the gold.

We filtered for pages with high impressions, very low clicks, and high bounce rates. Ultimately, Google was giving people access to these pages, but users were ignoring them or clicking through and then taking a hard pass on what they found.

By hopping into both Reddit threads and Quora chats in our client’s niche, we noticed that user intent had changed. People no longer wanted those 2,000-word Ultimate Guides published in 2018. They craved bite-sized formulas, quick hacks, and clear answers. Our comprehensive content was so much fluff to them.

We used the Content Consolidation Strategy.

We identified 40 separate blog posts that all vaguely covered the same topic but were underperforming.

We deleted the weak ones and merged the useful bits into a single, high-value Power Page.

We updated the format based on that forum research: less text, more bullet points, and downloadable templates right at the top.

It was definitely scary to rip out 30+ pages from the site, but the data wasn’t lying. In three months, that single page had more visitors than the previous 40 pages combined. Even more significantly, the conversion rate on that page tripled.

A time when data analytics truly impacted my SEO strategy was when I dug deeply into last year’s organic traffic data. I uncovered a slow decline of around 12% over two months, with no significant algorithm change to account for it. I pulled performance data from our Search Console, focusing on queries, CTR (click-through rates), and page-level performance.

What caught my eye was the number of high-intent keywords that were still ranking but whose CTRs had deteriorated by around 18%. This told me that lack of visibility wasn’t the problem; rather, it was likely a lack of relevance or presentation. I cross-referenced competitors for the same queries and compared the pages of top performers with my own. I saw that my titles and meta descriptions weren’t closely aligned with user intent and lacked clarity and benefits.

I rewrote the titles and meta descriptions to be clearer, more benefit-driven, and aligned with the searcher’s phrasing. Within six weeks, CTR improved by 22%, and those pages collectively saw about 15% more organic traffic.

My biggest takeaway was that small shifts in user behavior show up early in the data. By focusing on these micro-metrics, I was able to adjust before it became a bigger ranking issue.

I used Search Console and Google Analytics together to decide which pages to prune and which to refresh on a content-heavy site.

I exported Search Console data for two time periods (months 1–8 and 9–16), then matched pages in Google Sheets using VLOOKUP to compare impressions, clicks, CTR, and average position over time. I flagged URLs that once had solid impressions but had almost dropped out of search in the more recent period.

Next, I pulled a landing page report from Google Analytics for the same period and joined it to the sheet, so every URL had SEO metrics plus sessions, engagement, and conversions.

From there, I split pages into:

• “Update” pages: historical impressions and conversions but declining visibility. These got intent-aligned rewrites, fresh content, and stronger internal links.

• “Prune/redirect” pages: no impressions, no traffic, and no conversions; these were merged or 301’d to stronger pages.

That cleanup led to fewer indexed URLs, a higher average CTR, and a lift in organic sessions without publishing lots of new content.

One instance in which I employed data analytics to inform SEO strategy occurred during the assessment of a large e-commerce client, where we noticed that organic traffic had plateaued.

Through the use of Google Analytics, Search Console, and Looker Studio, I discovered that one of the category pages was experiencing:

• A 35% decrease in click-throughs over the past 60 days.

• A bounce rate that increased from 48% to 64% in the same timeframe.

Search Console indicated that the page was still within the top results; however, the click-through rate had dropped from 6.4% to 3.1%.

Additionally, keyword queries had shifted to less desirable, lower-value variations, such as “budget” and “affordable.”

Using Hotjar, I saw that users were only scrolling through 40% of the page and were ignoring the key filters.

The search intent had shifted toward affordability and comparison.

There was a content mismatch between the position of the page copy and the premium and luxury content.

The page had a UX issue: the filters were placed below the fold, thereby reducing user engagement.

Assisting users through a value-driven experience required content updates to ensure alignment with user intent.

The title of the page had previously read: “Best Luxury Furniture Sets.”

This was updated to: “Best Affordable Furniture Sets – Compare Prices & Top Offers.”

We included additional budget picks, a price comparison table, and targeted long-tail queries in the FAQ section.

UX was further enhanced by moving the filters from below the fold to above the fold and by creating a sticky filter bar that incorporated structured data to surface discounts.

Revised meta tags:

• Old: “Premium Furniture Sets Online”

• New: “Affordable Furniture Sets – Compare Prices & Best Deals 2025”

Results (30 – 45 Days):

• Clicks on the category page: +42%.

• Impressions: +27%.

• CTR improved from 3.1% to 7.2%.

• Conversions increased by 18% due to better UX.

• Bounce rate decreased to 41%, and time on page was up by 31%.

Analytical data pinpointed shifts in intent, content omissions, and UX challenges. Redirecting the page to better match users’ behavior led to positive, significant improvements in CTR, engagement, and conversions. This reaffirmed that the foundation of optimization for search engines is understanding intent — not just keywords.

One of our most effective data-driven decisions recently was to redefine our entire SEO strategy by combining BOFU keyword analytics with LLM visibility signals.

Instead of relying on generic keyword research, we identified every bottom-of-the-funnel query with real revenue potential and evaluated each across four dimensions:

• conversions and purchases it generated

• traffic volume and CTR

• performance of the pages ranking for it

• our current positions in search results

Then we added a second layer: AI visibility.

We generated target prompts derived from these BOFU keywords and used our internal AI tooling to collect what users actually ask on Stack Overflow, Reddit, and other developer communities. We also mapped which of our pages appear in LLM outputs and under which potential prompts. Of course, we can’t know the exact prompts users type, but we can reasonably infer them. Whenever our internal monitoring showed even minimal LLM visibility, we added those prompts to our target list as well.

Based on this combined dataset, we built our annual plan:

• a defined set of target keywords

• a corresponding set of target prompts

• the pages associated with each

• their current rankings and LLM visibility

• the visibility targets we want to reach next year — both in traditional search results and in LLM answers

This approach gave us a roadmap grounded entirely in real user intent and actual performance signals. It helped us focus our SEO and content investments only on keywords, prompts, and pages that influence revenue and where we have a realistic chance to win — across both search engines and AI-generated results.

Recently, one of our clients experienced a dramatic 75% drop in traffic and impressions shown in Google Search Console during the third week of August 2025. We conducted a site and code audit but found no malware or code-related issues. We then conducted a backlink audit and found over 100 toxic backlinks from dozens of porn sites, likely a negative SEO attack by a competitor. All the links were created in July and appeared in Search Console in August. We updated the Disavow file with all the new toxic backlinks on Monday, August 18, and within eight days the client’s site impressions had returned to pre-August levels. Google’s AI says disavowing toxic links is no longer required, but that is 100% incorrect!

We create content at scale for our 1,400 WordPress customers. By building our own backend, where prompts, content, and Google Search Console data come together, we always know which content and prompts perform best across industries. Using this data strategically, we found a way to build topical authority again and again, quickly, by connecting to customers’ existing rankings during onboarding and extending horizontally and vertically into clusters and subclusters. Data drives every decision, and AI is the engine. This strategy and framework have guaranteed our customers an average 60% increase in organic traffic in their first 12 months with us (at the portfolio level). Despite AI Overviews and AI Mode, we continue to increase organic traffic instead of losing it, and this is not about reacting to change but about evidence from our own and customers’ first-party data, showing which prompts, content, and frameworks work best, building resilience through all kinds of search (algorithm) updates and user behavior changes.

We ran a series of AI-powered SEO experiments on our blog to understand what truly drives visibility and clicks. Instead of constantly creating new content, we focused on optimizing and reusing what we already had.

We tested four main tactics:

• Rewriting titles with AI

• Shortening and improving content with AI

• Updating publication dates

• Merging similar posts

What worked best: Shortening content with AI. Concise, well-structured articles with clear key takeaways increased clicks by over 340% and impressions by 70%.

What worked partly: AI-generated titles improved visibility but not always CTR. Context matters more than clickbait.

What did not work as expected: Merging posts increased impressions but reduced clicks, showing that strong headlines and structure are still key.

Overall, we achieved a 20% increase in clicks and 50% more impressions within four months.

The key takeaway: AI can enhance SEO performance when used thoughtfully — to clarify and strengthen content, not just to automate it.