Not Provided Keywords: What You Can Recover

Not provided keywords are the organic search terms Google has hidden from analytics tools since 2013, when it moved to secure search and stopped passing keyword data through referrer strings. You cannot see them directly in Google Analytics. What you can do is piece together a reliable picture of which keywords are driving your organic traffic by combining Google Search Console data, on-page content mapping, and a small amount of structured analysis. It takes about an hour to set up properly and gives you far more signal than most marketers realise is available.

I want to be honest about what this article is and is not. It is not going to promise you a workaround that magically restores the keyword data Google removed. That data is gone from your analytics. What this is, is a methodical walkthrough of how to recover enough keyword intelligence to make good decisions, using tools and methods that are available to any marketing team right now.

Why Google Removed Keyword Data in the First Place

In 2011, Google began encrypting search queries for signed-in users, citing privacy. By 2013, it had extended that encryption to all searches, which meant the keyword data that had previously flowed through referrer URLs into analytics platforms was now replaced with the phrase “not provided.” The official reason was user privacy. The practical effect was that marketers lost direct visibility into the organic search terms driving traffic to their sites.

There is a reasonable argument that this decision also had commercial logic behind it. Paid search keywords remained visible in Google Ads. Organic keyword data disappeared. The gap between what you could see in paid and what you could see in organic created an incentive to spend more on paid. I am not going to overstate that point, but it is worth naming.

The result for marketers was a significant reduction in visibility into what was actually driving organic performance. For anyone who had built their SEO reporting around keyword-level traffic data in analytics, it was a genuine operational problem. Most teams adapted, but many adapted by simply ignoring keyword-level analysis rather than finding a workable substitute. That is the gap this article addresses.

If you are thinking about this in the context of a broader organic growth strategy, the articles in the Go-To-Market and Growth Strategy hub cover the commercial decisions that sit above this kind of tactical channel work, which is useful context for prioritising where keyword intelligence actually matters.

What Google Search Console Actually Gives You

Google Search Console is the most direct and accurate source of organic keyword data available to you. It is not a perfect substitute for the old referrer-based keyword data, but it is far more useful than most teams treat it.

Inside the Performance report in GSC, you can see:

• The queries that triggered impressions for your site in Google Search
• Click-through rate by query
• Average position by query
• The pages that received clicks from each query

That is a substantial amount of keyword intelligence. The limitation is that GSC aggregates and samples data, applies some privacy thresholds that suppress very low-volume queries, and does not connect keyword data to on-site behaviour like conversions or time on page. But for understanding which search terms are driving traffic to which pages, it is the most reliable data source you have.

To get the most from it, filter by page and export the query data. This lets you see, for each landing page, which queries are sending traffic. Do this for your top 20 organic landing pages and you will recover a substantial proportion of the keyword intent driving your organic channel. The long tail will still be missing, but the queries that matter commercially will be visible.

One thing I have found useful when working with SEO data across different clients is to look at the ratio of impressions to clicks by query. A high-impression, low-click query tells you that you are appearing in search but not compelling enough to earn the click, which is a title tag and meta description problem, not a content problem. A low-impression, high-click-rate query tells you that when you do appear, the intent match is strong, which means there is likely more traffic available if you can improve your ranking. Those two patterns require completely different responses, and you can only see them if you are working with GSC data at the query level.

How to Connect Search Console to Google Analytics 4

Linking Google Search Console to Google Analytics 4 is the most impactful single step most teams can take to recover keyword-level context. Once connected, GA4 surfaces a Search Console report that shows organic search queries alongside landing page data, giving you a bridge between keyword-level search data and on-site behaviour.

To connect them, go to Admin in GA4, find the Search Console Links option under Property, and follow the prompts to link your verified GSC property. Once linked, the Search Console reports appear in GA4 under Reports, Acquisition, Search Console. You will see organic Google search queries, landing pages, clicks, impressions, CTR, and average position, all in one place.

What this does not give you is a direct connection between specific keywords and conversion events. That gap remains. But it does let you see which queries are driving traffic to which pages, and you can cross-reference that with your GA4 conversion data by landing page to infer which keyword clusters are commercially valuable. It is inference, not direct attribution, but it is honest inference based on real data rather than guesswork.

I have spent a lot of time over the years working with teams who treat analytics as a source of truth rather than a perspective on reality. The not provided problem is actually a useful corrective to that instinct. It forces you to think about what you are trying to understand and whether the data you have is sufficient to make a decision, rather than just pulling a report and assuming the numbers tell the whole story.

Using Third-Party Tools to Fill the Gaps

Google Search Console covers your own site’s search performance. Third-party tools like Semrush and Ahrefs extend that picture by providing estimated keyword data based on their own crawls, clickstream data, and modelling. It is worth being clear about what that means: these are estimates, not direct measurements. But they are useful estimates, particularly for competitive analysis and keyword discovery.

Semrush’s Organic Research tool, for example, shows you the keywords a domain is estimated to rank for, along with estimated traffic volumes and position data. Semrush’s work on market penetration illustrates how this kind of keyword-level data connects to broader commercial opportunity mapping. You can use the same tool to look at your own domain and see which keywords you are estimated to rank for, which gives you a broader picture than GSC alone, particularly for keywords where GSC suppresses data due to low volume.

The practical workflow I recommend is to use GSC as your primary data source for understanding what is actually happening with your organic traffic, and third-party tools as a supplement for keyword discovery, competitive gap analysis, and validating whether your GSC data looks directionally correct. When the two sources broadly agree on your top-performing keywords, you can have reasonable confidence in the picture. When they diverge significantly, that is worth investigating rather than assuming either source is right.

Tools like Semrush’s growth tools also provide keyword clustering and topic modelling features that help you group related queries into content themes, which is useful for planning content that covers a topic comprehensively rather than targeting individual keywords in isolation. That shift in approach, from keyword targeting to topic authority, is partly a response to not provided and partly just better SEO practice.

Content-to-Query Mapping: The Most Practical Recovery Method

The single most practical method for recovering keyword intelligence at scale is what I call content-to-query mapping. The principle is straightforward: your top organic landing pages are already attracting search traffic, and GSC tells you which queries are sending that traffic. By systematically mapping your content to its associated queries, you build a working picture of keyword intent across your site without needing to reconstruct individual keyword-level analytics data.

Here is how to do it in practice. Export your top 50 organic landing pages from GA4, ordered by organic sessions. For each page, go into GSC, filter by that URL, and export the query data. You now have a spreadsheet that shows, for each major piece of content, which search queries are driving traffic to it. Group those queries by intent (informational, navigational, commercial, transactional) and you have a keyword intent map of your organic channel.

This process typically takes a few hours the first time and much less on subsequent runs. What it gives you is actionable: you can see which pages are attracting informational traffic that is not converting, which pages are capturing commercial intent, and where there are gaps between the queries you are appearing for and the content you have on the page. That last point is particularly valuable. If a page is appearing for queries that the content does not directly address, you either have a content gap to fill or a relevance problem to fix.

Early in my career I spent a lot of time optimising for metrics that felt precise but were not actually connected to commercial outcomes. Keyword-level data in the old analytics world had some of that quality: it felt like signal, but a lot of it was noise. Content-to-query mapping forces a more honest question, which is not “what keywords are driving traffic” but “what intent is this page serving, and is it the intent we want to serve.” That is a better question.

Landing Page Analysis as a Keyword Proxy

If you cannot see which keywords drove a session, the next best thing is to understand what the landing page tells you about likely intent. A session that starts on a product comparison page almost certainly came from a commercial or transactional query. A session that starts on a how-to article almost certainly came from an informational query. The page itself is a proxy for keyword intent.

This is not a perfect substitute for keyword data, but it is a useful approximation that requires no additional tools. Segment your organic traffic by landing page category, map those categories to intent types, and you have a working model of how your organic channel is performing across the funnel. Combine that with conversion data by landing page and you can start to see which intent categories are commercially productive and which are not.

I have seen this approach used effectively in situations where teams have limited tool budgets or are working in markets where third-party keyword data is sparse. It is particularly useful in B2B contexts, where keyword volumes are low and third-party tools often have unreliable data. In those situations, landing page analysis combined with GSC data is often more reliable than anything a third-party tool can give you.

The broader point is that not provided is a constraint, not a dead end. The teams that handle it well are the ones that treat it as a prompt to think more carefully about intent rather than a reason to abandon keyword analysis altogether. Understanding why a visitor came to a page is still possible. It just requires a slightly different method than pulling a keyword report.

What Paid Search Data Can Tell You About Organic Keywords

If you are running paid search alongside organic, your Google Ads search term reports are a genuinely useful source of keyword intelligence for organic strategy. The queries that convert in paid search are, by definition, commercially valuable. If those same queries are also driving organic traffic, you are getting signal from two sources. If they are not driving organic traffic, you have a gap to address.

The practical application is to take your top converting paid search queries and check whether you have organic content that directly addresses those queries. If you do, check your GSC data to see whether you are appearing for those queries organically. If you are not, that is a content or authority gap. If you are appearing but not ranking well, that is an optimisation opportunity.

This cross-channel approach to keyword intelligence is one of the more underused methods in most marketing teams. Paid and organic search are often managed by different people or different agencies, and the keyword data that exists in paid rarely gets used to inform organic strategy. That is a waste. The intent data in your paid search account is real, validated by actual conversion behaviour, and directly applicable to organic content planning.

I spent a significant part of my agency career managing large paid search accounts and watching the same conversation happen repeatedly: the paid team would have detailed keyword-level conversion data and the SEO team would be working largely in the dark on intent. Getting those two functions to share data was often harder than it should have been, partly because of organisational structure and partly because people protect their data as if it belongs to them rather than the client. Breaking down that silo is one of the simpler things you can do to improve the quality of your organic keyword intelligence.

How to Build a Keyword Recovery Dashboard

Pulling this together into a repeatable process requires a simple dashboard that aggregates the data sources described above. You do not need a sophisticated BI tool for this. A well-structured spreadsheet or a basic Looker Studio report is sufficient for most teams.

The core components of a useful keyword recovery dashboard are:

• GSC query data by landing page, updated weekly or monthly
• Organic traffic by landing page from GA4, with conversion data where available
• Paid search top converting queries from Google Ads, updated monthly
• Third-party estimated rankings for your target keyword set, updated monthly
• A content-to-intent mapping column that categorises each landing page by funnel stage

With those five data sources in one place, you have a working picture of your organic keyword performance that is substantially more useful than “not provided” in your analytics. It requires maintenance, but not much. An hour a month to update the data and review the trends is enough for most sites.

The discipline of building this kind of dashboard also forces a useful question: what are you actually trying to understand about your organic keyword performance, and why? I have seen teams build elaborate reporting infrastructure that nobody uses because it was built to impress rather than to inform decisions. A keyword recovery dashboard should answer specific questions: which keyword clusters are driving commercial traffic, where are the gaps between our content and the queries we want to rank for, and are we improving over time. If the dashboard does not answer those questions, it is not doing its job.

There is a broader point about measurement here that connects to how growth strategy works in practice. Go-to-market execution is getting harder across most categories, and the teams that cope best are the ones that maintain clear visibility into what is actually working, even when the data is imperfect. Not provided is a good test of whether your team can work with honest approximation rather than demanding false precision.

The Limits of Keyword Recovery and What to Do Instead

It is worth being direct about what you cannot recover. The long tail of low-volume queries that drove occasional traffic to your site is largely gone. GSC suppresses queries below a certain threshold, and third-party tools do not have reliable data for very low volume searches. If you are trying to reconstruct a complete picture of every keyword that ever drove a session, you cannot do it.

But that is the wrong goal. The queries that matter commercially are not the long tail of one-off searches. They are the keyword clusters that drive consistent, intent-matched traffic to your most important pages. Those are recoverable, and the methods described above will surface them reliably.

The more important shift is from keyword-level thinking to topic-level thinking. Rather than trying to track individual keywords, focus on whether your content comprehensively covers the topics your target audience searches for. That means understanding the range of queries associated with a topic, not just a single target keyword per page. GSC data, content-to-query mapping, and third-party keyword tools all support that kind of analysis, and it is a more durable approach than keyword-by-keyword tracking.

There is also a useful parallel here with how growth strategy works more broadly. BCG’s work on commercial transformation makes the point that sustainable growth requires understanding your market at a structural level, not just optimising what you can currently measure. The same logic applies to organic search: the teams that grow their organic channel over time are the ones that understand their audience’s search behaviour at a topic and intent level, not the ones that chase individual keyword rankings.

If this kind of channel-level analysis connects to larger questions about how your go-to-market strategy is structured, the Go-To-Market and Growth Strategy hub covers the commercial framework that sits above individual channel decisions, including how to think about organic search as part of a broader growth model rather than an isolated tactic.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions