Case Study 27.1 — "(not provided)": How Encrypted Search Made Search Console Indispensable

A real, documented episode from the public record that explains why the Performance report's query data is uniquely valuable — and uniquely irreplaceable. This case describes public facts about Google's move to encrypted search (2011–2013) and its effect on keyword data. Evidence note: the sequence of events is well-documented and dated; where this case gives proportions ("single digits at first," "effectively all organic keywords" by late 2013) it states the well-established direction and scale, and does not invent a precise percentage. The SEO reading of the consequences is labeled as analysis.

Background

For the first decade of SEO, marketers had a luxury they barely appreciated: when someone searched Google and clicked through to your site, the search term they typed was passed along in the referrer data and showed up in your web analytics. You could open Google Analytics and see, visit by visit, the exact keywords that brought people to each page. Keyword-level organic reporting was simply there, for free, in the tool everyone already used.

Then Google closed the window — for good reasons, and permanently. In 2010 Google introduced an encrypted (HTTPS) version of search. In October 2011, it began redirecting signed-in users to encrypted search, and a side effect of encryption is that the search query is no longer handed to the destination site. In analytics, those organic visits stopped showing a keyword and started showing a now-infamous placeholder: (not provided). Google initially framed this as affecting only a modest, single-digit share of searches.

It did not stay modest. Over the following two years the encrypted share climbed, and in September 2013 Google moved essentially all search to HTTPS — signed in or not. Within months, (not provided) went from a minor annoyance to effectively the entire organic-keyword picture in Google Analytics. The keyword-level organic report that SEOs had relied on for a decade was, for practical purposes, gone.

The SEO issue: the one window that stayed open

Here is the pivot that makes this a Search Console case. When (not provided) swallowed keyword data in Analytics, one source of query data remained: Google's own Webmaster Tools (renamed Google Search Console in May 2015). Its "Search Queries" report — the ancestor of today's Performance report — kept showing the actual queries a site appeared and was clicked for, because that data comes from Google's side of the transaction, not from the referrer that encryption stripped.

But — and this is the part that ties directly to §27.2 — Google did not simply reopen the old window. The query data in Search Console is aggregate and anonymized by design. You see the queries and their impressions, clicks, CTR, and average position; you do not see which individual visitor searched what, and rare or potentially identifying queries are omitted entirely (the data anonymization of §27.2, which is why the query rows never sum to the totals). Google's message, in effect, was consistent: you can have aggregate query intelligence to improve your site, but not the individual-searcher keyword data that raises privacy problems.

So the tool the whole industry had underused quietly became the only free source of the single most valuable input in SEO — what people actually type to find you — and it delivered that input in a privacy-preserving, aggregate form.

What it shows

• Search Console's query report is not a convenience; it is the last free window into organic search terms. Everything a keyword tool "estimates" about your own terms, Search Console reports — because it's Google's data, not a model of it. After 2013, there was no free alternative for your site's real query data.
• The data is aggregate and anonymized on purpose. The same event that made query data scarce also set its terms: privacy-first, aggregate, no individual searchers. Reading Search Console as an aggregate, privacy- respecting instrument — rather than expecting per-visit keywords — is reading it correctly.
• "We don't have the data" is often "we never opened the tool that has it." Countless site owners spent the post-2013 years lamenting lost keyword data while never verifying the free property that still reported it.

Outcome

(not provided) is permanent; encrypted search is simply how the web works now, and no analytics configuration recovers per-visit organic keywords. The durable outcome is a shift in where SEOs look: Search Console became the canonical source of organic query data, and "connect Search Console" became the first step of every serious SEO engagement. The paid keyword tools that dominate the industry exist, in large part, to supplement and contextualize this data (search volume, difficulty, competitors) — not to replace the query report, which they cannot, because they don't have Google's record of your site.

The lesson

The most valuable data about your search performance already exists, for free, in a tool from Google — but only if you verify your site and learn to read it. Encrypted search took away the casual, per-visit keyword data and left behind something more disciplined: an aggregate, privacy-respecting, authoritative query report that tells you what people really search to find you. Treat Search Console as the source of truth it became by historical accident, respect the anonymization that comes with it, and never again say "we don't know what we rank for" — because the answer has been sitting in the Performance report the whole time.

Discussion questions

1. Explain, to a colleague who misses the old Analytics keyword report, why Search Console's query data is better for SEO decisions in some ways and more limited in others. Name both.
2. The same privacy principle that produced (not provided) also produces Search Console's data anonymization (§27.2). Why is it consistent for Google to give aggregate query data but withhold individual-searcher keywords? How does that shape what you can honestly claim from the report?
3. A vendor's keyword tool and your Search Console query report disagree about which terms bring you traffic. Which do you trust for your own site's actual queries, and why? Where is the vendor tool still useful?
4. Rivertown's owners assumed their old vendor's slide deck told them what they ranked for. After reading this case, what is the first thing you'd have them open instead — and what would you expect it to reveal about the branded-vs-non-branded split (§27.2)?