What Keywords Meant to Google in 2017

Yes, Google used keywords for SEO in 2017, and it still does today, but the mechanism was already very different from the keyword stuffing era that many people had in mind. By 2017 Google had spent years moving from literal string matching towards understanding entities, synonyms and intent. Hummingbird had reshaped query interpretation in 2013, RankBrain had been folded into the core ranking system by 2015 and was described as one of the most important signals, and mobile-first thinking was reshaping how pages were evaluated.

In practical terms, 2017 was the year many marketers finally accepted that a page did not need to repeat an exact phrase a set number of times to rank for it. Instead, a page needed to demonstrably cover the topic behind the phrase. Google was matching meaning, expanding queries to related concepts, and rewarding pages that satisfied the underlying need rather than pages that mirrored the wording.

How AAMAX.CO Can Help With Your SEO

We are AAMAX.CO, a full service digital marketing company offering Web Development, Digital Marketing and SEO Services worldwide, and we still meet businesses running keyword strategies built on 2017 era assumptions or older. Our approach replaces density chasing with intent mapping: we research the questions your audience actually asks, group them into topic clusters, build content that answers each one completely, and support it with clean technical foundations and internal linking. If your rankings have plateaued because your keyword approach never modernised, hire AAMAX.CO for professional search engine optimization and we will rebuild your strategy around how search engines interpret language now.

The Tactics That Were Already Dead by 2017

Meta keywords had been ignored by Google for over a decade and provided zero ranking benefit. Exact match keyword density targets, the idea that a phrase should appear in a fixed percentage of your words, had no basis in how the algorithm worked and often produced awkward, low-quality copy. Exact match domains stuffed with a service and a city name had lost most of their advantage. Article spinning, footer keyword lists, hidden text and link scheme anchors were actively risky after successive Penguin and Panda updates, with Penguin having become part of the core algorithm in 2016 and operating in real time.

Anyone still relying on those tactics in 2017 was already fighting the wrong war. The sites that thrived were those investing in genuinely useful content, technical health and credible link acquisition.

What Actually Worked in 2017

Search intent came first. Understanding whether a query was informational, navigational, commercial or transactional determined the format of the winning page. A comparison query needed a comparison table, not a sales page. A how-to query needed clear steps, not a brochure.

Topical depth came second. Pages that covered the subject from multiple angles, answered follow-up questions and used the natural vocabulary of the subject outperformed thin pages built around a single phrase. This is where the concept of semantically related terms proved its worth, not as a checklist to sprinkle in, but as a natural consequence of writing knowledgeably.

Placement discipline came third. Including the primary phrase in the title tag, the H1, the opening paragraph, the URL slug and a subheading or two was, and remains, sensible practice. It tells both users and crawlers what the page is about without distorting the writing.

Mobile experience came fourth. Google had begun rolling out mobile-first indexing experiments, and the intrusive interstitial penalty landed in January 2017. Sites with slow, cramped or pop-up-laden mobile pages lost ground regardless of keyword work.

Finally, structured data and rich results were becoming a genuine differentiator, expanding how a page could appear on the results page beyond a plain blue link.

Carrying the 2017 Lessons Forward

The direction of travel established in 2017 has only accelerated. Subsequent years brought BERT for natural language understanding, helpful content assessments, and now AI-generated answer summaries. Every one of those developments pushed further away from string matching and further towards evaluating whether a page genuinely helps a person.

So the modern version of keyword research is not a hunt for phrases with the highest volume and lowest difficulty. It is a process of understanding a market's questions and building an information architecture that answers them. Choose one primary topic per page. Identify the cluster of related questions that belong to it. Write for a reader who wants a real answer. Use the phrase naturally where it fits. Interlink related pages so both users and crawlers understand the relationship between them.

Track performance by query group rather than by single keyword. Search Console query data shows you dozens of variants that a single well-written page ranks for, and that spread is a far better indicator of topical strength than one tracked position. Because discovery now also happens inside AI assistants, structuring content with clear headings and direct answers has become part of the same discipline, which is why we combine traditional optimization with GEO services for clients who want to be cited as well as ranked.

A Simple Modern Keyword Workflow

Begin with customer language: support tickets, sales calls, reviews and forum threads. Expand with search suggestion data and question-based queries. Cluster everything by intent, then decide which cluster deserves its own page and which belongs as a section of an existing one. Draft the page to answer the primary question in the first hundred words, then go deeper. Optimize the title and description for clicks, not just for inclusion. Publish, monitor impressions and average position, and refine based on the queries you begin to appear for rather than the ones you guessed at.

Support the whole programme with technical hygiene: fast pages, crawlable architecture, correct canonicals, working internal links, and a sitemap that reflects reality. Keywords cannot rescue a site that search engines struggle to crawl or that visitors abandon before it renders.

Conclusion

Google did use keywords in 2017, and it uses them now, but always as an entry point into understanding meaning rather than as a matching game. The tactics that felt clever in that era, density formulas and meta keyword tags, were already irrelevant. What worked then and works now is the same: identify real intent, cover the topic thoroughly, present it in a fast and accessible page, and earn credibility. If you want that discipline applied to your site by a team that has watched these shifts happen year by year, we would be glad to help.