Ranking For Writers

Simple, actionable advice for doing well in search engines without compromising the quality of your writing.

Version 1.2, updated Mar 15, 2021

Search Engine Visibility

This guide is for writers who want their content to rank well in search engines. Ranking well means being found towards the top of Google's results for relevant search terms.

Unlike many other sites about SEO (search engine optimization) that focus on technical details of linking or keyword usage, I want to give you actionable advice for how to think about search engines – and your content and audience – from a sustainable, long-term value perspective. I will not recommend tricks that may backfire in the future when Google changes its algorithms, nor will you learn magic ratios for how many times you should repeat a keyword in the first 200 words of an article. Such ratios do not exist, and any silver bullets will eventually run out.

Instead of silver bullets, we'll start by figuring out what the search engines (especially Google) are trying to do in the first place. We'll then consider some easily accessible and insightful tools. Finally, we'll look at a method of writing – including research, outlining, and publishing – that you can start following today.

Mikko Järvenpää

About the author

I'm Mikko. I've worked in the business of technology my entire career. My first job out of college was with Google, where I was a Product Marketing Manager (I quit to become a tour manager for a death metal band, but that's a story for later). More recently, I've been the CEO of Infogram, founded Sentient Media, and launched a couple of search-related products, recently Candle. You can find me at @mhj.

Contents of this site

How to think about search

The Google algorithm attempts to model the behavior of a human reader as accurately as possible. If you learn only one thing from this adventure, let it be this. The repercussions of this fact are that as the content creator, you should always write for your human readers, not for the search algorithms – even if you want to rank high in Google.

For the most part, Google is motivated by giving its users what they want. The search is good because it provides people with higher quality likelier than lower quality. There are some cases where Google's strategy attempts to balance the writer's interests against the economic interests of Google, and that may cause complications for the writer. We'll look at some of these briefly, but for our purposes, these are exceptions.

Google attempting to model human behavior means that it will try to understand everything it can that influences people's content preferences. Does the site load slow? You'll lose readers, and you'll lose in Google. Is your opening sentence lackluster, and people click back from your article? You'll also lose in Google. Did you start with a vacuous paragraph stating the obvious in multiple different phrasings? I hope you lose in Google (that's to all the badly-written, keyword-stuffed SEO articles out there).

If you want to do well in Google, write as if you would to your target reader who is intelligent, informed, pressed for time, easily distracted, and critical.

A process and a method

The research and outline method in this guide is designed to make it easy to write long articles that perform well in search engines. Below is an overview of what it looks like. We will jump around a little as we build familiarity with the process and the tools, but by the end of this guide and after applying the learnings a couple of times, you will already be well on your way to good SEO writing.

Mikko Järvenpää

Before diving into the process – and even if you don't follow the process or use the outline method – there are a few rules of thumb that I recommend you follow.

Rules of thumb

Write for people, not for algorithms.

Focus on the first 0.5 seconds

The most crucial sentence in a piece of writing is the first one. The second most important sentence is the second one. First impressions count, and they count real fast.

When a reader arrives at your article, essay, or blog post, they'll decide whether your content is worth their time. That decision is made based primarily on what is "above the fold" (an expression from the newsprint days), perhaps accompanied by a fast cursory scroll a little further down. You should assume this all will take less than one second.

By this time, the reader has gathered enough information to decide whether they'll stick with your content or bounce. There's a more technical definition for "bounce rate" in web analytics, but this characterization is complete for our purposes.

Suppose Google sees that a reader has clicked through to your site from search results or another Google-tracked source, and they leave the page right away. That's a negative signal about the quality of your content, at least for that context (if a user is looking for information on jaguars but ends up on your page about classic Jaguars, that's an example of an easy context for Google to figure out). Google has enough behavioral data to extrapolate quality assumptions from just a few signals, and combined with the semantic understanding of the indexing algorithms, it will quickly start to favor content that is engaging to users over content that is not – other things being equal.

If you're looking at your Google Analytics, don't worry too much about high bounce rates. An article may have a bounce rate in the high 80% or even 90% range but still do well in search results for the crucial searches that you target. The people who stick with you over the first impression hump; those are the people that you want to keep engaged.

Sometimes, you will have to experiment and test your assumptions. With Sentient Media, we had an internal debate about paragraph lengths and reader engagement, so we ran some tests. We picked news articles (1,000–1,500 words) that did reasonably well in Google's index and formatted them differently for two tests. For two weeks, those articles had paragraphs consisting of four to six sentences – typical paragraph lengths for a newspaper or a magazine article. After that, we split the articles into many much shorter paragraphs, from one to three sentences.

While keeping an eye on other metrics to ensure the audience and traffic sources remained unchanged across these two tests, we had our results: the same articles with the shorter paragraphs had approximately 100% improvement in engagement. In terms of bounce rates, for example, if we were previously getting 15% of the traffic from search engines to stick with the articles, we were now reaching 30% with the articles with shorter paragraphs.

This particular learning may not apply to you, but it illustrates the point: a simple change in our editorial style resulted in doubling our content impact with a particular article type.

Article length and time on page

There is no magic word count for articles that do well in Google's search results, but if there were, it would be higher than you'd like to hear. The longer the article, the better its chances of ranking well.

A long article contains more stuff. Even if an article is about any one thing, no matter how specific and niche, a longer article will contain more topics adjacent to the main topic; more ways of discussing it; more examples; more of anything relevant. This is great for SEO because it creates more semantic gravity for the article. Google very likely already knows the topic you're writing about. It also knows what other topics are related to it and how relevant each of those is. These are good signals.

A long article keeps the reader's interest for longer. This trivial statement makes sense in the context of our first rule of thumb – Google uses time spent on content as one signal of its quality. Don't consider this deterministic, however. There are many ways to keep people on a page, and Google will figure out what the ones that don't add value are.

Finally, a long article is perceived as higher value by readers than a short version. It is likelier to get shared, it is likelier to get comments and discussion (on your site or outside of it), and it is likelier to be linked to as a source of value. These dynamics feed each other.

But you still want to know how long is long?

Take a look at the first page of results for the search you are targeting. Open the results, and copy their body text into a writing app that gives you a word count (or use a word count browser extension). Figure out the average word count of the top five or top 10 articles on the results page (leave out clear outliers, anything ~80% off from the average of the remaining entries). Eyeball the quality of the competing articles. Consider if the site where the article lives is well-known or not among its intended target audience.

Take the average and multiply it by two (2x) if there are no large, well-known sites in the running. If all or most of the articles' quality is low compared to your output, just multiply by 1.5. If there are well-known sites in the competition, multiply by three (3x).

This means that if the average word count of first page articles is 1,000, and there is quality there, and there are large sites in the list, your word count target should be 3,000 and up. When in doubt, make it longer.

Daunting? I know. You can certainly write shorter pieces, but as discussed, this is likely to reduce your probability of making it towards the top of Google's results. Shorter pieces of course do make it to the first page of Google – but much less likely if they are competing against a long, well-written, valuable piece of content like the one you're about to create after reading this guide.

Think of your writing as you would of constructing an investment portfolio where most of your investment consists of the time you spend writing.

You can diversify like you were a seed-stage venture capitalist, placing small bets based on limited information all over the playing field (or mostly whatever is hot on Twitter and Clubhouse; same logic seems to apply to picking startups and picking what to write about in some circles!). If something catches on, double-down and follow up on that topic.

Or you can do the opposite and model your strategy after Warren Buffett's, placing highly researched and very large investments in carefully selected and often overlooked companies. This works best for a writer who knows what they are doing. They are an expert in what they write about, and they know the target audience very well.

Alternatively, and this where most people find themselves, you can do something in between while keeping in mind that you need to be in the game long enough to see some results. Calibrate your approach to the level of information you have, how much time you are willing to invest, and your level of expertise and recognition.

The process discussed in this guide will work for any of these approaches.

Datestamps and SEO

This is a more trivial nugget of information, but keep it fresh if you put a date on it. Of course, Google will figure out when an article has been published or updated without your help. That's not what the date stamp is for. Instead, it signals to your human readers that the page is fresh and up to date. On the internet, new beats old almost every time. Google knows this because it has learned that if there are old articles in the results, they get less engagement than more recent, equally informative articles.

If you look for product information or comparison articles, you often see such pages mention when they have last been updated. For highly competitive search terms, publishers like to update them every month. Because, if it were March 2021, you'd be likelier to click on the result "Best Cameraphones Compared (Updated March 2021)" than on "Best Cameraphones Compared."

Platform: speed and mobile-friendliness

In this guide, I will not go into meta tags, descriptions, or types of HTML elements, other than saying they matter less than the SEO profession would like you to think. There are two things that I implore you to pay attention to for SEO purposes.

The first is speed. The site must be fast. Pages must be fast to load, images must be as small in file size as possible, and you should use any reasonable technical solutions like caching and CDNs (content delivery network). Google cares tremendously about speed because your audience cares about speed. Speed is also a crucial part of the first impressions mentioned earlier. If the site loads slowly, the reader has already lost some of their patience with you.

The second is also speed, combined with legibility and mobile responsiveness. Google is mobile-first also in its algorithmic thinking, and the experience of the mobile Google user is priority number one.

If there were a third thing to focus on in a publishing platform, it would be speed.

The good news about site speed is that most publishing platforms – WordPress, Substack, Squarespace, Letterdrop, Ghost, Carrd (you're reading a Carrd site right now!) – are fast out of the box. If there are settings for using asset caching or a CDN like Cloudflare, you should probably use those. If you know enough to disagree, then you have a special case and enough expertise to do things your way, which is legit – you do you.

Prepare and research

Planning is indispensable even if things rarely go as planned.

Plan your keyword wishlist

This guide assumes that you are a writer: you create multiple pieces of content for a brand or a website, for example. The step we're in now pertains not to an individual article but a collection of writing like a thematic site or a personal brand. A collection of writing never covers just one narrow thing – if it did, you'd be just doing rewrites of the first article.

Instead, we're now interested in your overall strategy. What do you want to be known for? What are all the topics you'll be writing about, or may write about, or could write about if you had the time? This is the fun part of the process: make a wishlist of all searches that you want to rank well for in the medium-to-long term say 1–5 years from now. Aim for 20–50 entries on your keyword wishlist for starters.

You don't need to be too granular – don't include minor variations of searches, like "management accounting courses" and "management accounting MOOC." You'll get to drill down to that level later. For now, just list topics that you think people would want to search for and read about. I recommend you aim for two-word searches at a minimum, with a preference for three-word searches.

Find the semantic overlaps

You have a wishlist of searches. Now, let's group them into one or more groups of semantic relevance. You're going to dip into these buckets of semantically connected search terms in your research and writing, but they are also crucial for planning your writing outside of the one piece you're working on now.

Google is interested in expertise because Google users are interested in expertise. One way to signal expertise – and to actually build it – is to cover a topic comprehensively, in-depth, and in various contexts and contingencies. When you have a list of related search terms, you can use that list to develop ideas for more articles and other writing projects. By writing about related topics, you're educating Google about your expertise in all of these topics. This is what semantic relevance means in our context.

Once you start to have more than one written piece in a semantic bucket, you can also begin to cross-link them. For example, if you write about plant milk and already have an article on deforestation due to animal farming, link back to that article when you discuss plant versus dairy milk's environmental impact. There's no need to obsess over cross-linking, by the way. Just do it in a way that would be helpful for a human reader. Google will figure out your expertise even without you having to litter your articles with links back and forth between them.

We'll use our intuition as our primary tool at this stage. Some of the tools we'll look at next can inform this process, but if any changes are required, you can make them later.

Ask yourself these questions when considering semantic overlap. What terms are related to each other? What could naturally be discussed in a single post? Are there groups of search terms that necessarily be mentioned in a comprehensive article about any one of the terms?

Here's an example of bucket construction, a segment from a keyword wishlist:

how to learn accounting online
management accounting courses
corporate strategy jobs
strategy career education

There are two immediate semantic buckets here. The two first terms are more closely related with each other than with the latter two. Someone searching with one of those searches could quite possibly be interested in the other one, too. The latter two terms are related, but the search intent is different, though depending on the focus of the writer and the audience they cater to, the fourth term might be more closely related to the first bucket.

Outline method for faster writing

The outline writing method may not suit everyone's style, but I recommend you give it a try. You may like it even for other kinds of writing.

Mikko Järvenpää

You can access a Google Sheet of the outline template here. Just copy it into your own Google Drive, and keep reading to learn how to use it.

The Outline template

Primary keyword: this is what your article is about, and this is the search term you want your article to rank for.

Target word count: as discussed in rules of thumb.

Searches related: here, we'll list related searches that we will use as ideas for the article structure.

People also ask: similar to the above, this is where we'll collect questions related to the primary keyword, and we'll use these for our outline.

Google's first page results: this is optional but highly recommended. Here you'll collect the results on the first page of Google for your primary keyword. It is also useful for figuring out the target length of your article.

Tip: do not trust private browsing or the incognito mode on your favorite browser to give you objective results. Not only is there no such thing as purely objective results in Google, but there are many ways that Google will attempt to tailor the results to you, even if you use a private browser window. One way to go around this is to have a separate browser for just doing this Google research. Go ahead and download Firefox, Chrome, Brave, Opera, Edge, or any of the other less popular browsers, and use that. Don't log in to any accounts in that browser, don't install extensions or plugins, and don't visit Google results from that browser. Think of it as your "burner browser." Essentially, try to give Google as little in terms of behavioral data as you can, and the results you see will be as close to "default" as you can easily get.

SEO title ideas: brainstorm some article titles here. Attempt to include your primary keywords, as long as you can do so in a natural (copywriting) sentence. If your primary keyword is "animal agriculture and climate," your title could be "Animal Agriculture's Massive Impact on the Climate Crisis." Other best practices in title writing apply – make it informative, snappy, and highly clickable.

Article structure: this is the big one. We're getting close to starting on our article, but before that, let's give ourselves a solid structure. The article structure is a hierarchical outline, and you are likely familiar with its siblings, like the indented ordered list in a word processor or HTML.

The article structure maps to HTML headers. Your article's title is usually H1, the top-level header. This header should contain your primary keywords if they are possible to write into a natural sentence. The top-level header can be the same as your title tag, and it can be the same as the URL slug (for example, amazingwritercontent.com/your-article-title-here). However, they don't have to be the same as long as they don't deviate from each other too much. As long as they are semantically relevant and highly relevant to your article's content, you can experiment with some variance between the H1, title, and URL slug.

If your publishing platform does not make it clear what header-levels are at your disposal, do not fret. After all, is it important for the human reader whether a particular sub-heading is inside an H2 or an H3 tag in the HTML? It is not. Google will figure this out. But if you want to play it safe and have the ability to decide what header tag to use where, I recommend using H1 for the title of the article, H2 for the main topics in the article, and H3 tags for subtopics nesting inside the main topics. This concludes the HTML portion of our tutorial.

The structure can help you immensely in tackling the word count target you have. For example, if you aim for an article of 3,000 words and build an outline with 20 topics, you only need 150 words for each topic to hit your target. Breaking the big challenge into multiple smaller challenges makes it easier to make and track progress.

But right now, to not get ahead of ourselves, we'll now need to equip some tools.

Tools and opportunities

To keep things simple, we'll focus on a handful of free and easily available tools. SEO-specific tools like Ubersuggest have paid plans, but you can get pretty far with the free version.

The most important tools are not SEO tools at all, but features of Google's search results. That's what makes them even more fun to use.

Google: "People also ask..."

One of Google's first page features that often shows up is a box that suggests other related questions to the user's search. It is my favorite feature. Not only does it give us a glimpse into the live zeitgeist of the Google-using populace, but we can also use it to provide Google with more of what we want. This feature does not trigger for all queries, so poke around with differently worded searches around your topic to increase your probability of seeing it.

The "People also ask" (PAA) feature is a list of collapsed questions. Clicking on a question expands the answer Google has picked as the best for this question. But it also, at least as of writing this, usually loads two or more additional questions into the PAA feature.

Google does this dynamically: the additional questions loaded into the PAA are relevant to the one you clicked on. That means clicking on the question most pertinent to the topic you are writing about yields additional questions that are likely to be relevant.

Mikko Järvenpää

Using the PAA feature this way, you can nudge Google into giving you more questions that can help you build your outline. Feel free to expand into adjacent topics or sub-topics – that's just one way of building the semantic gravity around your main topic.

Google: "Searches related to..."

There is more than one kind of related searches (RS) feature in the Google results, and there may likely be experiments around these features. As the name suggests, RS gives you ideas for searches that you could find valuable. That information is collected from actual user behavior and the semantic structure of the web.

Using PAA and RS

You can use the material from these two features as inspiration or direct input into your outline structure. Record everything you find from these features on the outline sheet and attempt to address them in your article outline.

Sometimes that can mean taking a question from PAA directly into your outline as a topic or a subtopic. For example, if the search term you are targeting is "management accounting courses" and one of the PAA entries is "How do I become a management accountant?" and it is something you want to cover, you can take that directly as one of your main topic (H2) or subtopic (H3) level headers. But if the question format crimps your style or doesn't work in this case, feel free to write it out as a declaration instead.

You can also use what you find in these features to inform and grow the semantic buckets we mentioned earlier. If "How do I become a management accountant?" is something you think you could write an entirely different article on, by all means, save it into your keyword wishlist and the relevant semantic bucket. This way, the tactical research you do for individual articles also works to deepen your long-term strategy.

Search suggestions

These are the suggestions that dynamically populate in the search field when you start typing something in. They can be helpful and cover some additional topics, but I'm leaving this feature out because I haven't seen it vary in a useful way from the two above features. And you can't copy-paste the list of suggestions directly into your spreadsheet.

Google Webmaster Console

The GWC is a gem for understanding how your site and your articles perform in Google's index. It also flags any site performance or mobile usability issues that can hinder your performance in the index. Set GWC up for your site, refer back to it to track your progress, and see what keywords you are ranking for and producing traffic to your site. For writers, GWC is at least as necessary as its better-known sibling, Google Analytics. There are good video tutorials on getting started with GWC; just head over to YouTube.

GWC is the source you want to rely on when figuring out where you rank in your target keywords' search results. Often you'll see your content ranking even lower than you see in your burner browser. This could be because you are on your way up (yay!), and GWC doesn't reflect that yet. Or it could be because, despite your diligent use of a burner browser, Google has figured out that you are especially interested in your own content and seeks to boost that in the results it offers you.

Google Analytics

GA is the perennial favorite for analyzing your site's traffic, audience, and content performance. Going in-depth with GA is outside this essay's scope, especially as so much has been written about it already.

Ubersuggest (or similar)

Ubersuggest is a user-friendly SEO toolkit that lets you figure out how difficult it might be to compete for placement for your target keywords, what other sites are competing for the same (that you might not see in your own results!). It contains nice tools for finding keywords and topic ideas. As of writing this, there is a powerful free version available. Other good tools that give you similar assistance are Ahrefs and SEMrush, and there are many more even more weirdly-named tools out there. Use these tools to supplement the above tools and to check your assumptions.

Write and publish

Fill out the outline

Write! Having completed the above steps, we know what we're writing about, we know what related topics we'll touch on, and most excitingly – we have a ready outline of a substantial article, just waiting for us to fill it in.

Keep the outline handy when you write. You can refer back to it, check some of the materials you've saved there, and tweak the structure easier than in most text processors. As you write, you may come up with new ideas, and you can easily save these into the outline.

Publish, share, celebrate – repeat

A writer's job is ever only temporarily done. Your hard work will eventually be rewarded by a published article, a submitted manuscript, or a popular blog – but you'll keep writing, right?

Let me know if this guide has helped you in any way. You can find me on Twitter at @mhj.

Bonus content

Video interview

Brian Kateman from The Reducetarian Foundation interviewed me about how to apply these principles for advocacy-related writing. The interview was some time before I wrote this guide, so I talk of some of these things in different terms and from a different angle. For that reason, it may also offer something additional – high-level, not as tactical – as this guide.

Want to work together?

If you're interested in more detailed in-person coaching, have a content strategy you want to build, or want hands-on help creating the outlines or even writing your content, do get in touch.

© Mikko Jarvenpaa. All rights reserved.