This subject line/title made me wince a little so then I tried wrapping the term “SEO” in quotes to distance myself from it, like that would reduce the cognitive dissonance of writing this because optimizing a specific piece of content for search is my least favorite part.
How teams typically approach optimizing content
It goes something like this.
In the first part, you choose a topic doing keyword research where everyone is focused on getting traffic as a numbers game. You have already decided success will get measured as how much you can dilute the quality and relevance of users you’re sending to a piece of content.
In the second part, you write the article from a keyword rich outline. You rely on a rote task-based approach to optimizing marketing content for robots. Keyword in heading? Check. Title tag rewritten? Check. Commonplace because it’s easy to grasp adding keywords you want to rank for to your content and simple checklists can fit easily into even the most frazzled content team’s workflow.
In the third part, you amplify, sharing on social. In the best case scenario you hit publish, and the search engine gods shine traffic upon you, no additional work required.
Your new user starts reading. They see something akin to, “Dear Robot, here is a tasty keyword for you… Dear User, fill out my popup CTA.” Sensing the content is written for robots, they quickly scans for anything useful to their cause and then races to the back button, safer now, because it is easier to get a relevant answer in Google.
And so this is the status quo we are up against.
Instead of all that, you can gain a meaningful advantage by writing for your audience, creating website resources that will make competitors talk about you in meetings, and win by moving the needle of users supported, which includes improved conversions. And all you have to do is approach SEO’ing content as contextualizing content as solutions to problems, as an opportunity to account for their needs in search in the way you serve that audience.
Some tips to that end:
When doing your keyword research
Do keyword research with people
You can’t assume you know the intent behind keywords when doing keyword research. These are all recent things said to me that I bring different meaning to than they did:
- “How many blogs should we have?” referring to articles, not blogrolls.
- “We should talk to the web firm” referring to a web dev agency.
- “We are making mood boards” referring to graphic mockups of the above fold area of a homepage.
- “We need another media buy” referring to increasing social ad spend.
- “Will your work affect the sitemap?” referring to the header navigation.
The way people in your audience/market search and talk will be different from you based on their industry, geography, and level of depth of understanding about what you do. You are playing the long game so it is worth investing energy here. Pay attention to how they talk about things and account for that in your writing.
When there is a mix of ways people talk about things, it is okay to provide alternate titles, AKAs, and subject + verb for x / y / z. That helps users and search engines with information scent.
Spend some time on it. Explore synonyms and think about underlying intent of keyword phrases you are evaluating. If it feels like a slog, remember this counts as market research too. You can’t be an expert and not have an awareness of the existing demand on your subject matter and services. And you might even uncover problems that result in new ideas for offerings.
It’s not really research if you don’t look at the Google results (SERPs)
A lot of searches on a very specific phrasing might indicate college students Googling for a homework assignment. So look at the actual SERPs, that will be your best hint as to what Google thinks it is about. The SERP structure is Google’s best guess based on lots of user data as to the right mix of intent to satisfy for that page.
Probably ignore keyword metrics
Don’t pay too much attention to the metrics on a term-by-term basis like “keyword difficulty”, or “CPC” or “search volume”, these are all pretty flawed. At most, search volume from third party tools like SEMRush should be used for ball parking or sorting lists, not a yay or nay for a term making it onto your keyword sheet. Let everyone else chase the same “low difficulty” “high CPC” “high volume” terms.
When you see search volume estimates 10x to 30x off the actual impressions showing in GSC or G Ads all the time, you start to trust common sense things like using keyword specificity, keyword length, and SERP features present to estimate potential search volume over third party search volume estimates.
Exceptions abound here, like getting stakeholder buy-in, layering in a bunch of real PPC data, or ecom, but for your purposes but I bet you do or would rely on those metrics to your detriment.
Mapping keywords to content
Think about search intent first, content topics ideation later
Your experience should serve you here. You often already know the approximate value of a user interested in x topic, because you (ideally) have experience with prospects this way.
After collecting your terms list, cluster your related rows of terms data together going down the line with the shorter head terms at the top and longer terms at the bottom of those groupings. The more similar in meaning the closer they should be together, and if they are same intent, you can collapse all but the clearest one.
If it gets unruly, split off into tabs. At the end of some good keyword research you should have a rough understanding of the natural edges on topics and subtopics that would make sense for a given piece of content.
Primary keywords and secondary keywords
Resist the urge to think in terms of a single primary keyword and a single secondary keyword for a piece of content. Matter of fact, resist the temptation to think about grouping keywords for making a given piece of content at all during this stage. Initially, you are simply trying to find rows with similar and related search intent that add up to patterns with relatively easy to find edges for pieces of content.
How long should the page be?
When ideating for a list of keyword rows as mapping to a piece of content you plan to create or refresh, the right level of breadth and depth matters. You’ll hear lots of rules of thumb based on (bad) usually context-specific correlational research. Ignore those. Use your head instead.
The answer varies based on context and what works for your audience. I have had many people tell me my emails are too long. Should I adjust for them or focus on serving the audience who wants to go deeper? Probably both.
The more competitively dense pages serving a cluster of keyword terms happens to be, the more important it is to factor for the length of what is ranking where into your decisions.
If the search intent can be answered in two sentences, don’t write an article about it when it’ll go to the rich snippet knowledge box on the Google result anyway. Alternatively, if you just want the rich snippet for the two sentence answer to a popular search intent, write about it and plainly put the summary / definition / most important part upfront.
Go as deep as the intent demands
The spectrum varies for how deep someone would naturally go given some search intent. “When was Lincoln shot” has a single date as the answer. “Why was Lincoln shot,” implies a much more comprehensive treatment is required.
If you pick a popular topic, you have to go all in, but that does not necessarily mean writing a 10,000 word guide. Instead, you could curate how to get around your site for topics you’ve covered that would relate to that guide and end up with a great alternative Google will rank to all the 10k word guides out there on x.
The optimal level of granularity
The level of granularity and the edges on a given piece of content are what make it different. So you can expand on a topic as a section in a longer guide type article and still have your shorter article exist, just don’t link to the shorter article from the section on the longer guide. That’s not an SEO trick, just common sense – you don’t need to point users to read a more succinct version of what they just read.
The opposite is also fine, and even optimal for search: write a guide as a sort of expanded table of contents with summaries of sections and curated links for more information to those corresponding pages. Depending on what you are trying to accomplish, there is no wrong format here. Both ways can be optimal for users in different situations with different intent.
On writing guides for SEO
When ultra long is better for guide style content
A common mistake to avoid: know that a deep dive on a topic in a long comprehensive guide is more of a bottom of funnel piece of content than you think of it as. Someone has to be committed to learning about a topic to read something that will take 20 minutes, so it probably won’t rank for the head term on that topic unless you are really backing it up with site-wide E-A-T and already the accepted authority, or you have some rich snippet and people also ask answers that can drive traffic to the site.
The freshman textbook trick
When you do write guide-like content, imagine you are writing a text book for a broad range of people who will only have a very limited to approximate knowledge of the topic they are searching on. Understand that there is an array of backgrounds, varied experience and biases/assumptions they are bringing to your subject matter.
Picking a primary keyword for titles and headings
I know I said don’t think in terms of primary keywords, but for titles (page title tags, heading tags) you have to choose a primary.
Even though search engines are “semantic,” meaning they account for synonyms and context to make sense of words of topics, they still like to return results in the pattern of language the user is using. A bit like how salespeople mirror you to make you feel comfortable, search engines need to signal that the results are relevant to a person who doesn’t have expert context on something they are searching.
So if I’m searching “how to differentiate my services” Google probably wouldn’t return a “comprehensive guide to positioning,” unless there was insufficient results to serve up using the language of differentiation of services. It’s doable, but cost prohibitive and increases your risk over time because you really have to load up the signals.
Uncovering opportunity from SERP structure and competitive density
Ignore the “about 1,000,000 results in 1.03 seconds” thing at the top of the Google results page. It is not a useful heuristic for competitiveness. But do look at the titles and descriptions for the results. What patterns do you see? What sites do you recognize? How does the structure change throughout the Google results page, the hierarchy of content?
When you look at depth of competition on the first page, sometimes there is a clear line between second and third position, or fifth position and everything below it. Sometimes the rich snippets are bad or there is a high up people also ask section. Sometimes there are video card results. These inform where the opportunities exist.
If the top 2-3 results in Google for a big term are more sizable competitors, you may need extra effort, but probably less than you would think. It is often totally doable to dislodge them from that position and/or be fine with the 4th position.
When 4th position is better than 1st
I suspect you are playing the long game so you may need to plan for a longer term answer to that query. But don’t put off what could getting traffic this quarter over it.
Remember we are looking for the right amount of effort to returns ratio over the right period of time. If a contrarian piece can land you in the 4th spot with not much additional effort in a month or so because it provides Google the opportunity to add some diversity in those SERP results, then writing a similar textbook style article as the first three results is a waste of your time.
If there are three strong results, then you should be able to get in the top 4th to 6th position. That is still worthwhile, even with lots of SERP features because you cannot know the how varied the amount of personalization is
You are competing for attention
Related, if you’re going to do SEO, be a competitor.
Experts don’t like to think of other experts in their field as competitors. Maybe they were bad at sports in high school or read a book about having an abundance mindset in the early 2000s. Whatever the reason, the fact is that user attention is a finite resource, as humans we have not solved multi-threading, we cannot read two articles at the same time, and the higher you are in the SERPs the more likely your content will be read, and someone else’s won’t. So in SEO, you have competitors. If that really bothers you, take the competitors you are closest with and treat it like you’re all on the same team, band together at the top of the SERPs, refer each other, work together, but get to the top.
On repetition of content across pages
Don’t worry about repeating yourself across pages. It’s fine to be repetitive as long as the intent or value is different. When you hear talk about duplicate content, it should really refer to 100% plagiarism or the whole page being copy-and-pasted or dynamic templating that generates duplicate content. Even then, we are pretty certain Google does not penalize sites for this anymore, they just pick a primary page and ignore duplicates.
When you’re trying to refresh content you already wrote
(or you want to amplify something you are about to publish)
I find asking myself these questions helpful:
- Zoom out topically for a moment. Where does this piece of content fit in the ontology of your target expertise? Now go put it there.
- What other content does this support? Link to it in an additional resources or related resources section on the page. If you want to “boost” the page you’re linking to, then link to it inline and higher.
- What other content would support this thing you just wrote? Identify it and link to it from those pages where it makes sense for the user.
If you review old content and find hard-to-organize patterns or formats, like you write a lot of list-type posts, ie. here are five things to think about when you x, where y is an array of topics you typically create content about, then it is okay to keep that content and have five clear places to link.
Know where this content fits in the bigger plan
Most experts have not architected the ontology of their expertise, an exception might be writing a book, but even then a book has much tighter edges than a body of work churned out over a career of cultivating expertise.
Once you physically map your taxonomies, and label your content appropriately, you will see glaring holes in what you should be writing about next. Not because a keyword tool said it was a good opportunity, but because it connects the dots for users on a path and you just realized you don’t have an answer, at least not on your site.