loader image

Welcome back to our multi-part series on the basics of SEO. In our last post (here) we looked at the fundamentals of technical SEO, and now we’re going to move on to look at keyword research (finding search terms to optimise your site for) and on page optimisation (tagging and structuring the page in the best way for both users and search engines) as well as some guidelines on producing content for onsite SEO.

Keyword Research for Onsite SEO

What kind of keywords are there?

Generally speaking, SEO keywords are divided into three main formats with 3 different types of user intent as well. The main (length based) types are:

  • Short tail
  • Medium tail
  • Long tail

These can interchangeably refer to the raw length of the keyword entered into Google or the search volume. Longer, more detailed queries will almost always have a lower search volume. For example, the short tail keyword ‘Ferrari’ in the UK has a monthly search volume of around 135,000, whereas the longer tail keyword ‘How much does a Ferrari cost?’ only has a search volume of 390 searches per month.

User intent, on the other hand, describes what users want to get out of their query and these can be classified as:

Navigational queries

This is when a user types a query in with an expectation to find a specific website or part of that website. For example, ‘Wikipedia’.

Informational queries

Users seeking specific information are more likely to use what/when/where/why and similar queries. as far as online commerce is concerned, this might revolve around product reviews, media, FAQs or finding out about a company. For reasons we’ll get to shortly, these tend to be more appropriate to target than the commercial queries described below!

Commercial queries

It would be sensible to assume that in order to do well out of Google, you just need to get your main landing pages and/or product pages showing on page one, right? However, as we recently wrote, this has become a lot more difficult in recent times:

Recent data from Sparktoro and Jumpshot shows that a staggering 50.33% of searches now result in no clicks, with organic click share dropping from 54.01% in 2016 to just 46.12% in Q2 2019. Also worryingly for organic search marketers is the trend towards more paid ad clicks on a year by year basis, particularly on mobile where paid CTR, as a percentage of all clicks, rose from 3.29% in 2016 to 11.38% in 2019.

Let's work together

If you want to talk to our specialist team about how we can help you with your digital marketing, talk to our team today.

The enormous number of searches returning no clicks relates to instant answers on mobile (the example we usually give being a query like “what age is Brad Pitt?”), so this aspect isn’t something to be unduly alarmed about, as most marketers will have no need to rank for these sorts of queries anyway. It is worth noting though that ads are becoming more and more prominent in search results. Therefore ‘ranking on page one’ no longer generates the same proportion of traffic that it used to in most cases. In order to combat this, we can instead:

  • Target low-competition, informational and content-focused keywords with non-commercial content (which is easier to rank for and will still capture interested users)
  • Use content pieces to drive top-of funnel visitors towards landing pages i.e. visitors at the start of their buying journey

We’ll look at how to do this in the next step:

Keyword research for informational content

For both keyword research and the on page optimisation example below, we’ll assume that we’re an online retailer selling products for dogs, who also produce content. We’re going to focus on informational content as mentioned above. Whilst everyone else will be scrambling around trying to rank for ‘dog toys’ against a huge number of ads and high-authority domains, we can work on building a steady stream of organic traffic through providing useful secondary content to our target audience.

Researching keywords with forums

As we’ve mentioned before on the site, forums are great for keyword research because:

1. Forums are full of your vertical’s consumers and enthusiasts.

2. Forums tend to be where lots of questions get asked and  reveal what your audience desires, has problems with and fears.

3. They tend to be well-organised into subsections which helps us perform research more easily.

4. Few marketers use them for research, particularly in terms of finding keywords, so untapped keywords, content gaps and content opportunities may present themselves more easily.

Although this particular guide of ours focuses more on how to scale keyword discovery across an entire forum or several forums, in this example, we’ll just have a browse manually to see if we can discover any opportunities. To do so, we’ll be having a look at petforums.co.uk:

Just by typing ‘toys’ in the search bar, we’ve found:

From this, we could come up with a few article ideas – obvious ones would be roundups of the best durable, outdoor or puzzle-related toys and similar ideas, but we could also look at topics such as toy safety or what to do if your pet starts choking.

The only thing to be careful of here is to make sure we don’t write about topics or search terms where there’s little or no search interest. To check this, we can either use the Google Keyword Planner (if you have an Ads account) or alternatively look at search interest over time on Google Trends. It’s worth mentioning at this point that it’s best to cluster several longtail queries together. Writing one piece of content for each individual longtail query that you research could lead to thin content problems (which we’ll cover below).

Keyword research for commercial content

Researching keywords for products and landing pages can be a bit trickier than with content. Although this should be simpler for ecommerce sites, as the targeted keyword will usually just be the product or category name! For lead generation, it’s often the case that you may have a branded product or service. When this happens, it’s usually best to include the generic terminology in the content. For example, if we had an imaginary service providing accounting services called Easybooks, our main service landing page would need to make mention of the core service (accounting software) in prominent places such as the title tag (see below) rather than having meta and heading tags that just made references to ‘Easybooks’.

On-page optimisation (Onsite SEO)

Now that we’ve chosen the keyword (s) that we’re going to incorporate into our content, it’s worth going over some basic on-page optimisation tips to help give your content the best chance of competing in search results. The below checklist applies mostly to written content in HTML format (blog posts, whitepapers and so on) as SEO for images, video and unique file types like PDFs require a different approach.

Page URL

This is the only part of the on-page process that ideally needs to be correct the first time around. Everything else can be changed easily, but redirects will need to be set up if the URL changes subsequently. If that happens multiple times, it can be a pain to manage and can also impact SEO. Webpage authority and site speed can both suffer if search engines have to trawl their way through several different URLs to get to their final destination.

What makes a good URL?

Good URLs tend to be:

Short, readable for humans and descriptive of page content

How your page URL displays in both search engines and browsers is important. Who wants to be deciphering a long, random-looking string of letters and numbers? Try and keep your URLs fairly short and easy to read. For example


In a sensible structure

the structure of your URLs is important as well. Although this will usually matter less for content and lead generation sites as the site structure will tend to be simpler. Search engines will start at the ‘root’ of your site and then work their way down. For example, an ecommerce site might have the following structure:


Root folder (home page) Main category Subcategory Product URL
Example.com /dogs /dog-toys /inflatable-dog-toy

As a general rule, try not to have too many subfolders if you can (this is also known as having a ‘flat’ site architecture).

Let's work together

If you want to talk to our specialist team about how we can help you with your digital marketing, talk to our team today.

Forbidden characters

It’s advisable to avoid the use of these sorts of characters in your URLs, as it will otherwise not function correctly due to compliance standards:

Space  ”  <  >  #  %  |

Title Tags and Metadescriptions

Title tags and metadescriptions comprise of the text-based parts of a search result that a user will see when your site shows up in Google search. In the above example, the bold blue text is the title tag (also known as the meta title) and the paragraph is the meta description. Although title tags have some weight in the Google algorithm, metadescriptions do not. Therefore, there’s no specific need to include keywords in them (although Google will often put in bold words that it thinks are relevant to the query in the search results, as can be seen above).

We’d recommend the following checklist for title tags:

  • Between 50-60 characters in length
  • Aim to have your main keywords at the front of the tag if possible
  • Having a brand affix at the end (in the form of | brandname). It generally looks professional and helps you stand out more as a brand in the search results
  • Emojis can be used, just don’t go overboard with them!

And this one for metadescriptions:

  • 130-155 characters
  • A metadescription is essentially a piece of ad copy for your site in Google. Try to provide the user with clear USPs (unique selling points), a call to action and accurately describe the content of the page. If possible, try to give them a reason to click rather than just begging them to visit your website instead of someone else’s!

What about meta keywords, don’t I need to add those as well?

Meta keywords have been ignored by the major search engines for some years now. This is because unscrupulous individuals were using them to stuff large amounts of irrelevant keywords onto their pages in an attempt to rank for phrases they didn’t deserve to. Search engines, as a result, decided over a decade ago to start ignoring the meta keywords field. If your site outputs meta keywords automatically this shouldn’t be a problem (unless it’s generating thousands of them per page which will bloat your site’s code), but otherwise it’s not worth the time or effort to put them.

Heading tags

These are essentially the same as what you might see in MS Word when formatting a document:

Heading tags (otherwise known as headings and subheadings) are great for both users and search engines:

  • For users, they help to break content down into more manageable chunks, making it easier to read whilst clearly marking out different parts of a piece of content
  • In a similar vein, search engines find it helpful when a document has a structure to it, as they can crawl it and extract information from the document more easily

Headings are classified from H1>H6, where H1 is the main document heading, H2s are subheadings, H3s are sub-subheadings, and so on. Generally, one H1, several H2s and possibly some H3s under the H2s should be sufficient. However,  this will depend on the format the document needs to take.

Word counts, thin and duplicate content

It’s worth mentioning at this point, three terms that have been hot topics in the SEO community the last few years and all are pertinent to creating content. We’ll look at how many words you should be writing for a piece of content and also discuss ‘thin’ and duplicated content.

Word counts

Unfortunately, there isn’t a simple answer to the question ‘how long should my content be?’ However, the most important aspect in terms of content length is this:

What would be appropriate to meet the needs of the searcher?

What do we mean by this? Well, an ecommerce product page selling paperclips probably wouldn’t need a lot of unique copy, whereas a product page for an iPad or a BMW lease will require a lot more detail. Similarly, someone wanting an in-depth, content-based guide to a topic probably isn’t going to be satisfied with a vague introduction that only comes to a couple of hundred words. On the other hand, if the searcher is just looking for a glossary definition of a specific industry term then a paragraph of text on a longer glossary page would be sufficient. Here’s some suggestions for what to aim for in terms of word counts, by type of page.

Page Type Word Count Notes
Ultimate/in-depth guide to a topic 2500-7500 words Topic dependent
Ecommerce product page for a simple product 100-200 words E.g. Paperclips
Ecommerce product page for a complex product 500-1000 words E.g. iPad or BMW leasing

Enabling customer reviews and/or product FAQs can help with this

Lead generation page Dependent on the page and type of lead being generated E.g. a SAAS lead generation page will typically rely much more on design, imagery and tabulated data than lots of text, so may only need 250-500 words

It’s worth saying that writing large amounts of unique copy is only really suitable for sites that are either small, or have substantial dedicated content and product teams. Don’t feel like you have to write hundreds of thousands of words of copy if you’re running a store by yourself!

Thin content overview

The opposite of the above is what’s known as ‘thin’ content. This is where there is insufficient text and/or imagery on the page to meet the needs of the searcher. This can happen due to a number of factors, for example:

  • Page template setup
  • Spam (could be external or user-generated e.g. on a forum)
  • Short-form, low-quality news articles
  • Unanswered Q+A topics

We’ll cover how to identify and deal with both this and duplicate content below.

Duplicate content overview

Duplicate content can take many forms. For example:

  • Products duplicated in multiple categories (could be listed in a main category, offers and a collection)
  • Different colours/sizes of product being of different URLs, but everything else being the same
  • Large sections of articles being copy/pasted into other articles

As far as we’re concerned in this article, the main thing is to not intentionally duplicate content on a site. The main body content of an article or content piece should be 90%+ unique content (duplicated products are often best solved with the canonical tag).

Let's work together

If you want to talk to our specialist team about how we can help you with your digital marketing, talk to our team today.

Why are thin and duplicate content bad?

Although there isn’t a Google-specific penalty as such for having thin or duplicated content on a site, it does cause a number of issues:

  • Dilution of the authority of the site’s pages – they will find it harder to rank for target terms than if they were concentrated in one more powerful version
  • Site reputation and authority – this will suffer with large amounts of duplicate or low quality content, resulting in traffic and leads/revenue being left on the table
  • Crawl inefficiency – if search engines are coming across large amounts of duplicate or uninteresting content it may be harder for them to access your unique and valuable content and they may not try as hard to find new content. Similarly, if content pieces are substantially duplicated they may be ignored altogether

Dealing with thin and duplicate content

Happily, there’s a great free tool that can be used to locate both thin and duplicate content on your website called  Siteliner.com. Once you’ve got to the site, simply paste in your website (it has a limit of several hundred pages, so works best on smaller websites).

Once it’s finished running,  go to the ‘duplicate content’ section as below. This will tell us:

  • Which pages on the site have duplicates elsewhere
  • The number of words that are the same
  • The overall percentage match
  • How many pages the page duplicates against

This can be very useful for identifying duplicate content and coming up with a plan of action.

Thin content

If we navigate to ‘your pages’ the number of words on the page is listed as well, which is great for establishing how lengthy a given page is:

What actions should I take?

This will essentially depend on how the thin/duplicate content was generated:

Small number of thin/duplicate pages

Updating these pages to be more detailed/removing duplicate content will usually do the track for smaller sites.

Larger (several hundred>thousands of pages) 

When dealing with large amounts of undesirable content at scale, it’s often easier to redirect or delete the content (apply 404 or 410 status codes. Ask your developer on this). Just be sure to check the sections of the site where the content is coming from in Analytics first to make sure that you won’t be deleting anything that brings in traffic or revenue.

Keyword density and keyword stuffing

Keyword density simply refers to the number of times your target keyword (s) appear on the page. Historically, when search engines were a lot less advanced, simply ‘stuffing’ your keywords into as many places as possible on the page and its metadata tended to be all you needed to do to rank highly! Now though, it’s much more sophisticated. Search engines use a wide variety of text, sentiment, entity and keyword recognition/analysis algorithms to the extent that they can usually tell what a page is about quite easily. They can also detect when words have just been stuffed onto a page easily as well and are likely to penalise a website for this! To make sure that we have an adequate, but not spammy, placement of our keywords on the page, all we need to do is the following:

  • Mention your main keyword (s) in the page (meta) title and H1 tag (main heading)
  • Do the same for the first few paragraphs of text on the page (i.e. mention your main keywords a couple of times naturally in the introductory copy)
  • Have related (not repeated) keywords in your subheadings
  • Write about it naturally –  as long as you know your subject, the search engines will pick up on the topic (s) discussed through various means (such as understanding semantic keyword relationships)

Internal and external links

Linking internally to your own pages can be a powerful way of directing visitors and generating link equity (if the page has links pointing to it) elsewhere on your site. It can often be a good idea to link internally to relevant pages or products. This is also a great way to send traffic to otherwise hard to promote products. If such products are featured in a post on the site’s blog for example.

Linking to external websites can be a good idea as well, depending on who it is that you’re linking to. Academic references for a medical article would be an example of good usage. Just remember to set the links to open in a new tab or window, so your users don’t forget about your site if they visit somewhere else in the meantime!

Optimising images

Optimising media can help to give an added boost to your page. For images this will generally entail adding ALT tags (which is simply writing a natural description of your images as though you were describing it to someone) and compressing them. Here’s some of our existing content on the subject of compressing images, which is important for pagespeed:

How to improve website page speed through image compression

Page indexability (once the page is live)

Wouldn’t it be embarrassing if the great post you’ve created couldn’t be found by search engines? Thankfully, there are a number of free and easily available add-ons for Chrome and Firefox that can check that a page is indexable. These add-ons can also break down useful features like the meta title and metadescription all in one place. One extension that we use and like is the Detailed extension. Upon clicking the icon, you’re presented with a helpful summary indicating the status of the page (if the page is not indexable, it will explain why).

That’s all for this week. Thanks for reading. Make sure you don’t miss out on the next instalment by liking our Facebook page or follow @Agency51UK on Twitter.

Ben Henderson

Ben Henderson

Ben Henderson is a SEO specialist at Agency51, and enjoys working on and writing about all aspects of technical SEO for a wide variety of websites and industries.

    10 Great North Way, York Business Park, York YO26 6RB

    © Copyright Agency51 2019. All rights reserved. All trademarks acknowledged. Agency51 Limited, trading as Agency51, is registered in England & Wales: No. 05516494. VAT Registration Number: 192 0540 24. Agency51 is part of the PureNet Group of Companies

    Privacy Policy     Cookie Policy