On a given Google search, users are 20x more likely to click the organic search result over a paid ad. According to Google’s Andrey Lipattsev, backlinks and content are the most important, impactful ranking factors. Welcome to our next enterprise SEO training. Last month we talked about Enterprise SEO for Developers (if you’re new to enterprise SEO, content at scale and/or a developer we recommend reading that first). This time, we’re focussing on content at scale.
Content at Scale: Agenda
Content at Scale
Optimising for Human Outcomes
Perhaps the biggest change with SEO is that in the past the algorithm had more holes in it. SEO would focus on finding these holes and exploiting them. Over the past few years, machine learning has made the algorithms more and more complicated and complex, meaning there are fewer holes to exploit.
SEO tactics that once worked, no longer work. SEOs can no longer simply outsmart Google and so in some ways the SEO of the past is dead. BUT, SEO is still alive and has simply evolved to embrace machine learning and optimise for human outcomes, for example, getting the user to find a restaurant they wanted or to rate something higher, or to buy something. While there may not be many holes in the algorithms anymore, you can optimise for people and get users to convert.
The way that I think about the difference between SEO then and now is that rather than optimising for the algorithm, you’re optimising for human outcomes and getting people to do more of what Google is looking to optimise for.
Challenges of Content at Scale
SEO Pyramid of Needs
Content at Scale
- Clear keyword and search intent targeting
- Minimise confusion among similar/duplicate pages
- Comprehensive pages with supportive content
- Topic expertise shown by breadth and depth of content
SEO Product Strategy Pillars
Conceived through the global team for a client, we collaborated together to define and align on product strategy pillars and techniques.
SEO is about Relative Performance Against Competitors
- What is your relative authority compared to other sites that are ranking?
- What is the content quality like in comparison?
- Are you providing a good user experience?
- What is your conversion rate?
Each of these dimensions are inputs that decide whether you rank and to what extent you rank. Whenever you’re looking at a query and trying to figure out where you should stand, consider these dimensions and compare yourself against the other sites that are ranking.
Content from Google’s Goggles
Google is evolving. From a Search Engine …
To an Answer Engine
Google algorithms have evolved from keywords to intent and context. Technical SEO, keywords and links still matter but now context and fuzzy logic is also used to determine rank.
Queries Become Entities
Google breaks a query into a series of entities.
Entities are Extracted from Content to Inform Ranking
Google Understands Patterns of Quality
Google is a machine. It does not understand quality. It interprets patterns. It matches against quality benchmarks (direct human input or derived from machine computation and algorithms) and relies on user data for validation.
Google Uses Term Frequency in its Index
Term Frequency – Inverse Document Frequency measures the importance of a keyword phrase by comparing it in a large set of documents.
Synonyms and Close Variants Should be Included
Search engines possess vast corpuses of synonyms and close variants for billions of phrases. This allows us to enrich content with natural text to provide greater meaning, using Steming + frequency (TF-IDF’s) and Lemmatization to group words like “better” and “good”.
Content Duplication & Low Quality Pages
Content duplication and low quality pages can affect your rank. When you’re looking at content at scale for enterprise businesses, this is a very common problem. Though common, one that should be fixed.
How to Fix Duplicate Pages
So, how do we fix all these duplicate pages? By effectively showcasing the existing supply to satisfy non-trivial audience demand.
- Balance opportunity to provide specific landing page experience vs the number of pages we want to provide to Google
- Look at how many searches exist for a given category/subcategory/attribute/location combination
- Use internal search data and patterns to build rules
Reliable Solutions
- Avoid it! Internally and externally (Double Serving, Affiliate Content, Syndication)
- Identify it! (Ryte Reports, “Quotation Searches”, HTML Improvements in Google Search Console, etc.)
- Rewrite or enrich content
- NOINDEX
- Enforce Canonical URL via 301 (lookup, fix, truncate)
Post & Pray Solutions
These might or might not work:
- Canonical Tag
- Google Search Console Parameter Handling
- Robots.txt
Put Your Site on a Diet
Source: Philipp Klöckner
Cut the crap
If your website has more than 100,000 pages it probably needs to go on a diet. Millions of indexed HTML documents are not an asset but a liability. Indexing everything is inefficient by definition.
Over 80% of your website is usually deadweight which pulls down your best pages. Non-transactional content is (most likely) overrated. Inventory equals content. What can you do? Analyse your potential with an organic landing page report.
Identify Low Quality Pages by Page-type
Easy NOINDEX Targets
- „no results“ pages
- Few results pages (revise item threshold?)
- Single review pages, other low-quality UGC (user-generated content)
- Bulk product pages
- Any duplicate pages
- Faceted search without search demand
- Expired offers/ads
- Parameters, etc…
Data to Support Page Quality Decisions
- Revenue distribution on landing pages (Google Analytics)
- Engagement and commercial metrics per page-type
- Conversion rate related to inventory count
- Demand-Data (search volume, PPC traffic, navigational traffic)
- „Indexation Gap“ (sitemaps, submitted vs indexed)
- Crawling activity (server logs)
E.A.T. – Expertise, Authority, Trust
On top of the constant changes that are made in Google’s algorithm, they also employ humans to assess the quality of the search results. Google contracts with over 10,000 search quality raters worldwide to evaluate its search results.
Raters are given actual searches to conduct, drawn from real searches that happen on Google. They then rate the quality of pages that appear in the top results. Quality raters use a playbook to help them review whether a site meets the needs of those who perform searches.
Quality raters keep EAT in mind to judge how well a site or page provides what users need whilst looking at the overall user experience and whether the content meets their standards.
How Do You Stack Up?
When you’re looking at implementing EAT, consider the reason why users should choose your site over your competition. Ultimately, EAT could still have a direct impact on how well Google receives, and ultimately ranks your website.
If your accountant lacks expertise, authority and trust, you’re likely to look around for an alternative. Google’s users are the same. When they find a domain or page that lacks expertise, authority and trustworthiness, they’ll look around for an alternative in the form of another page.
Google is also training their algorithm to assess EAT, using them as signals to determine how much faith they should put in a businesses website to satisfy user needs. If your website fails to satisfy user needs, Google will choose someone else for the job.
Put simply, if Google finds another domain or page provides a better customer experience than your own, Google will promote them instead. This translates into lost rankings, lost traffic and lost revenue.
Content at Scale – Relevance vs Reach
Content for Relevance
Provide and answer queries.
Content for Reach
Maximise brand recognition and engagement
Relevance vs Reach
Find Content Associated with Core Keywords
Time to look at keywords beyond your core keywords. Your objective should be to find terms that will drive either long-tail searches or secondary information.
Build Hub and Spoke Content at Scale
Google has changed its algorithm to favour topic-based content. As a result, we as SEOs should explore a new way of link related content under a “topic clusters” model.
What this means is linking all internal content to a topic or a “pillar page”. The advantage of this is that all relevant resources should point towards a strong core keywords page.
Content at Scale – Client Example
For a client, Open Colleges, we created a hub and spoke content strategy. SEO needed to strengthen its authority and capture a larger chunk of the overall customer journey. We saw an opportunity to create middle of the funnel content revolving around industry advice and interviews. This would bring traffic and leads earlier in the decision-making process.
Middle of the funnel content is often neglected as it doesn’t drive many leads, sales or links, reach and shares. However, this is still a vital part of the user journey and leads to the bottom of the funnel content (money pages).
For Open Colleges, we built content hubs to target different audiences. We created the careers site. A site specifically created for prospective students. Typically targeting higher-level consideration.
Instead of targeting keywords close to the field of studies, we identified key drivers such as career development, better salaries and increased professional fulfilment.
We looked at resources people would go through.
Creating Career Hub Content
We limited our focus on content production due to priority being placed on optimisation and a streamlined journey to conversion.
Bottom of the Funnel
We created course pages to facilitate lead generation.
Middle of the Funnel
We created career content including:
- Employment outlook
- Expert series
- Career resources
- Quizzes
- Student stories
We also created an InformED (educators content) piece outside of business as usual to obtain links to our content and resources.
Each page of the hub has the potential to rank for a set of keywords.
Relevance – Assist the Customer Journey / Improve UX
Target by Intent, Not by Keyword
Everything starts with intent and one of the first things that I do when I’m building out strategies is to look at all the keywords and try to categorise them by intent. Usually, I’ll come up with these four intents for commerce, particularly in marketplaces.
- Learn
- Browse
- Filter
- Buy
Leverage Proprietary Research and Visuals
In-depth data is always in high demand, and many companies don’t want to invest the time to pursue it or create it themselves.
That’s where proprietary research can help distinguish your SEO efforts from those of competitors.
This is a great way to create wholly unique content that’s highly relevant to your audience and searchers. It’s certainly not easy, but it’s a worthwhile investment.
You can gather data by sending surveys to your followers and aggregating the data or compile industry statistics or expert opinions. Data-driven and unique content is incredibly linkable and because it’s unique it’s the only link option for those searching.
SEMrush found that creating cornerstone assets centred on data is the number 1 most efficient link building technique.
Manual vs Automated Content
Data across 40k Unbounce landing pages show copy was 2x as influential as design in why visitors converted. The copy is the most significant factor when determining whether users buy or not. Enterprise brands need to invest in better copy by writing better copy and optimising what they already have.
FAQs
Including FAQs on product pages can enrich content and provide answers to common questions your users will look for before deciding to buy. Use keyword research and sites like answerthepublic.com to find the most common questions and provide relevant high-quality answers.
Product Classification
Better product classification enables better pages
The more we can group products with certainty, the more opportunity we have to create valuable content for said product. It can also help manage price range more effectively and open opportunities for more interesting data analysis.
Better classification can improve UX dramatically
Simple extra info gathered from your site data can enhance the user journey and add valuable info for Google too. Being able to tell users whether this is a good deal or a common item can substantially improve the experience.
Automated Content
Automated content can help fill a gap.
Automated content can be a good way to fill in gaps but it doesn’t always equal high quality, entirely relevant content. For example, the most obvious thing to talk about in the example above is the amazing view. Instead, the automated content is covering meeting rooms. Still relevant, but we’re missing some important information.
Level Up Your Automated Content
Aggregate your customer feedback to level up your automated content. Use common attributes in reviews to create unique content.
Beyond Text
Visual Search
Image search usually accounts for a small percentage of the total for enterprise businesses. While web and video search should be focussed on, it’s often a good idea to also look further into image search, especially for eCommerce.
Image Search is Monetised
Image search is becoming monetised. Rich snippets are now showing on image searches too.
Enterprise Video Marketing
Video makes up 80 percent of all web traffic and is clearly the way of the future, too many organisations still flounder with their video marketing.
Bonus Video Content
Transcribe your videos for extra content.
Here is an example of Google Search Console data from 2 different transcribed video posts for a client, picking up organic traction.
Voice Search
The average Domain Rating of a voice search result is 76.8. Therefore, Google’s Voice Search Algorithm may emphasize trusted, authoritative domains.
Pages using Schema only account for 36.4% of voice search results.
Nearly 10% of voice search queries begin as actual questions – who, what, when, where, why, how – as compared to only 3.7% of text-based searches.
Content that’s easy to read and understand ranks well in voice search.
Backlinko found that the average voice search answer is just 29 words.
Longform content tends to rank best in Google Home searches and no you do not need to have the exact keyword in the title. In fact, the exact keyword appears in the title just 1.71% of the time.
Featured Snippets
Appearing in featured snippets appears to be extremely helpful for voice search SEO. In other words, the higher a page ranks for a specific keyword, the more likely Google will choose that page as the voice search result.
In short, if you want to appear in voice search:
- Load quickly (under 4.6s)
- HTTPS
- FAQ page
- Long-form content (2.3k words or more)
- Contain concise answers (29 words or less)
- Written at Year 9/ 9th grade/ Secondary School level
- Highly shareable (evergreen works)
- Ranks in top 3 desktop search (generally function of above)
- Appear as featured snippet
- Page is on a high authority site (Domain Authority over 70 for most cases)
Content At Scale: Enterprise SEO Training
Google wants the same things as human users. It just has different accessibility needs and processes information differently. Improved UX for users that Google also understands, often results in better rankings and organic growth.
Reduce your workload and improve efficiencies, by allowing your SEO team to tweak and adjust site performance.
Need some help? We help many enterprise businesses and their developers with SEO. Get a free consultation.
