- Many SEO signals and perceptions can make it difficult for content managers to evaluate their on-page performance.
- For industries with niche semantics, this problem becomes more complex and larger.
- They present highly specific scenarios to content planning, each with their own lexicons and semantic relations
- Zach Wales, Senior SEO Strategist at Brainlabs uses results from a competitive analysis to help him understand how to assess your on-page game.
Search engine optimization is more difficult for industries that use niche terminology such as medical or scientific ecommerce brands. They present highly specific scenarios with unique semantic relationships and lexicons to the content planning process.
SEO is complex. There are many layers, starting with technical and ending with content. All of them aim to optimize for various search engine ranking signals. Some of these are moving targets.
How can you approach on-page optimization in such a complex space? Recently, we had the pleasure of performing a long competitive analysis for one of these clients.
We came away with a repeatable process that allowed us to analyze on-page content in complex semantic spaces.
The challenge: Putting findings into practice
It is important to clearly define the problem at the beginning of any analysis. Our goal was to translate our findings into actionable on-page actions, with priorities.
We would then compare the keyword ranking performance for our client’s domain with that of five of its chosen competitors.
We needed to find areas in the client’s website that were being lost to competitors in terms of keyword rankings. To prioritize, we had to identify the areas of client’s website content that were losing to competitors in keyword rankings.
Two additional sub-challenges added to the complexity:
- Volume of keyword data. However, scientific industries are different. They are considered “niche” because their semantics are not easily accessible, including keyword research tools. However, their keyword potential is extensive.
- The market was already dominated by our client. Using keyword gap analysis tools, it was clear that our client was dominating the market in all product categories. They were still losing traffic to these five competitors, even though it seemed like they were dealing with a random and spread-out amount of cases. These losses, taken together in an incremental fashion, had significant effects on their web traffic.
This is where you’ll find the needle in the haystack analogy.
We had to:
Determine the areas where keyword rank loss is having an incremental effect — this will help us prioritize;
Map these keyword trends to their appropriate stage of the marketing funnel (from the informational top-of the funnel to the transactional bottom.
Remove off-page factors such as backlink equity, Core web Vitals, and page speed metrics in order to…
Cases in which competitor pages were ranked higher than our client’s based on their on-page techniques and
In order to help our client adapt their content to an effective on-page strategy, identify the most successful techniques.
How to spot trends among a sea of data
It doesn’t mean that the data sets you work with are too large or lack apparent trends. This is only a sign that you need to change the way you view the data.
We are not claiming that our approach is the best. It was one that made sense to address another challenge, which is also common to the industry. Ahrefs and Semrush have different intent measures — “Informational,” “Navigational,” “Commercial,” and “Transactional” — that aren’t very reliable.
We used this approach to spot trends in a sea data:
Step 1. Step 1.
Numbers don’t lie. We trimmed the dataset in half due to lack of reliable intent data. Keywords with MSVs greater than 200 were excluded. Keywords below 200 had their MSVs reduced to 200. These were graphed and it produced a classic long-tail/short-tail curve.
This provided a proxy for funnel mapping. Short-tail keywords (high-MSV and broad focus) could be most closely associated with the upper funnel. Long-tail keywords were more searched, but they are also more focused and could be used as a proxy for lower funnel.
This helped us manage the multi-million-plus keywords our tools generated for the client as well as five of its competitors websites. Even though you can export data in batches using Google Drive, Google Drive and your device’s RAM will not want to deal with this much data.
Step 2. Step 2.
The term “keyword-operative root keyword” refers to root words that are similar to all or most keywords in a particular topic or content type. The common root word for most keywords is “dna”, which refers to DNA lab products that our client sells. The root word “protocols”, which is used to describe many keywords found in informational, upper-funnel content, is also a common root word.
This list was created by combining our long- and short-tail data (exported using Semrush’s Keyword Gap Analysis tool) into two spreadsheets. From these spreadsheets we could view the shared keyword rankings for our client and five competitors. These spreadsheets were equipped with data filters and formulas to score each keyword relative to the six web sites analyzed.
We also compiled a list from our client’s product groups and brainstormed possible keyword-operative root keywords. We then filtered the data and identified trends such as the number and sum of MSVs for each keyword.
To quantify the importance of a trend, we used a calculation that included average position, MSV and industry click-through rate. We could assign a numerical value to a competitor if they had a keyword ranking advantage in a particular subset of keywords.
Step 3. Step 3.
It is important to understand page templates’ role in guiding you to your goals, such as mapping keyword trends to your marketing funnel. Why?
Page speed is an important ranking signal to consider. Ecommerce websites often use content templates that correspond to each stage of the funnel.
All six of their competitors had different templates for top-funnel, middle-, and bottom-funnel content in this instance:
- Top-funnel Text-heavy informational content in what was formerly called “Learning Resources” or similar;
- Middle-funnel template: This page contains text-heavy information about a product category with links to the products and visual content such as diagrams, videos, and other content. It is essentially called the Product Landing Page (PLP).
- Bottom-funnel template: Product Detail Pages (PDP), with concise, conversion-oriented texts and buying calls-to-action
Step 4. Step 4.
Following Step 2, we began to see keyword ranking trends. We just needed to map them to the appropriate funnel stages.
This was made easier by having content templates identified and the data being divided into short- and long-tail. We focused on the trends that were outranking the client’s website.
The added benefit of identifying content templates was the ability to see where our clients are outranked on a particular keyword. For example, our client’s winning webpage was built using a PLP (content-rich, optimized PLP), while our client was using a PDP.
Step 5. Step 5.
Our goal was to analyze and identify on-page techniques. We had to exclude off-page factors such as link equity and page speeds. We were looking for cases in which one page ranked higher than another on a common keyword, despite having lower link equity or page speed scores.
Despite all the advances Google has made in processing semantics (e.g. BERT, or the Helpful Content Update), there are still instances where pages with short text content can outrank pages that have longer, more optimized text content. This is due to link equity.
We created a “SEO scorecard” for each page under investigation to rule out these factors. The scorecard listed the rank-signal-worthy attributes that each page had in its SEO favor. These included Semrush’s page authority score and the number of external vs internal inlinks. The presence and types Schema markup and Core Web Vitals stats.
Scorecards also include on-page factors such as the number of headers and subheaders (H1,H2, H3 …), use keywords in alt tags, meta titles & character counts, and even page words count. This gave a high-level overview of the on-page performance, before we dive into the actual content.
We only selected cases in which the SEO scorecard of our client’s pages was lower than its competitors when comparing them to their competitors. These are some of the most notable findings.
It really works to add H3 tags to product names
OrangeValley’s Koen Lemans published a Semrush article titled SEO Split Test Results: Adding H3 tags to product names on Ecommerce Category Pages. This study was especially timely, since it confirmed what we had seen in our competitive analysis.
For those who are familiar with on-page SEO, it is wise to place keywords in the h3> HTML formatting (or any other level of h …>). Google crawls the text before it reaches the paragraph copy. This is a well-known ranking signal.
Ecommerce clients are prone to abandon the product name when it comes to SEO-informed content plan. This is despite their best intentions. They want to find the perfect on-page recipe to use a non-brand keyword. Because the brand assumes that it will rank higher than others for its product names, the value of the product name becomes an issue.
Editors may decide, in the course of this thought process, to list product names on a PLP bolded p> text instead of as a h3>, h4>. This is an obvious missed opportunity.
We found that this on page tactic worked even better when the h> tag product name was linked (index and follow) to the PDP AND accompanied by a sentence description below the product name.
This contrasts with the product landing page (PLP), which has plenty of supporting copy and lists products only as hyperlinked names without any descriptive text.
Word count is important, but h> count is very likely to be an important factor
It’s not unusual to find PLPs in the ecommerce market that haven’t been visited by the content fairy. Uninteresting grid of product names and images.
However, when two PLPs from this variety were matched-to-match over the same keyword in a case, the sheer quantity of h> tags appeared to be the only factor that ranked one of them above their competitors’ PLPs. These PLPs also had higher link equity.
This is the take-away: If you don’t have the time to update your PLPs with landing content, at the very least, set product names to h> tags. You can also increase the number of hyperlinked products (e.g. set the page to load 6 rows instead of 4.
What about word count? Google’s John Mueller has confirmed that word count does not affect the ranking of search results, but this is still a topic under discussion. Our competitive analyses cannot provide any conclusive evidence about word count. We can only say that it is a component of what we found that…
Your content will define the topic.
Brian Dean, a Backlinko user, proved that a single page can rank for hundreds of target keywords. This is true if your copy covers all aspects of the topic that unites hundreds of keywords.
This approach is useful for long-form content marketing, but it’s not as applicable to ecommerce. Alternative to this is to create pages that are interlinked intentionally and logically (from an UX standpoint) that cover all aspects of the topic.
This content should address the questions that people have at each stage of the awareness-to-purchase cycle (i.e., the funnel). It should be able to define niche terminology as well as spell out acronyms. It should be easily accessible.
One standout case in our analysis was that a competitor’s page held position 1 for a lucrative keywords, while our client’s website and those of other competitors could not even achieve a page 1. The keyword was being addressed head-on by all six websites. They also had better link equity.
What was the advantage that the winner had over the rest? It happened that in this lone instance, its product was being marketed to a high-school teacher/administrator audience, rather than a PhD-level, corporate, governmental or university scientist. This virtue alone made their marketing copy far more accessible to laymen and was approved by Google.
It is important not to simplify technical industry jargon. It does highlight the importance of telling every aspect of the story within the topic vertical.
SEO experts who are specialists in the biotech and scientific industries place a lot of emphasis on content planning from a top-down perspective.
After completing the competitive analysis for my client, I found these posts. Because the “Findings to-Action” section in our study recommended something similar, this topic-takeover emphasis was validated.
Map subjects to the funnelspan styling=”font-weight 400 ;”>. Before you do keyword research, map the broad topics and subtopics to their respective spots in the informational and consumer funnel. Identify:
Questions to ask & problems to solve at each stage of the funnel
Keyword possibilities that lead to these respective stages
How many pages are needed to rank for these keywords
These are the best website templates to accommodate this content
The header and internal linking strategy between these pages
In scientific industries, there are two distinct audiences. This is in contrast to other industries that use common language. The first is the AI-driven audience that search engine bots use to scan this complex semantic terrain looking for meaning and symmetry. The other, however, is a human with an already-mastered understanding of this symmetry and a highly skilled mind capable of deciphering it.
Content planning and delivery must be well-organized to make the best use of time and provide the best user experience. As an organizing tool, the well-known marketing funnel works exceptionally well. All that is left is the hard work of applying this content-rich, topic-coverage approach.
Zach Wales, Senior SEO Strategist at Brainlabs.
Sign up for the Search Engine Watch newsletter to get insights into SEO, the search landscape and digital marketing.
Join us on LinkedIn or Twitter to continue the conversation.
The post Is your on-page point in a sea of signals? Search Engine Watch first published this article.
Did you miss our previous article…