7 Steps To Semantic Content Excellence
What is semantic content optimization, why do you need it, and how can you implement it? Columnist Eric Enge answers these questions and more.
Lately, I have seen more and more people write about “Semantic Content Optimization,” the practice of tuning your Web pages to satisfy a larger percentage of visitors.
I’ve written about various aspects of it, too, including my article on the 100-User Model here on Search Engine Land.
In today’s post, I am going to outline a seven-step process for Semantic Content Optimization for one target Web page.
Why Does It Matter?
As I outlined in the 100-User Model post, Google is doing its best to evaluate content quality and user engagement with pages it chooses to rank high in the search results.
Higher user satisfaction with pages that rank high in Google’s search results means that those users will be more satisfied with Google’s search engine. In short:
The quality of your page is a ranking factor.
With that in mind, it makes sense to closely examine the most important pages of your site with a laser focus on their quality. As we have all known historically, this is something that will drive conversion and customer loyalty, but now we know that it will also drive SEO.
For more perspective on this, check out Rand Fishkin’s SlideShare deck that discusses what he calls a 2 Algorithm World. I particularly like his concept of “task completion.”
Consider the notion of 100 people coming to your Web page on a given search query. What percentage of those people will complete the task they had in mind?
The task they are trying to complete may be broader or far more complex than your page is currently designed for.
7 Steps To Semantic Content Optimization Greatness
Now that the stage is set, let’s start walking through the process to enhance the Semantic Content Optimization for one of your target Web pages.
1. Decide on the focus for the Web page. First of all, the process I am outlining here is not trivial. The first pages for which you pursue this effort should be the most important pages of your site. In short, your “big money pages” are the initial targets. It follows, therefore, that the terms you are targeting for those pages are very high-value ones — either high in volume or premium from a pricing perspective.
For example, if you have a page that sells hammocks, you don’t optimize for the phrase “backyard snoozing sling.” That is already obvious to anyone with SEO experience. What’s new is the focus on other areas related to hammocks that the page should address and what it is that differentiates it from other pages on the Web selling the same thing. Understanding how you differentiate is a key part of the process, because clearly expressed differentiation can drive user engagement with a page.
2. Identify the customer needs the page should satisfy. Once you understand your basic level of differentiation, you now need to start figuring out what other things a person looking for hammocks is looking for. Examples might be:
- Various options for styles and sizes
- Spare hooks for when one of the hooks used to hang the hammock gets misplaced
- Advice on where in your backyard to put the hammock
- Tips on how to store your hammock during the winter
- A way to order the hammock
- Shipping information and options
- Special offers and discounts
- Links to pages where you can buy other backyard furniture, such as outdoor chairs and tables
- Info on ice/drink caddies to go with your hammock and a way to buy one or more of them
- Links to pages on yard care, landscaping, backyard design, etc.
These are just examples, and you probably should develop a deeper list. It’s a great idea to start by brainstorming this internally with your team before going to the next few steps.
3. Talk to customers and validate your need profile. Now that you have generated some ideas, talk to potential customers. If you have a brick-and-mortar location, ask people in your store. If not, reach out to people who have become customers, and ask them if they would be willing to provide some input. You can also ask a random sample of people on your website or simply put a non-intrusive survey on the side rail.
You can also use a service like UserTesting.com to test the pages of your site. Or you can set up and run a survey through Survey Monkey to collect data. These are all great options to get a ton of info about what people would like to see on specific pages on your website.
4. Use TF-IDF analysis to analyze the top 10 pages for the main search term. One very important concept in information retrieval theory is Term Frequency and Inverse Document Frequency (TF-IDF for short). Essentially, what this does is measure two things:
- What terms and phrases your page emphasizes
- Which of those terms define your page’s unique aspects
We can use a form of this analysis to look at the pages Google ranks in the top 10 for a search query, to see what terms related to your main focus (in my example, “hammocks”) Google considers most important.
The way this starts is by looking at those top 10 pages and seeing what their pages appear to emphasize. Does every page in the top 10 offer spare hammock hooks? If yes, then you probably should, too. Do seven out of 10 offer drink caddies? Might be a good idea to do that, as well.
These are hints that these are seen as providing users a better chance of completing their entire task.
Your analysis should not stop there. You should look for opportunities for uniqueness, too. Differentiation is highly valuable, particularly when it involves an important related need for that hammock shopper.
Perhaps only one of the top 10 ranking pages offers any related articles designed to answer user questions, or maybe none of them do. That could represent a big opportunity for you.
5. Invest in usability and design. Usability, design, page layout, calls to action and so on are all important parts of the process.
If your page has everything the user is looking for, but they don’t like the page, don’t see what they are looking for even though it’s there, or if they find it hard to use, they could be out of there in a heartbeat. All your hard-won knowledge from your research efforts will be for naught.
Similar to what we did in trying to identify user needs, you will want to get user input to your page design. Many of the same solutions that we talked about in step 3 should go into this stage of the process, as well.
6. Implement what you have learned on your page, and measure engagement. Next up, take all of your ideas and integrate them into your page.
Push the page live and see how users respond. What do people end up doing on the page? How long do they stay there? How many convert? How many pages do they consume on your site? What’s the bounce rate? These are all good things to track.
Your Web analytics software can provide you a lot of this data, but consider doing more advanced things such as mouse tracking to see where a user’s mouse typically goes on the pages of your site. You can use this information to make decisions on how to change and better optimize your page.
7. Test page performance using A/B split testing. Last, but not least, run A/B tests of various page options on your site. A/B split testing is its own science, but undoubtedly you had to make many arbitrary choices when putting your page together. Now it’s time to consider alternative layouts, designs, content or cross-linking approaches on the page.
Create multiple different versions of the page, and see which versions provide the best results for user engagement and resulting revenue. Be careful to not tilt the scale too far in the direction of revenue optimization at the expense of overall engagement.
What I’ve described above can be a very involved process. But in a highly competitive world, gaining extra advantages with the quality of the key pages of your site helps us with our reputation, our overall visibility, user engagement, and yes, SEO.
This part of your website strategy can no longer be ignored. Doing this well has become essential.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.