When Google initially burst onto the internet scene, it was a search engine with a simple purpose: Try to figure out what someone is trying to look for, and connect them to the result or website that might have relevant information. Over the years, with more and more people accessing search engines, the “Try to figure out” part has gotten exponentially smarter.
Engines now try to set the context for your searches based on your browsing history, your geographical location, and multiple other factors. This is why search trends and the processing mechanism have progressively moved towards the Natural Language System, one of the most efficient language processing strategies. The system itself is pretty simple: Look up what you want exactly the way you think about it. If, for instance, you need a car for a weekend getaway, you can now search for “Where can I rent a car for the weekend?” instead of “Weekend car rental places”.
The process of making search engines more linguistic has two parts to it:
1. Making search queries more conversational.
One of the easiest ways to go about it is to rely more on voice searches, auto-suggest related questions, and so on — things that Google, for instance, has already implemented. This also means focusing on mobile-based searches, and therefore the way search engines now crawl through and index websites, takes that into consideration. Google already introduced a mobile-first indexing approach (where any website’s mobile version is used to rank its content), making it extremely important that businesses don’t treat the two versions of their website differently.
Using high-quality images, the same structured data, and providing all the content of the desktop version in the mobile website are just a few of the best ways to make your site rank higher. This brings us to the second part:
2. Making websites more conversational
This affects websites themselves in the order of ranking, too. Content is now drifting away from plugging in keywords and towards writing in simple, reader-friendly, jargon-free terms that are presented in a way that’s easy to consume. Instead of paragraphs about what a passport is and how to get one, an article would do better to rank higher by just jotting down crisp, step-by-step points instructing the reader what to do.
The key to making this experience as “human” as possible is understanding the intent behind a search. Looking up “Ginger” would give you entirely different results depending on whether your previous search was “list of vegetables”, “types of cats”, or “grammar apps”. And the way Google understands the context for each search is with BERT: Bidirectional Encoder Representations from Transformers. (Thanks, Google. We’ll stick to BERT.)
Before BERT, Google tried to identify what might be the most important term of the query and load results based on that understanding. Now, it tries to understand the entire search term before providing results. Say, you search for “I need a crane for constructing a building”. Google reads the search both ways (bidirectional), and decides based on “constructing a building” and “I need a” that the crane referred to isn’t the bird, and is the heavy equipment instead.
So with all this in mind, how do you optimize your website and make it more visible in search results?
- Understand what a potential user might search for. Sculpt a bunch of user personas and contextualize what they might be looking for.
- Write content directly for these personas: Use a conversational tone, a set style that appeals to them, gauge their level of knowledge and keep things accordingly simple. This helps in predicting the way in which your prospective readers might search for terms, and therefore bring them to your website.
- Keep the content short and simple. Emphasize only one topic for each landing page. Answer one question per paragraph. Include subheadings that trail from the title and stay on course. This not only makes it easier to read (minimizing the chances that a visitor would bounce off the webpage), but also helps your content rank better in search engines.
- Check your content thoroughly. Make sure there are no typos or factual errors. A typo might throw BERT off in identifying it as a conversational term (“Is search egnine optimization an actual thing?”), while providing incorrect information might throw off its ability to understand the intent. With Google starting to provide fact-check status and information in search results, the last thing your webpage would need is a negative fact-check verdict in fine print underneath.
- Understand that all search engines really want to do is make asking questions easier, and provide better answers. NLP or BERT or mobile-first indexing isn’t here to particularly punish your website. It’s just more efficient in the long run for everyone involved.
If you’re looking for an experienced, knowledgeable team to handle your organization’s SEO, we’re right here! With over 2000 keywords listed on the very first page of Google results, we have a track record of doing a great job. Write to us now!