November 15, 2013
Hummingbird is the biggest iteration of the Google algorithm since Caffeine in 2010. It is, however, going to have less impact immediately on search marketing than the Panda and Penguin updates. Amit Singhal, of Google, said 90% of searches have been affected by this algorithm change. It was interesting to note, however, that very little was said within the industry about affected rankings before the announcement was made.
The details of the Hummingbird amendments are provided below, detailing what has changed and how it impacts search marketing strategies.
In general terms, Hummingbird is a step towards the Google algorithm becoming artificially intelligent. Google’s goal is to be able to serve content based on what the searcher wants. A key performance indicator of this development is surely the number of searches a user makes in a short space of time (the higher number of searches denoting that the content is not relevant to their query) and also the number of pages a user visits per search term.
For example, 8 years ago 30% of searchers would look at search results past page one of Google. Today, it is only 10% and it is likely to decrease as Google improves the indexing of searches against the real intent of the search (thus making it even more significant to increase SEO rankings).
So let’s have a look at Hummingbird in detail. What are the changes and what does this mean for search marketers and website owners.
Semantic search will be discussed more and more in regards to SEO over the coming years, particularly as search devices change, search queries evolve with the technology, and Google and information retrieval technology adapts to changing hardware.
Semantic search is where an algorithm, like Google’s, retrieves information based on the meaning of the search as opposed to matching the content against the search term. Google’s aim is to index all content based on semantics and it will continue to improve in this regard.
Before Hummingbird, Google was indexing content by using pattern match to marry up content within its database against the search query.
Today, Hummingbird is more intelligent, and now indexes content based on the query intention. This includes some key elements such as the context of the search and the searcher’s requirements. For example, the context of the searcher can include variables such as the device a user is searching from, the time the search was made (local) and/or the frequency of similar searches from the same device, allowing Google to understand more about the search habits of that user. With a better understanding of searchers, Google can make better judgements about search intent and improve the content served from the index.
As Google continues to improve its semantic search algorithm, websites will need to continue improving content for end users.
Providing the best content for basic brochure websites may be the hardest since a large majority are relying just on textual content.
Gone are the days when SEOs could focus on keyword density and target keywords within alt tags, meta tags and H tags. More granular research is needed into multiple keyword variations, although this recommendation isn’t a result of the Hummingbird update.
Webpage content breaks down into only three core sections – images, text and video – and a combination of all three determines the theme of the page. It is important that companies, SEO practitioners and in-house marketing teams see that each page has its own theme of which all content types contribute to improving.
For example, if I have a page about “Mailing bags” then I want to give as many signals to Google as possible that this page is highly targeted for all relevant search terms. Therefore, from a contextual point of view, I want to ensure it is optimized for all search variations for this product, including mailing bags, mailing cartons and mailing pouches. I also want to make sure the text content covers important practical information – for example, the uses of the product and application. This is particularly important with the Hummingbird update because Google is specifically looking at how it can serve content better for practical searches beginning with things such as “how do I… ” or “what is the… “. More information on this below.
Secondly, you want to ensure that you have relevant images on the page, for example, maps. 20% of all searches have “local intent” in the UK, and it’s even higher on mobile. So maps are a significantly important element to add to websites.
Another important factor to help serving content against intent is schema markup, specifically for videos.
All of these elements contribute as signals to Google to say “Hey, this page is very relevant to that search term”.
Material Searches and Application Searches
This is an aspect we have been working on with clients for a while and we are not surprised that Google has covered it in the Hummingbird update.
We have worked with clients to optimize their websites which contain multiple bespoke products and have seen that there are clearly two types of searches being made. Material searches are searches for a certain product based on what it is – so for example, a search of “cardboard boxes” is a material search. The searcher is clearly describing the product that they are looking for in the search term. Similarly, a search for “coffee table” is someone looking for a very specific item.
Application searches, on the other hand, are queries made by users who are looking for something or someone to do a required job, but the searchers are not sure exactly what they are looking for. For example, the search phrase “packaging material for moving house” may be used by someone wanting to find packaging solutions to meet their house moving requirements, but the searcher may not know exactly what is needed. The searcher is, in effect, asking Google to provide the answer. This is where Hummingbird is enhancing its semantic understanding of search queries and improving the indexing of content for each user. Google may consider factors like past search history, the device used, and even elements like personalized search with “Search Plus Your World” which is another change similar to Hummingbird that also contributes to better indexing of content based on a user’s needs.
So, in short, focus more on searches to do with the usability and performance of your products and services (applications), rather than material searches which just describe what you do.
Knowledge Graph Tap In
Google is also ranking sites based on information in your Knowledge Graph. This is probably more relevant to larger companies or brands. Do not, however, underestimate the significance that local search and Knowledge Graph are likely to have on impacting rankings.
Use of Mobiles
More and more people are speaking searches into mobile phones which is where Hummingbird determines the meaning of a search. These spoken searches are invariably longer search phrases and mainly contain a question. Google’s search results need to be able to answer these questions so the focus for Hummingbird is to understand the semantics of the search term, be it by location, device and/or search intent of the text query, to serve the most relevant results.
Use of Local Searches on Mobile Devices
40% of searches made on mobile have local intent. Therefore, think about the content on your website and how it appears on mobile. Does it format correctly? Is it easy to use? By this I do not mean is it responsive and built in HTML5 because that is only a temporary solution. Consumers on mobile are in a different buying mode, so content on your desktop website may not be relevant to a consumer on a mobile device.
For example, your store locator links are going to be very important to mobile visitors but, if your store locator links are buried in the footer of your desktop site and they are regurgitated the same way through your responsive design, then this just isn’t good practice.
Google will be making changes in the coming year to give increased weight to mobile optimized content. Hummingbird is, effectively, the first step towards this goal. Google wants to see content created to best serve a user’s “intent”. This includes the device used to perform searches.
Long and Short Tail Keywords
We are going to see Hummingbird have a more significant impact on long tail search terms (longer string search terms where there is a question or element of uncertainty within the searcher’s mind) than on head terms (short tail). Head terms and generics may be affected within certain industries, but greater variations will be seen for longer terms.
This is because of conversational search.
Use of voice search creates longer search query strings and it is important to serve content to match these searches. Think about user guides and how-to guides, if you have products with multiple applications.
Ikea is a great example of a website that has good video content showing how to put together each furniture type. This is great content to index for conversational search.
Some people recommend making text content more conversational. We do not, however, see this making a difference in regards to ranking effectiveness.
Make sure your content is geared towards a user’s intent. Think not only about what is being searched for, but also about what searchers are trying to find based on contingent factors such as location, search device and time of search. Match your content and increase semantic signals to assure better indexing.
Hummingbird is the beginning of some major changes to content indexing. Google will become better at determining user intent and semantic understanding, and with changes in consumer habits across multiple channels – particularly mobile – search marketers will need to adapt as well.
Article by Geoff Ian Parker. We are a leading search engine marketing company for small to medium-sized companies. We provide SEO, PPC and social media marketing for companies across the UK via three different delivery vehicles; managed services, consultancy and training courses.