Without fail, every few months there is yet another breathlessly published blog post or article proclaiming that “SEO is Dead!” Typically this dire proclamation regarding the future of SEO is accompanied by some variation of the following:
- Content is king!
- Social media is the new SEO!
- Link-building is dead!
In reality, SEO will never die. At least not as long as there are search engines. While Google certainly has diminished the amount of real estate for the average search query in the “traditional” organic search result, “natural” search results are critically important for search engines. If, as some have speculated, Google went to a “paid” results-only model, over time I suspect that the average web user would lose trust in the quality of their search results.
However, there are by contrast many exciting developments in the field or search that will likely have a huge impact on SEO in the coming years. They may not make as good of clickbait as “SEO is dead,” but the possibilities are intriguing to say the least. Here are 3 predictions for the future of SEO:
1) Machine Learning & RankBrain
Ok, so in this instance the future is now. Google now serves an increasing percentage of their results based upon conclusions reached by self-learning computers managed through an ambitious artificial intelligence project. In reality, this is not a totally new concept.
Going back to 2013, the Hummingbird algorithm represented a fundamental shift in how Google processes search results. Rather than traditional metrics like keyword density and such, Hummingbird worked to understand the intent of a query, rather than just parsing and matching words from the query itself. For example, a search like “What is the Capital of California?”, the results on page one won’t list that exact query in the page title (as you might have expected several years ago). Instead, the algorithm recognizes the intent of what the question in looking for (the City of Sacramento), and serves up results accordingly.
With RankBrain, the fundamental difference is that the computers used continue to gather and interpret new information over time, and effectively teach themselves how to further interpret user queries over time. This tool has shown to be particularly effective with hard-to-understand or poorly formed search queries. A recent study by Stone Temple showed that of 163 queries where Google’s baseline algorithm did not understand the question, RankBrain improved the results over 54% of the time.
— Mark Traphagen (@marktraphagen) April 28, 2016
Expect machine learning to play a larger and larger role in the future of SEO. As the machines get smarter, it will become virtually impossible to “game” the system from an SEO perspective, as techniques like keyword stuffing, alt text SPAM, and the like will be easily detected by search engine algorithms.
2) Automated Image Recognition
This prediction dovetails off of machine learning, but it’s a component that is not talked about all that much. Outside of Google, there are many other companies exploring the possibilities presented by Artificial Intelligence and machine learning. One is the Dreamcatcher research project currently run by Autodesk. The goal of this generative design project is to enable “designers (to) describe the forces that act on an object and then lets computers go off and make it.”
From an SEO perspective, the most attention-grabbing feature of this project (to me at least), was the super-computer’s ability to automatically generate captions for photos, simply by scanning them. Millions and millions of images were served to the computer as a training set, and from that knowledge it was then able to create the following labels:With search engines able to automatically recognize images and interpret context, the whole concept of image optimization will need a re-tooling. Many fundamentals of image optimization, from naming files to using keyword-rich alt text, have remained as best practices for roughly a decade. More recently, optimizing images for reduced file size, and enabling social sharing (via og: tags etc.) has come to the forefront, but there has not been a dramatic shift.
In the future, with search engines able to understand and accurately process images, using image-based text not be such a bad idea. Finding highly relevant and useful images to serve with content will likely become much more important. For industries heavily invested in a strong presence in the image results of search (think sign companies and others), these changes will be critical to follow closely.
3) Real-time Algorithm Changes & Negative SEO
Remember the big Penguin 4.0 update that was poised to hit late last year? Me too… Even as recently as this week, there was revived speculation about when the long-awaited update will occur.
Thankfully for those of us sick of waiting a year or more to see the fruits of a long and painful link clean-up project come to bear, future updates such as Penguin appear poised to happen in real-time. In a nutshell, rather than reacting to changes by the search engines, then anxiously awaiting the next update to see positive changes, webmasters *should* expect to see much more instantaneous gratification in the future of SEO.
However, this change will likely cut both ways. Though the much-ballyhooed concerns about Negative SEO never really came to fruition following the initial Penguin updates, real-time updates could change that reality. If sudden influx of hundreds or thousands of crappy forum profile links or SPAM directory listings could be processed by Google in real-time, theoretically it would make taking down a competitor a much more feasible option for quickly gaining ground in the organic results.
Negative SEO is about 100x more common than Google pretends like it is. I've seen it at almost every company I've worked for.
— Austen Allred (@AustenAllred) April 26, 2016
^^ Real-time updates will make this much worse!
… So there’s my $.02 (likely worth even less than that) on the near-term future of SEO. What are your thoughts on what Google has in store for us in the coming years? Sound off in the comments.