In the world of search engine optimisation, there are a wealth of tools which produce analytics and SEO reports after changes have been made to your site. But what if you could predict how your site's ranking would change before you’d made any alterations?
In the world of search engine optimisation, there are a wealth of tools which produce analytics and SEO reports after changes have been made to your site.
But what if you could predict how your site’s ranking would change before you’d made any alterations –and see the impact on traffic, ROI and more?
Sounds too good to be true, right? But ALPS, a new platform from iQuanti, sets out to do exactly that.
ALPS, which stands for Analytics Led Platform for Search, is a “best-in-class analytically driven platform”, in the words of CEO Sastry Rachakonda. The platform “allows you to set your SEO strategy with a deep understanding of your competition as well as your business.”
The aim of the platform’s predictive capabilities is to let SEOs understand exactly what impact a certain change to their site would have, before they invest the money and time in making it. But it also gives an in-depth insight into exactly what competitor websites are doing with their SEO, allowing users to adapt their strategy accordingly.
So how does this tool actually work, and how accurate is it really? I asked CEO Sastry Rachakonda for some insight.
As you might imagine, a platform like ALPS is built on an in-depth knowledge of SEO, a lot of data, and a lot of research.
“I used to be a marketer in large Fortune 500 companies,” explains Rachakonda. “Having looked at the SEO space from the other side, I found there were a lot of gaps in the existing tools, and that was pretty much the genesis of ALPS.”
Existing SEO platforms have a good level of analytics and reporting, says Rachakonda, but as of yet, nothing predictive.
Building a tool like ALPS practically required Rachakonda and his team to build their own search engine – or at least to understand how the theory behind them works. They plumbed the industry research and patents available – including a number filed by Google – in order to understand the factors that go into making a search engine.
“At the core of ALPS is a desire to get a deeper understanding of how the algorithms work,” explains Rachakonda.
Image by iQuanti
Using this knowledge, they were able to build a model which could simulate how a search engine would respond to various changes on a website, and alter the site ranking accordingly.
The ALPS tool uses 105 different factors to model search rankings and predict SEO. While this might sound pretty complex, Google is rumoured to use between 150 and 180. Of course, Google has a lot of internal data at its fingertips which outside parties could never hope to replicate, much of it accumulated over decades of learning and tweaking. But iQuanti did its best with the information that was available, and while some of it was purchased, a surprising amount is publicly available for anyone to use.
ALPS aims to replicate Google’s search algorithms as closely as possible, but it works for other search engines as well.
“We looked at Google primarily because that has the most volume, but the variables remain the same,” says Rachakonda. “There isn’t a dramatic difference between search engines. In our roadmap, we are looking at tweaking it to come up with a secondary model that will more accurately replicate Bing’s search engine ranking.”
While the platform obviously can’t match Google one hundred percent, it comes pretty close, says Rachakonda – and it’s the most extensively-researched and modelled tool of its kind. “Is it perfect? No, but I would say this is the most far-reaching effort in that direction, and we have been successful in driving results.”
At the core of ALPS is its scoring engine: the higher your score, the better your SEO. The ‘ALPS score’ is made up of four components: on-page, off-page, social and technical SEO. The platform also gives you your Google search ranking for a particular keyword – users can choose the keywords they want to target when they onboard with the platform.
You can then compare your SEO score in various areas with competitors who rank above and below you for the same keyword, see what they’re doing better than you (such as having better on-page SEO), and use the tool’s predictive function to forecast how altering different parts of your site will affect your score.
Of course, SEO nowadays isn’t just a keyword game, and a lot of the factors that are now key to SEO rankings are more subjective and difficult to quantify – like content quality. So how does ALPS account for changes to something like the quality of your site’s content?
Ajay Rama, Senior Vice President of Product at iQuanti, explains,
“There are two aspects to content quality that we look at: A, if the page is relevant and meeting the primary purpose it was meant to serve; and B, whether the content is from an authoritative or trustworthy source.
“Our algorithm analyses the purpose by looking at the mix of terms that are being used and not just exact word combinations. It looks at synonyms and topically similar words. It also looks at whether the links that the site is getting are provided in the same context as the page content, and then assigns a relevancy score to it.
“To determine the trustworthiness of a site, we look at the nature of links that the content has, and whether they are from trusted sources or domains.”
ALPS also has a dedicated section for mobile SEO, which looks at how pages and keywords rank differently in mobile search compared to desktop.
Another feature that many SEOs would find handy is its ability to account for Google penalties for something like failing to nofollow ‘freebie’ links by bloggers. So you can simulate the impact of disavowing various links on your site, and then watch your ranking respond accordingly.
Why has there been so little development in the predictive SEO space? | Image by nvodicka, available via CC0
The ability to simulate how changes to your SEO will affect your ranking before you make them is obviously incredibly handy in the search industry. So why aren’t more companies doing this?
I asked Rachakonda why he thinks there has been so little development in the realm of predictive SEO.
“It requires a combination of strong, data-driven folks, engineering, as well as strong marketers – typically, a lot of tools come from very strong engineering companies, but I think that there is a strong overlay of marketing and data science that you need [for SEO],” he replied.
“SEO is a bit of an art. A lot of times, the investment in paid search is much more than in SEO, because of how predictable paid search is. And we hope that we can transform the industry with this tool, by making SEO a lot more predictable and results-driven.
“Could others do it? Obviously – but this is the first, and I wouldn’t be surprised if there are followers. This is a space that is really ripe for innovation, and for really making data work a lot more. This is one of those corners of digital marketing that is still very much an art, not as much a science, and hopefully this tool will take a big step towards making this a lot more of a science.”
Image by iQuanti