Three years ago, we set out to rebuild our existing Apollo technology with a goal of hunting down and combining every available piece of data relating to the online presence of our customers and putting this information into one single system.
We knew that by doing so, we would save the massive amounts of time we spent piecing this disparate information together manually; and we also know that we would generate new and amazing recombinant insights that would yield unprecedented results for our customers.
In order to collect this data at the scale that we needed, we built a robot – or more accurately, a small army of robots that all have a single purpose – collect, catalog and combine the vast amounts of tiny pieces of information available. We call this group of data gathering robots – Harvesters (aka Harvestor)
An army of Harvesters
Our Harvesters crawl sites, access dozens of 3rd party services such as Moz and SEMRush, search Google and Bing, catalog page details such as content, meta tags, import Google Analytics and Google Search Console data, and access social media networks. They currently generate 10 gigabytes of data each week (and growing).
Once we had this data, we were able to create the impressive dashboards and deep analysis screens that exist today in Apollo. You can overlay impressions with visits and see what parts of your website are working and which relevant topics are not covered by your content. Insight that would be difficult to achieve by using isolated, detached segments of data.
Early last year, we started the next phase of our “robot revolution”. We wanted to move beyond descriptive analytics and build computer algorithms that matched the time-consuming and complicated analysis performed by experienced marketing experts. We wanted to build a diagnostic and prescriptive capability into Apollo that would reduce the cognitive load on our analysts, eliminate human error completely and provide an unrivalled level of thoroughness that just could not be achieved manually – or anywhere.
The diagnostic capability is delivered by two different types of robots that we call – Auditor and Advisor.
Auditor is continuously assessing the state of your online presence (web sites, social media, etc) and digital marketing campaigns overall. It performs hundreds of integrity checks, it verifies the data, it assesses every relevant word and every page. If you have 100,000 pages on your website, Auditor will find the dozen or so that have a problem.
Advisor is our expert system. It is with Advisor that we automate various types of analysis that to date has only ever been performed manually, by experienced practitioners who’s only limitation is time and the variations of being human (sleep, coffee, Mondays, etc).
Everybody understands the importance of assessing and potentially eliminating inbound links (aka backlinks) to a website when it comes to improving SEO. But this process requires expertise in understanding what qualifies as either a positive or a negative backlink and it also requires a lot of time.
For example, if your site has 5,000 backlinks and an experienced SEO specialist spent just one minute assessing the merits of each link, then this would require 5,000 minutes or 83.33 hours or 10.4 days to complete. The reality is that this level of scrutiny would never get done manually – purely because of the scale of the task.
We have an Advisor algorithm, however, that does just that. For every one of those 5,000 backlinks, it looks at the quality of the site, the quality of the page, the number of external links, the size of the page, compares Majestic’s Trust Ratio with AHRefs Domain Authority with Moz’s Spamscore, and a long list of additional checks. It does this tirelessly every single week. It then delivers the dozen or so backlinks that is determined to be suspect so that the SEO specialist can verify and disavow.
Heading towards prescriptive analysis
It is with Advisor that we are delivering on our goal to provide Prescriptive analytics too. A system that not only knows what is happening (descriptive analytics), and knows what the problems are (diagnostic analytics), but also provides advice to our marketing specialists on the best way to resolve the issue or it guides our content team on the content topics that need to be written (prescriptive analytics).
We are leading the way in our industry with our approach to algorithmic analysis via Apollo. Our robots are improving all the time too with machine learning and artificial intelligence projects underway to take these concepts even further.
These robots are working tirelessly for you, our customers – they will not become our “robot overlords”. They augment and enhance our teams of marketing experts by liberating them from the manual labour of data collection and analysis so that they can spend more time actually fixing the issues, or writing the content, and generally excelling at what humans do best – understanding your own customers and finding creative ways to reach them.
Need help using Apollo?
If you’ve not used Apollo yet (or recently) and you’d like some help navigating your way around, drop us an email at firstname.lastname@example.org and we’ll be more than happy to help.