GPWA Times Magazine - Issue 41 - July 2018

Recently I have been working on some SEO projects, and I thought I would share some of my insights based on what I’ve actually been doing. Here’s a summary: • Good links give a website an initial ranking “bump.” • A good link is from a ranking site. • Ultimately, rankings come down to how satisfying the content is. Before I get into the article, bear inmind that SEO is very holistic. In other words, it’s hard to create an absolute rule that covers everything – that’s why there’s always so much deliberation within SEO. I may make an assertion about something, but invariably there will be some kind of anomalous example that somebody could cite. However, I will do my best to keep things on point. To start, let’s deal with the part that says “long live links.” Then I’ll go into the part where I believe “links are dead.” Google always believed a link should be a “vote,” and then people like us came along and hijacked the voting system in our favor. Google realized that links were a huge fail, and has been attempting to fix the link spam problem ever since the page rank algorithm was first thought up. Google is looking for a differentiating signal so it canmake decisions on whether it should rank that website or leave it in the abyss. In my experience, for a fresh domain, the best signal Google can work with (apart from solid on-site quality/index-ability) is links! WHAT’S A GOOD LINK? Majestic, Ahrefs, Moz and all those similar services have no chance of accurately replicating Google’s evaluation of links. Why? Firstly, artificial intelligence will just find its own course, just like a human would evolve a set of opinions. And secondly, the sheer scale of the internet means that unless you have vast infrastructures, you can’t analyze it like Google can. To give you some perspective on the scale of Google, it is es- timated to have over 1 million servers and use about the same amount of power as the whole city of San Francisco, with its population of around 900,000 people. When we think of the causes and effects of links, we know links help rankings (in some circumstances) but we don’t know exactly which links help rankings. We rely on link metric tools, or we put forth our own opinions on what links are best. Here’s how I evaluate links. I start off with a simple premise: “Google trusts the websites it ranks.” If Google didn’t trust the websites it ranked, it would have a serious problem. If a link is a vote, then a vote is only worth something if it is trusted. I look for links from ranking websites, because I believe Google trusts those sites. I then look for the “power” of a website, which I think comes from its rankings. If I had to choose one core metric to mea- sure rankings, it would be the commercial value of a website’s rankings. But if I’m placing links, I’mOK with “traffic volumes” and “proportion of traffic from search” metrics. WHAT TOOLS DO I USE FOR EVALUATING LINKS? AND HOW DO I USE THEM? It’s iGaming, therefore, I buy links. I get offered lists of domains from various sellers and I run these domains through a tool called Digimetr. The metrics I’m interested in are “traffic for six-month” and “traffic source.” Once I have run these sites through Digimetr, I filter out all domains with no rankings, and then order the sites by vol- SEO SEO is very holistic. In other words, it’s hard to create an absolute rule that covers everything – that’s why there’s always so much deliberation . W W W . G P W A T I M E S . O R G 34