DIRECT MEDIA United Solutions
By: Darko Jovančić, Digital Group Account Manager at DIRECT MEDIA United Solutions
In the information age we live in, it’s next to inconceivable to build a website without using analytics to collect and process all the data about users visiting the website.
As the name suggests, analytics is used for collecting and processing user data, providing insight into the results achieved, analyzing the target audience visiting the website, and so forth. I say “and so forth” deliberately because there are enough options to make your head spin. As such, website analytics is constantly being improved and perfected, because of the ever-increasing competition frantically fighting for its piece of the pie. All these strides are largely thanks to the development of artificial intelligence, which permeates virtually every inch of our lives. Its development has contributed to things evolving many times faster than what people are capable of with all their abilities.
Already after reading the introduction, you can surmise that you’ll be coming across the name Google much more than any other word in this piece.
Google’s baby is the well-known Google Analytics. And if the available data are an indicator, Google Analytics’ global market share stands at some 75%. A staggering 75% — three quarters of the global total! It’s quite amazing that at the time of present-day technology expansion, this huge superiority isn’t dwindling down. They are 5.5 times bigger than Facebook, which takes second place in this arena. And as many as nine times compared with the Russian Yandex!
To put it plainly, they were the most proactive at the right time. They were two (big) steps ahead of others, and by the time other companies realized the potential, functionality, and monetary value of data, it was already too late, and Google was so superior at that point that the others stood no chance. And it was the first to recognize itself in the new mantra, “The world’s most valuable resource is no longer oil, but data.” It should come as no surprise then that the most valuable companies in the world are tech giants that stockpile information about billions of users that they monetize and exploit in different ways.
After all, the Google team should be congratulated on their excellent idea and decision — it’s little known that Google Analytics was virtually created by Google’s acquisition of Urchin Software. It was the first step towards the monopoly position they now maintain, because that’s how they got in the game during the biggest expansion of the internet around the world. They offered users a free tool and sufficient amount of needed information. With that amount of data, you were able analyze a website in detail, as well as the channels through which users came to your site. You had general information about the users and the amount of interactions and conversions taking place on the website. All this for free. Truly excellent!
Although, all this sounds too idyllic to be entirely true. Well yes, there is another side to the story where Google Analytics relies on sampling. When analyzing your website, Google doesn’t take into account every single visit that takes place but only some of them. And that analyzed sample is used to paint the whole picture. And yet, it remains the tool that gives you the best value for money. Of course, there are plenty other website analytics tools, both paid and free-of-charge, but in the end it all comes down to Google’s tool.
“Half the money I spend on advertising is wasted. The trouble is, I don’t know which half.” The marketing pioneer John Wanamaker’s famous words haunted us for a long time. The advent of online advertising brought about various models and measuring that relatively precisely let us know about which channels, formats, and websites worked well and which failed to meet expectations. Digital is a medium where everything can be measured. The accuracy of the metrics and data that we measure depends on the software’s sophistication and ultimately on the price you’re willing to pay.
One of the models that can be used to gauge the impact of each digital channel, website, or format is the attribution model. What exactly is the attribution model? Practically, it’s a process used to determine — with more or less precision — the value created by each of the digital channels through which a user came to a particular site and performed a specific action or conversion.
There are several different attribution models: Last Interaction, First Interaction, Linear, Time Decay, Position Based, Data-Driven, etc.
In addition, we certainly have the option to create a custom attribution model. This custom model is quite limited in nature and offers little room to reinvent the wheel because after all, the machine can analyze each channel more and better.
The most common model used for attribution is Last Interaction. In short, the model attributes all the credit to the channel or website or format that delivered the last click after which the action on the website took place. And if we look at this model, we’ll see that the dominant channel is direct traffic. Because when users see a product on a banner and click to get more information on the website, after that they most often go to a search engine to get more information and analyze the product. If after that they decide to buy the product, they again type in the company’s URL in the search engine and go directly to the website and make the purchase. In this example, the banner that first conveyed the message and played a very important role will get no credit. Instead, all the credit goes to direct traffic.
On the other hand, the attribution model that provides the most accurate results is the Data-Driven model. This model is one of the rarely used, which is counterintuitive. In essence, the Data-Driven model builds a custom model in line with the data to pinpoint which channel had the highest impact on a conversion/action. This shows that direct traffic has an average of 30% less conversions compared with the Last Interaction attribution model. The model is based on complex algorithms that take into account conversion data to determine the impact of channels in the conversion process. Data-Driven models are characteristic for being unique to every advertiser.
One of Google Analytics’ shortcomings is that it only relies on post-click attribution. And all of the attribution models mentioned here are based on post-click attribution. This means that Google Analytics attributes credit only to those channels/websites/formats that users clicked on to go through to the website. But not to those that conveyed the message to the user or the videos that created awareness, after which the user got further information through a search engine. These awareness formats play a very important role in all this. All of the above excludes online advertising from awareness-oriented campaigns and groups it under performance.
All in all, Google has brilliantly integrated the entire digital technology and closed the whole ecosystem to its advantage. As the saying goes, they have their own walled garden. However, in the midst of the burning issues of online advertising — transparency and viewability — even Google didn’t manage to dodge that bullet. The question is how much exactly can we trust the giants Google and Facebook and the information from their platforms?
When setting up a campaign on YouTube, Google doesn’t allow to set up third-party trackers to monitor campaign results. The only ad server whose trackers you can use is Double Click, which is again owned by Google. You practically have no alternative but to take them at their word.
Facebook has been at the center of global scandals and questionable metrics that have discredited the social media platform. But despite all that, we see even more investments and advertisers’ confidence in those channels. Finally, the question comes up: Are smaller players in the market collateral damage by default?