Avinash Kaushik is one of the major references in web analytics that exist. He wrote two best-selling books on the subject ( Web Analytics an hour a day and Web Analytics 2.0 ), has a blog where he writes articles dense and complete and even a newsletter which discusses practical issue related to analysis and metrics.
The curriculum would go away but as Avinash himself says, he works at the intersection of Marketing and Analytics, as well as artificial intelligence applied in this context. And that’s what was his lecture.
Avinash started his speech by making a provocation: little has changed in Marketing in recent decades. From the earliest days, the marketers were using demographic and psychographic analysis to make decisions and reach potential customers. In other words, try to find out if these people go to school, have children, they like to read the paper.
To demonstrate his point: Avinash called a person on stage and gave an unusual example, said Bruno (participant who took the stage) like to buy lingerie. He is rich and like to buy dozens of them every month. Going by demographic and psychographic analysis, no company will offer lingerie for Bruno.
The point is, according to what he seeks, which he shares what he read, the sites on which he sails – that is, your intentions and your behavior – you can make the most relevant actions and focused on who really has interest to buy.
Another example is the Facebook ads. Avinash said that Facebook is the company with more data about people who exist in the world. Even if you’re not on Facebook, he knows a lot about you. All this information is available for you to use and target your ads, the problem is that many advertisers do not know how to use and end up delivering irrelevant ads to the wrong people.
If you search for “artificial intelligence” in Google, you will find several images of robots in human form. That’s the scary least and a wrong way of looking at the AI.
In fact, the artificial intelligence is already applied in different contexts with different goals. Avinash brought some examples:
To make it clear from this point, Avinash explained simply these concepts.
Avinash brought an example of application of very interesting neural network. Imagine teaching a machine as a table. There are thousands of different tables and display one by one to the machine is a work almost endless. For this, from a few sample “of what is a table”, the machine understands and starts to identify herself tables ever someone showed her.
No need to go far to start applying artificial intelligence in your business. To show this point, Avinash brought some application examples with tools that already exist.
1) Analytics (GA)
The GA has improved a lot over the past few years, but still a great tool to simply “play” information on you.
However, recently they added a feature of Analytics Intelligence, which analyzes the data and the GA brings relevant insights into the information, making it more actionable.
Also I said something that I found particularly very relevant: how is the distribution of time spent on analysis today and as it should be:
2) Artificial Intelligence in ads on Google
There is a network of hotels in the US Red Roof, which builds the next major airports. They found that a fairly common public in hotels is losing the flight or the flight is canceled. To get an idea of volume, 90,000 people go through this situation every day in the US, which has about 5000 airports.
To be able to reach these people when they need it, they used a combination of several variables, behaviors and intentions, such as weather and delayed flights index, to know when demand would appear.
Then combined this with segmentations throws automatic creative and automation to create automated ads when people pesquisassem for hosting.
The result was a 60% increase in bookings from announcements made “by machines,” without the intervention of a professional ads:
To end the presentation, Avinash left some goals for 2018:
– Have 5% of employees trained in ML
– Having 25% of the most important business processes aided by ML
– Having 50% of the company’s Digital Marketing aided by ML