How use Google Analytics on media publishers?
If one media publisher hire you to be a Data Analytics, you need understand the specification of this market. Google Analytics is a very good tool to start.
What are different on media publishers?
The biggest difference between media publishers and other website is monetization. On media the most important revenue source are sell ads, with this we don´t have the traditional Funnel to one goal, the goal is increase the amount of ads we can show to user.
Other big difference are the content, the amount of content and their lifetime. Media publishers are a lot of content and is impossible to analyze all pages, we need to group the pages and their life-time is very short, this depend of type of content, if the content are a news, normally the life-time is 2 days or less, if is about famous people or a investigate, can have 7 days of lifetime.
The Platforms where uses see content are different, in most cases have website and/or Apps, this moment media publishers need be website, Apps, Accelerated mobile pages (AMP) and Facebook Instant Articles (FBIA).
Most important metrics
The most important metrics to media publishers are Page views and Visits, this because this 2 metrics are related with ads. In some publishers, we can need one more metrics, the Video Play. We don´t have the Video Plays on Google Analytics but we have one type of metrics that we can use, the events. In this case we need collect when the player start the video to send one event to google analytics with specific category and we use the category to filter this data. We need count Video Plays on publisher that produced video because this are related with ads, the pre-roll.
The analyses that we need to do
The publisher need the big numbers, daily and/or monthly Pageviews, Visits and Video Plays for all pages on project (first difficulty, how to consolidate data of all platforms if Apps have different metrics?). To resolve this I use DataStudio, I collect all data in one Google Analytics propriety and on DataStudio I create new metrics and dimensions to see all data.
The source of visit, one of most important job is analyze the source of visit because this tell us where user came from. Media publishers don´t spend money to increase visits because their revenue is less that the cost and with this we need to understand what free source work better and what type of content work on this source to do more and try understand what have less traffic and if we can increase this.
- Social – We need understand what pages generate more visits and what content work on each page.
- Organic and Google News – We analyze if the content are good for SEO.
- Direct – Is very important source because this is a loyal users
- Email and Notifications – Is a direct communication between publisher and user and we need analyze if we alert user about correct content.
- Group traffic – Very publisher don’t have only 1 product, and have a widget with content from other project.
The content, publisher normally have a lot of content and we need group this content to understand what works better. One form is with page path or we can create a content group. Most time we need sent more dimension that google analytics collect to understand better the user behavior.
- Type of page
- Video Information
The properties and views
My approach is have one property for each project and views by platform or specific content. Is very important have one view with all data to can see consolidate number. First we create one view for each platform, website, APP, AMP and FBIA, with this we can create alert to see if have problem in one platform. After this we need understand if the newspaper have a subdomain, folder or partner that content are different and we need separate to analyze individual but on big number are included. One problem on this is if the project are a lot of website with different content, in this case we need understand the historical evolution, for exempla if we have only 5 active subdomain with different content but after a few months they change to other but the target audience is similar we create only one view for each type of target audience.