[Analytics Intern] How do I build website traffic dashboard with Google Analytics for NGO?
With the coming of autumn in Minnesota, I ushered in my fifth month at UMASH, an NGO under School of Public Health at University of Minnesota. After establishing the report automation for social media including Facebook and Twitter for UMASH, my next task is to look into UMASH website.
As an experienced web analyst back in Taiwan, as far as I know, website is a product that is way sophiscated and challenging than social media, and the metrics for website are different from social media too.
I made a comparison chart for website analytics versus social media analytics.
First of all, the fundamental elements for website and social media is different. Webpage are the basic element for website, while social media, either Facebook or Twitter, is consist of numerous post.
Second, they both can be measured by traffic and engagement metrics. Traffic is the inflow and outflow of user on the platform. Engagement is how user interact with the platform. And for each of website and social media, the traffic and engagement is measured by different metrics.
Last but not least, the expression is diverse as well. The expression here is indicated that if the data can be observed by third person. Say for Facebook fan page, a user can see how many likes, shares and comments on the timeline, but for website, a user will never know how many pageviews are created by other users.
Tools and skills
Tools: I use simple Google Analytics Dashboard function to create dashboard for the organization.
Skills: In this article, I will showcase and walk through the thinking process of how I structure the dashboard building.
Website Setting
UMASH website is relatively simple. It was built by WordPress with a clear structure. And google analytics tracking code is embedded in the WordPress back-end.
Outline
1. Analytics Funnel
2. Metrics selection and Execution
Analytics Funnel
The basic logic here is from macro picture to micro picture.
At the very beginning, I want to show to my boss and other colleagues user’s demographic. UMASH’s goal is to spread health info and make it easy for farmer around mid-west area to get access to handy info. Therefore, it is essential to see our user profile if we are reaching correct target audience.
Second, I try to show the overall performance for the website in both static and trendy manner. For static showcase, it provided a snapshot of how the website perform in a specific of time, while for trend, we can see the change within a period.
Third, I dig deeper into more micro segment: pages. As the minimal elements for a website, each organization has its focus on certain group of pages. For UMASH, our focus is ‘Farm Safety Check Pages’. Therefore, I put some highlight to indicate the performance of these pages.
Metric selection and execution
Owing to the data privacy of organization, I will not show the result of dashboard here. I decide to use some drafts for dashboard and screen shots for the widgets to take a walk-through of my execution.
Dashboard 0: Demographic
For most of the metrics here, I decided to depict the user as a key metrics for demographic dashboard. It shows the unique user id visiting the website within the period of time.
Gender and age is the basic dimension, while for user from different state is specific for UMASH. UMASH is mainly targeting users from mid-west America, total five states. My boss wants to track how it reach geographically, and that becomes the main reason I put map with different color blocks and the bar chart for users from different state at the same time.
Dashboard 1: Important metrics (Static)
This is dashboard is relatively easy. It is the showcase of all the important metrics for a typical website.
It has two functions:
- A snapshot of the period of time, and it grant the reader a sense of data what performance we are reaching at present time.
- It is easy for comparison. Google Analytics has a time window that is perfect for custom period comparison. With a easy and static metrics, the change, growth and decline can be simply shown.
The dashboard is also divided in two parts. The left hand side is traffic metrics, including user, session and page views, all of which are implying the user inflow and outflow of the website.
The right hand side is engagement metrics, including bounce rate, average session duration, average time on page and page/session. All these metrics are super helpful for web master and web designer to check whether the user experience and user interface of the website is good and engaging enough.
Bounce rate is to test if users just bounce without interaction with website; we are always pursuing longer time spent on website, and for each session, more pages are created by the users.
Dashboard 2: Important metrics (trend)
Jump from the static important metrics dashboard, I decided to extend that dashboard and make my third dashboard, which is showing the changes of each metrics.
I executed in two ways. One is using simple table to show the absolute number month over month; second is using bar graph to show the relative change across months.
Furthermore, I put the devices percentage here, both user and session basis. Last but not least, source/medium is also the important dimension for a website, where the traffic came from. I plot the donut graph for the proportion and the trend for two of the biggest traffic source of UMASH website: google organic and Facebook referral.
Dashboard 3: Pages
Finally I came to the final step of the funnel, the micro perspective: pages. Pages is kind of the minimal unit of a website. And its metric is usually page views. I divided my page dashboard into two part.
The first two columns on the left is the overall pages. I listed top of 10 pages in our website, and shown it in portion as well (donut graph), the daily trend of home pages, which give the reader an overall picture of the most frequetly-visited pages.
The last two columns on the right is designed specifically for ‘Farm Safety Check’ pages. With simple filter, I aggregate all Farm Safety Check page together and further show their monthly trend, daily trend, and top 10 in this category, which is very readable for social editor and my manager.
Limitation and future plan
The four dashboards I created covered most of the organization demand for website analyses. However, there are still some limitation there.
- It is the general data automation dashboard for the basic understanding of a website, but for more detailed and complicated user behaviors on website, I definitely need more tools to observe that, for instance, goal, event and segmentation, which I will write another article about all these feature in Google Analytics.
- Google Analytics tracking is majorly based on how the tracking code is embedded in the website. If the method of tracking code is wrong, the quality of dashboard will be affected.
My future plan for Google Analytics at UMASH is that I will try to pull out the data from Google Analytics with its API with Python and conduct more intricate analyses on the raw data. I will share more in the near future.
If you like the article, feel free to give me 5+ claps
If you want to read more articles like this, give me 10+ claps
If you want to read articles with different topics, give me 15+ claps and leave the comment hereThank for the reading
