Google has integrated a new way of querying data in GA4 where you simply type in a question or phrase, and it creates a dashboard.
Imagine if you could do the same thing with your SEO data – what would that do for your productivity?
In this article, you will learn how to configure your own dashboards with statements and questions, and create your own answer box using GPT-3 Codex and Streamlit.
Google Answer Box
In Google Analytics v4, you may have noticed a smart search bar that allows you to get:
- Instant answers.
- Configuration answers.
Prompt answers are especially helpful. By asking questions about your data, you get answers and – most importantly – ready-to-use reports.
There is nothing magical about it. This technology is based on Natural Language Processing (NLP), so you need to be precise about metrics, dimensions, and timing when asking for an answer.
For example, you can search for [conversions last week from the United States] And see the results in the search panel that appears on the left.
This new way of using a data visualization tool is incredibly powerful and will certainly be integrated into all solutions of this type.
The time saving for the user is impressive, as users do not have to dig through all the views of the tool and no longer have to configure the view.
Everything is done automatically based on the instructions given in the search box.
Can we easily do the same? What data should we use?
Smart SEO Dashboard
Before ordering a report, you need to think about the important data that needs to be taken into account.
I suggest you look into the concept of Smart SEO Dashboard.
- The first condition is Keep infographics simple and to the point. Less is always more.
- Next, file The coordinates or coordinates must refer to measurable data. Otherwise, it is impossible to see the evolution.
- In addition, Graphs should focus on meaningful parameters. It is useless to monitor parameters that will not affect your activity. Weather is an excellent example: it plays an important role in some locations and no role in others.
- Dashboards should always include relevant summaries To be quickly read and understood. In general, if it takes more than three seconds to understand the dashboard, it can definitely be improved.
- finally, The most important data is time. It is necessary to keep track of time data by comparing each day, month, year, etc.
Now, you need to identify the best technologies for creating this type of dashboard.
Codex Alimentarius GPT-3
GPT-3 Codex is a computer code generator created in August 2021.
Granted access to GPT-3 script is much faster than access to GPT-3.
Not surprisingly, GPT-3 Codex has been fed millions of high quality source codes available on GitHub – that’s over 54 million Github repositories.
Like GPT-3, it is a sophisticated neural network capable of self-learning.
Let’s take a closer look at this volume-to-cost ratio.
OpenAI experiments show that the ratio of volume versus performance in Codex follows a logarithmic scale.
This means that performance gains decrease gradually as the model size increases.
Therefore, the additional costs of collecting data, training, and running a larger model are not at all worth the small increase in performance.
All of these reasons explain why the model had only 12 billion parameters in its first version.
We have found an artificial intelligence to generate the code.
Now let’s look for the best framework available at the moment to do it all in a friendly interface with clicks and drag and drop.
Streamlined It is an open source technology that allows you to quickly build very advanced user interfaces.
Streamlit also includes several very useful components to get more interactions like:
- Session management.
- Password management.
- User Management.
Society Particularly active and involved in many dedicated SEO friendly modules.
To get started, we’ll use GPT-3 Codex to generate graphs with Streamlit, and then try to produce a Streamlit application that will automatically generate and run the code.
Two excellent examples of querying data
First, we need to create and run an app for Streamlit.
1. With OpenAI (semi-automatic)
The first thing to create is a Streamlit web application that retrieves all logs from May 1995 from NASA and displays the number of URLs crawled per day.
First of all, we need to recover the CSV file by specifying the columns name and formula if needed.
For example, it is important that the date is in UTC format.
Then you can ask OpenAI to display the graph of your choice, once it understands your data.
From this instructable, you will have a working code.
Remember, we don’t want to copy and paste code, but rather run everything through English instructions using a no-code approach.
2. With Streamlit (fully automatic)
Here is an open source example Based on a Streamlit app.
It is an application directly connected to GPT-3 Codex that generates computer code and allows you to execute it.
With Charlie WarnierWe did the same thing but for SEO use cases in an app called Codex for SEO.
With one click, you can import your data.
Next, you can describe the content of the imported file: What are the columns? What are the data types?
Then specify your instructions.
In our example, we ask it to group queries together and collect clicks and impressions.
We’ll tell it to keep only the columns we’re interested in (file Queries And Clicks Columns), and then click File Execute button.
No line of code needed to get results, everything is generated by OpenAI Codex and implemented by Streamlit.
So our proof of concept is validated by many different use cases.
Moreover, if you need help, everything you need is accessible through a training program with 150 minutes of video.
For educational and transparency reasons, we have provided the generated code as well as the results.
And with that, the SEO AnswerBox is there for everyone to create!
- How to set up Google Analytics 4 for success
- How to automatically generate quality FAQs and page layouts using Python
- Does Google Analytics use a search ranking factor?
Featured image: NicoElNino/Shutterstock