📑How to source & manage your data
Last updated
Last updated
Welcome to the fifth lesson in our pSEO course!
In this lesson, we’ll cover one of the most important elements of your programmatic SEO strategy: Data.
Having a purpose is key, your content structure has to be properly thought of, and your head term should be viable. But, at the end of the day, your pSEO collection will be as good as your data.
There are several paths you could take, depending on your collection’s topic and purpose.
You could:
Use an existing dataset
Create the content automatically
Leverage an existing data set
Produce your own content, manually
As we mentioned earlier, sometimes programmatic SEO serves as a strategy to put already existing datasets to use.
For instance, some of the best programmatic collections are the product of a user-generated dataset. These collections provide value by curating and organizing user-generated content, or data resulting from users’ activity on the platform.
A good example is ZipRecruiters’ Salaries collection. These simple pages aggregate data from job posts on the platform (plus third parties), to provide visual data on salary ranges for precise roles and locations.
These pages may appear simple at first sight. But they do their job. And thanks to them, ZipRecruiter ranks in the top 3 for over 84k keywords.
If you need to create all (or some) of your content from scratch, you may be tempted to automate this process. For instance, our WordPress plugins directory could use some automation. But ChatGPT-ing the whole dataset could have some unintended consequences, such as:
Low-value, generic content
Inaccurate data
Remember that GPT is a predictive language model. So, it predicts the most common answers to your prompts. Consequently, GPT is great for understanding what’s the most common perspective regarding certain topics. But fails to create fresh material.
With that in mind, we recommend using GPT to:
Generate accessory portions of your content
Extract specific information from a text (Remember what we did in the lesson on How to choose a head term?)
There’s an exception. You can create good content with GPT by feeding it your source material and using the model to remix it into usable content. However, human oversight is recommended.
Another way to populate your pSEO collection is to leverage an external data set, either by downloading it and tweaking it, or through an API.
This strategy isn’t bad. But remember that your focus should be on offering unique value.
You could:
Re-package existing data so it’s easier to browse and consume
Compile data to give context and authority to your original content
In some cases, your programmatic collection will require manual research and writing. This may produce particularly useful content - especially in overcrowded industries, where thin content and poorly curated resources abound.
However, it’s worth considering that this option is particularly expensive and time-consuming.
Integrate your data set to the spreadsheet you prepared
Test your content template! Choose any page in your collection and see if your data properly fits into the text you prepared. If not, make the necessary changes.
Once you’re sure that your data and your content template fit, it’s time to work on your design. This is probably one of the simplest steps in your pSEO project.