How I Would Design a Writing Advice App

There are several SaaS apps for writers that teach you how to write better articles. You might be familiar with tools like Yoast and Grammarly, for example. Some of these apps help you write articles that will rank for keywords in Google search, while other apps help you achieve goals such as getting the right tone and brand voice for an article. And after researching ways to write fintech newsletter posts that get more engagement, I may be able to create a similar app that gives you writing advice.

My Engagement Research Project so Far

With the launch of my newsletter, I’ve now studied the effects of five writing-related metrics on engagement. And my engagement models are getting more effective. But they’re still a long way from accurately predicting the number of likes and comments you’ll get on an article. Many factors could still be added to these models to improve their performance, though.

So I plan to continue my research by adding more metrics and seeing if they make the models more accurate. So far, they have. A few months from now I may have a model that uses 20 different writing-related metrics to predict the number of likes and comments a post will get, for example. I use linear regression to study the effectiveness of these models and use R squared scores to see how accurate they are.

For now, I’ve used these models to come up with insights that help me improve my own writing. I’ve also written about these insights in LinkedIn posts and my newsletter. But another way to provide these writing tips would be through an app or a WordPress plugin. Writers and marketers already use several similar services to improve their writing.

How the Writing Advice App Would Work

You might be wondering how I could turn something like this into an app. The idea is that the app would first analyze a collection (corpus) of your posts and generate a regression formula that predicts engagement on your posts. With this formula, the app could give you writing tips similar to the ones Yoast and Grammarly give you.

For example, the app might tell you that you’d get more engagement if you added more personal pronouns to your posts. This is because addressing a message to your reader makes it feel more relevant to them. It might also say that reducing the reading level of your posts would boost engagement. This is because readers are less likely to engage with your posts if they’re hard to read. Your audience might even stop reading them halfway through.

I don’t plan to use a one-size-fits-all model based on results from several different audiences. The audience for fintech newsletters is not the same as the audience for general news publishers or Twitter influencers, for example. So the advice that the app gives you would be based on the performance of your own posts.

As a result, the app’s advice could be more accurate, although it would also require a sample of your posts to function properly. The Shield app on LinkedIn works similarly. That’s an app that analyzes your LinkedIn posts and tells you how to write posts that get more engagement.


It’s unlikely that I’ll release an app like this in the short term. Even if I do manage to come up with a model that generates a high R squared score, I’ll still have to improve my programming skills to be able to turn it into a WordPress plugin or a stand-alone app.

But it does appear that making an app like this is an achievable goal. Similar apps already exist and have paying subscribers. So while my main goal with this research is still to improve my own writing, there is a possibility that I could create an app that gives you writing advice further down the road.

Leave a Reply

Your email address will not be published. Required fields are marked *