A: Hey Sumo-ling thanks for the questions!

I apologize for all the information but I just want you to have the necessary knowledge to make the right decision.

1 - Topic modeling is key to helping you write like an expert. At MarketMuse we value data quality over quantity.

Here’s a post I wrote on how MarketMuse topic modelling works
https://blog.marketmuse.com/how-marketmuse-identifies-topics-that-make-a-page-more-comprehensive/

It’s long so here’s the TLDR. Our algorithmic platform is a combination of:
* Bayesian statistical methods (a collection of algorithms that measure, for example, co-occurrence). Our methods are generally patterned from Latent Dirichlet Allocation which was invented in 2003 and is substantially different from Latent Semantic Indexing from the ’80s and TF-IDF from the ’70s.

* Natural language processing that measures, for example, the relationships between concepts in the English language and their specificity. For example – “dog” can be a pet, is a type of animal, has legs, etc.

* Graph analysis that looks at content as a collection of edges and vertices, in one document and across a collection of documents

* Deep learning – neural networks that look to learn and to understand documents similarly to how the human brain processes them

My understanding of SurferSEO, based on what I’ve read is that it’s a TF-IDF tool. Here’s some information about the drawbacks to this approach.

https://blog.marketmuse.com/why-tf-idf-doesnt-solve-your-content-and-seo-problem-but-feels-like-it-does/
https://blog.marketmuse.com/tf-idf-for-seo-faqs/
https://blog.marketmuse.com/why-your-content-optimization-is-failing/

Regarding LSI keywords. They don’t exist (read that first post).
Regarding NLP keywords. They don’t exist either. Maybe you’re referring to the NLP add-on that Surfer-SEO has tacked on top of their product at an additional cost?

From what I’ve read that still does not materially impact their topic modeling which is based on TF-IDF.

Page Optimizer Pro relies on latent semantic analysis. Here are the flaws in that method: https://blog.marketmuse.com/latent-semantic-indexing-seo-content-optimization/

2 - We analyze thousands of pages of documents to create a robust topic model, not just the top 20 Google results.

PoP and SurferSEO don’t do this - when they present bad data, their solution is to provide you with the option to remove it. Our data science teams’ solution to bad data is to improve our processes so it doesn’t happen.

3 - The key to writing like an expert is understanding what topics to cover, what questions to answer, and what internal links, external links and anchor text to use. That’s what we provide.

4 - As stated elsewhere, LSI and NLP keywords don’t exist. The links above go into more detail.