Saturday , April 21 2018 1:49 pm
Breaking News
Home / Analytics / SEO: Rank Indexing & Topic Modelling

SEO: Rank Indexing & Topic Modelling

The easiest way to optimize your SEO campaign is also the most flawed: individual keyword rankings.

Why are keyword rankings fundamentally flawed?

In the ‘good ole days’ of SEO, rankings were the same if you were logged in while in NYC or incognito in London. It seems hard to even imagine now but if you were winning a term, you were simply winning it.

Rankings now are highly personalized and depend on you: your location, search history, preferences, and social media usage. Winning a search term may mean you are #1 for millions or it may mean you’ve visited your own site a few thousand times.

Rank tracking software attempts to reduce these problems by using proxy servers and allowing for local specific results. This works to some extent but in many cases is just as flawed as manual search. Many country-specific versions of Google allow you to set location, which comes with its own set of reporting challenges. For ab enterprise client do you set to the company’s home location or where the company’s customers are?

(Good luck reporting to the Head Office in Sydney, owner lives in NY and they want to rank for “Perth pool chemicals”)

Keyword rankings as a metric also mean finding the best key phrases to track. Do you optimize for “Melbourne wedding photography” or “Melbourne wedding photographers” or “wedding photographers” with the location set to Melbourne? They’re all very similar but return unique results. If you optimize for all, you spread yourself too thin. Don’t optimize enough and you lose potential business.

So how do we optimize our sites without tracking the untrackable or optimizing a small number of keywords?

Rank Indexing: The way forward

A rank index is an aggregate ‘bucket’ of keywords around a specific topic. By tracking a bucket of many keywords you learn quickly whether your business is heading the right way as a whole.

Using our pool cleaners as an example, track a set of key phrases for each audience and compare business results and Google Analytics with each bucket. Assign dollar values to each based on actual revenue. Soon you are optimizing for revenue instead of keywords.

Let’s think about our photographer again. Maybe this time she tracks results for “Melbourne wedding photographers” and “Melbourne wedding photography” as well as “wedding photographers in Melbourne” and 10 similar keywords. By tracking across the whole industry segment, we see the whole picture. With enough resources we could also track and optimize suburb-specific keywords on this level.

So how do you optimize for rank indexing?

Doing on page SEO is often challenging because we limit ourselves to think about keywords. With rank indexing that changes toward a more honest page optimization. You no longer need to rely on a title tag to simply rank a specific keyword. Now you can focus on the content, writing sales copy, not search copy and setting meta tags that make sense to the page, not a specific single word.

The other main on page difference is how you approach topic modelling. A natural and high ranking page contains information about the topic but also it’s associated ideas and terms. You need to be very specific. Take an example:

If I say: “hoop, player, game, win, rebound, shot, court, coach, LeBron and playoffs” then you know I’m talking about the NBA without me specifically mentioning it by name. But why? Which word is the key? Or is it the sum of the whole?

Court and shoot by themselves don’t give you as much do they? It could be someone’s case about an armed robbery. Similarly, game and player could be about your kid’s soccer game or poker. It’s only when you flesh out a topic fully that you understand the whole thing. If you ignore a big part of your topic, you put yourself at risk of being misunderstood.

Let’s try one more example: “park, ride, ticket, window, drinking, roll, upset.”

Do I mean

“I became upset when my Sunday ride went wrong at the park. An officer asked me to roll down the window so he could give me a ticket for drinking.”


“I picked up my ride ticket from the theme park window. My friends and I were drinking and chose to ride a rock n roll rollercoaster, which upset our stomachs.”

Language is vague without context. Learn to be clear and specific. Write in detail and optimize for whole topics over lists of keywords.

About Matt Antonino

Matt is a marketing consultant in Melbourne, Australia. He teaches businesses to market smarter and faster through multichannel marketing including SEO & Content Marketing, SEM, Email & Social Media.

Leave a Reply

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


Share This