Internal Site Search Analysis

[BedeFahey]

Hi Folks,

I have about 6,000 internal site search phrases that I want to analyze. There are many variations and duplicates that have similar intent within the data, e.g. Employment, Employment Opportunities, Employment Application.

Does anyone have any thoughts on how I can aggregate the data to get an idea of user intent. There's a lot of long-tail in there.

The data does not come from Google's site search tool. I just have a spreadsheet of the terms and the number of times they were searched.

Cheers!

[CatalystSEM]

I recommend reading through the entire spreadsheet from start to end and color coding every different intent keyphrase that you come across. Any keyphrases which have identical or very close intent matches should be color coded together.

I do this process manually then I write down every intent variation that I came across. It usually ranges from 10 - 30 different intent buckets. Once I have the pieces of the puzzle, I start piecing together the search funnel for the business.

Who are your known demographics (the personas that your marketing team talks about or your target audience)? Are there intent buckets that do not map to these personas? If yes, then you may be attracting a demographic that is unaccounted for. See if you can figure out who this demographic is.

Is it a worthwhile demographic? Pull out your conversion data. See if you can use the master list of intent buckets to break down the search funnel into a flow chart for each demographic by using the keyphrases mentioned in the first paragraph of this response.

I'm assuming that you have some kind of an employment website like Indeed or Monster.

There are different kinds of personas that I can think about off the top of my head. For example:

1. Employed

2. Unemployed

3. College Student

4. Remote Worker

On 4 different documents, you'd write down the keyphrase intent buckets that you know for a fact are associated with each persona mentioned above. You'll want to use conversion data or pages on your website that are built for specific target audience groups to help you make this decision. For example, if you had a page targeted at a college student, then you could analyze the keyphrases that drive traffic to that page to determine whether your intent bucket is grouped correctly with your correct demographic. You could go one step further by looking at the keyphrases that possibly convert into an email subscriber whom you know is a college student with 100% certainty.

Once you have a bunch of intent keyphrases mapped to each demographic, then you would try to break down the steps. Put yourselves in the shoes of a single demographic and make a decision tree. Write down all the possibilities that one would take online in a sequential set of steps and see if you can map your keyphrases to that list.

There's more but this exercise will get you started and help you figure out the different target audiences that come to your website and highlight their search behavior from keyphrase to keyphrase. It will show you the best points to attack them from an online marketing perspective by tailoring your messaging to distinct steps in the funnel.

Please let me know if this was helpful.