How we use keywords and search data to improve our content
We have always used keyword data to create our titles and headings, and to mirror the language of our audience. More recently we have started using keywords to validate user needs, inform the scope of user stories and improve our existing content!
This post explains the tools we use to gather keyword and search data, and how we use that data to make decisions.
Tools we use
We use Keywords Everywhere and SemRush magic tool for most of our keyword data.
Auto suggestions from when you start typing a search term into Google can also help.
Google Search Console is good for data on pages we have already published.
SemRush is good for finding what keywords other websites are appearing for.
AnswerThePublic is another tool we can use but we don’t pay for it, so we only get limited access.
We use keywords to…
Check that the problem is something people look for help with online
Some problems cannot be solved by online content.
Before we started using search data to validate user needs, we relied on qualitative evidence of problems people face. During user research for our Disability Energy Support service, we talked to lots of people who found energy bills hard to read.
Because we write content which helps solve problems, we gave it a go and published a page about “How to read your energy bill”.
The page views were low and energy advisers weren’t using it to help our customers. If a bill is hard to read, so is a page about how to read your bill. Using keyword data to validate user needs limits our bias towards solving problems with content.
Sometimes people have a problem, and they want to speak to a person about it, rather than read about it online.
Often people are searching for contact information, for example “ESA contact”.
Sometimes people want to speak to others in their situation. For example, “disabled parents support group”.
These keywords suggest people want to speak to an expert, or someone with similar lived experience. So, we signpost to the helplines, forums or organisations people might find useful.
If there is no search data associated with an information need at all, we don’t write for it.
Find the language of our audience
We had a page called “Phone accessibility features”. We noticed the page views were low and used search console to find out why.
Keyword data showed people searched for “settings” more than “features”. So, we changed the title.
6 months later we checked to see if this change had made a difference, page views had increased!
Keywords also help us understand when to use jargon rather than avoid it. Our audience use “permitted work”, but they don’t often recognise “legacy benefits”. We also test these terms for understanding.
Inform the scope of content
We look at what information people are seeking and add to the scope of a user story if :
· there are high volume keywords showing an information need
and
· we can find qualitative evidence that matches
See the relationship between different information needs
If two terms are often searched together, we reflect this relationship in our content.
For example, we have a planned content item about mental health. We also have planned content on Employment and Support Allowance (ESA) and (Personal Independence Payment (PIP).
Keyword research found that people often search for ESA and mental health together. This showed a potential relationship between two content items. We looked at further evidence to see how they are related.
In this case, qualitative data showed two possible relationships between mental health and ESA:
· People want to know if a mental health condition makes them eligible for disability benefits.
· People find the assessment process for ESA has a negative impact on their mental health.
We have written a page on benefits and mental health which looks at these two needs together.
Improve our pages performance in search engine results
We have a way of using data to find out which content to improve. When a page has low traffic, we use search console to check for:
· Low demand. Low or no search volume for target keywords suggests it’s not something people are looking for.
· High demand but low ranking. People are looking for it, but we are low down on the Google search results page.
· High demand, high ranking and impressions but low clicks. We are appearing well in Google search and on people’s screens, but we aren’t attracting clicks.
This information helps us decide if:
we need to retire content because people are not looking for it
or
we need to change the keywords in our meta description, title or headings. So that Google, and more importantly, people, see us as relevant.
If you know of any other ways we could use keyword data to improve content, please get in touch!
