In my travels in marketing land, I’ve recently turned to tuning the web site to attempt to pick up some traffic via search engines.
Its funny how I thought traditional SEO rules just wouldn't apply to me because my app wasn't a web app. The traffic for my site currently comes from:
- Atlassian Marketplace
- Any other pages linking back to my site
You still want people to find your site and hence your app. Optimise all avenues.
The Bee site is built using Pelican which is a static site generator written in Python. All my pages are Jinja2 templates. This lets me reuse large portions of my pages and for a while, that's what I was doing.
I had three pages - one for each service that Bee supports (JIRA, GitHub, FogBugz). The problem was that they all shared the same common template and content. The only difference between the pages was a website comparison video placed up top, which was specific to each service.
As a result, these three pages (the most important pages to rank for due to their specificity), were not being ranked very highly on Google. Turns out search engines don't like duplicate content and heavily penalise your site for it.
The first step was for me to make each product page different with unique content. I also added a fourth product page which has specific content for JIRA Agile which although is part of JIRA, has its own behaviour that I felt could be promoted more.
This part was tedious because it meant rewriting all the copy to be specific to each service. In the end, I kept the same structure for each of the four service pages, but rewrote all the content and used different screenshots to illustrate each service's strengths. I automated a lot of this by just having a Python dict inside of each page which replaced content in-place on the main template.
I've also exposed the release notes page to search engines and I've noticed that it occasionally appears in search results. It's full of keywords so this should have been done earlier.
There is some truth to the old statement that gets paraded around whenever a developer talks about SEO: "just have great content". However just having great content isn't enough, you need to help search engines see it.
Apart from duplicate content, my other issue was that my pages weren't tagged up properly. I was using
h2 elements for their font size and style, rather than their semantic value. Search engines rely on elements to attribute weight to certain keywords in the page.
Its also advisable to only have one
h1 tag on the page and to use
h3 elements in a hierachical fasion.
Links which are in bold are also given more weight so each link to a service page is wrapped in
And although its remarkably straighforward now in hindsight, this was the realisation to me that SEO isn't just about having great content. Search engines are still computer programs that need as much guidance as they can get. Understanding how search engines crawl pages and how they interpret the semantic structure of your pages was eye-opening.
Meta description tag
Although meta tags are no longer used seriously for SEO juice, they are still heplful for defining the little snippet of text that gets seen by users when they search for a keyword related to your site. I updated each service's page to be more specific to each service and to provide a short summary of what Bee is.
My URLs did not contain the keywords I was aiming for. So I remedied the situation. I now have 'jira-client.html', 'jira-agile-client.html', 'fogbugz-client.html' and 'github-issues-client.html'.
Guess what my keywords are ;)
I also put in some
.htaccess rules to redirect current traffic going to my old pages to the new pages with a 301 which transfers the link juice.
My targeted keywords were not appearing in my images either which is also something that could be further optimised. Placing keywords inside the image's
alt tag is also a great way for your images to show up in image searches.
Targeted keywords should also of course be used within the rest of the page's content, however there is such a thing as too much repetition - also known as "keyword stuffing". There's a limit on repeating keywords for a page and keywords within image tags also count so be careful or you could be penalised.
My site doesn't get a whole lot of traffic but I can see a marked improvement in the amount of traffic coming from Google ever since I started down this route a month ago. Here's the traffic increase from Google within the last month: