Share with:


After the popularity of our last case study post, we decided to share more results, especially ones that involve sites not doing particularly well.

Most case studies you see online, particularly in this space, are edge-cases where somebody achieved fantastic results.

It’s rare to see people talking about the more frustrating side of internet marketing.

Today’s case study actually consists of both. It is definitely a success story with positive results, but it starts off with the more frustrating side of things.

Essentially, over the past 3 and 5 months respectively, we’ve taken two sites that were on a major traffic and income decline, audited them, made a lot of onpage changes, and waited patiently for the results to kick in.

Site 1 had been on a decline from the day we bought it in June, to the middle of March. We finished our revamp on this site in December.

Site 2 (below), we started working on for an Onfolio partner in December, finished it in February, and it also started turning around in late March/early April.

We’ve included keyword graphs as well as traffic graphs to show this isn’t just a seasonal traffic spike, and that the sites are both ranking for more keywords now.

Changes We Made

Both sites required a full audit before we actually started changing anything. Both sites also had a ton of problems.

Here’s a brief overview of some of the problems we encountered:

Site One

When we acquired the site, we were quite happy with the quality of the content as the previous author was an actual expert in the field, so we were confident we had the right foundation to grow the site.

However, as the site presented quite a few problems with a very outdated theme, poor mobile speed, and had serious cannibalization issues, both ad monetisation and scaling content were very difficult.

Also, the site entered a period of decline quite soon after we acquired it, so we bit the bullet and decided to shoot for an overhaul instead of going for patching leaks.

Here is how the traffic declined over time shortly after we took over the site:

Overview of steps we took to fix it:

  • Organise all existing content on a spreadsheet and decide what stays, what needs merging and what can be deleted from the site.
  • Plan out a siloed, tree-like architecture envisioning a site several times bigger than the one we had at the time, so we’re counting for potential expansion and building the grounds for it.
  • Set up an staging environment and start from a blank wordpress installation. 
  • Deploy the lightest possible setup for proper mobile performance. GeneratePress as a theme, with fewer than 20 plugins.
  • Design core “best xyz” page templates for a mobile optimised experience, including custom, clean looking responsible comparison tables
  • Design core single product review templates to standardize the user experience across multiple reviews and offer the best possible mobile experience.
  • Put all the content on the new blank site and test thoroughly. Added a few pieces of content to sections of the site where we saw we were missing sufficient topical relevance.
  • Added schema to all pages, as well as new rich media. 
  • Push to live, recrawl the whole site, clean any errors on GSC.
  • We also conducted a thorough backlink profile cleanup, and disavowed all spammy or otherwise risky links
  • Built a handful of white hat links to replace any lost link juice and channel some new authority to the site.
  • Wait. This is certainly the hardest part, as the decline continued even after pushing everything to live, but on the latest quality update the rankings and traffic have shot up, proving that sometimes all it takes is to go back to basics and ensure a site meets all the best practices and has a proper architecture in place. No hacks no tricks, no expensive offpage seo done.

Site 2

Our partner approached us with a CPA site in the medical space that they purchased off a broker in early 2018, which was losing a bit of traffic but still pulling a healthy income. The site then got hit in the Aug 1st update losing half off its traffic, and some point later in October where it got completely wiped out.

It was another case of a site with authority, great backlink profile and good content that was poorly built from a performance and architecture POV. 

We conducted an extremely similar process on this site as the site above, with the exception of not starting from a blank wp install, as the site didn’t have that much history and the database wasn’t as messy. 

The main takeaways of this rebuild were:

  1. Ensure siloing is spotless. As the offers were state-specific, we worked on each state starting from the most competitive to ensure the topical and local relevance were right. 
  2. Mobile first. Both from a monetisation and content organization POV, as the site previously had a design that was good on desktop, but very poor on mobile, hurting the overall performance and trust for mobile users
  3. Competitor analysis.  On this particular niche, every competitor ranks for the core keywords with the homepage. We spent considerable time reverse engineering their home portals and deciding where to place our maps, offers, and how to structure the content. Our in-house designer put several designs together and we worked on those until we were 100% certain this site’s homepage was better than every other site in page 1 by a long shot. 
  4. Keep it smart with the budget. Our client already was on a hard spot after purchasing this asset and suffering its decline and loss on cashflow. We wanted to make sure we got the site recovered within a budget that woudn’t dig our client even deeper into the hole. We spent some good time analysing all the angles and possible avenues for recovery trying to do what’s most cost effective. Hence, the entire project was kept very low cost, as we didn’t incur into expensive offpage link building and we focused on leveraging the existing content as good as possible. 80% of the SEO done on the site was on the highest tech-and-onpage SEO angles.  We still had to build some links to replace juice lost on the disavow. 

Since these were pretty big, wholesale changes, we used a staging server and didn’t push the changes live until they were all done.

This is pretty standard practice and allows you to get everything perfect, rather than updating things on the fly.

It also meant that it was going to take a while for Google to process all the changes, and even longer for them to kick in.

When making changes like this to a site, it can seem like nothing is actually happening at first, and there’s no guarantee it will, which is why we decided to share these results.

In our case, one site responded after about 4 months, and the other after 3 months. In both cases the response has been emphatic once it kicked in though.

What’s also worth noting is that because we improved the UX and the structure on the first site, the display ads have been responding with increasingly high RPM’s.

It’s often the case with display ads that as you get more traffic, your RPM’s decline at first, because advertisers need time for their ad budget to adjust to your increased traffic. With this site, we’ve just seen everything compounding and as traffic goes up, RPM also goes up, and therefore total revenue goes up twice.

Ignore the final day which hasn’t updated yet, there was no sudden dip.

This site has basically tripled its daily display ad earnings over the course of a month, with us doing nothing except watching Google finally reward the changes we made back in December.

Note: Early April is traditionally a low earning period anyway, because the first weeks in a new quarter are slow, as advertisers reset their quarterly ad budget. This means that it is somewhat inevitable that a site will increase its RPM throughout April. However, none of our other sites saw such a quick acceleration as this one.

How To Replicate This

It’s important to audit your sites regularly, and one of the first things we do when buying a site with investors, as part of our Full Service option, is to audit a site for issues and opportunities.

However, for most of you, you probably only have minor issues that need fixing. The sites in this case study were in need of some severe TLC and had been on a decline for some time. This is why we set up staging servers and proceeded with major overhauls.

Going into full details of the things you should check and exactly how to perform an audit is slightly beyond what this article is about, so here is an overview of the process.O

We run our standard audit process on the site as follows:

  1. Manually navigate throughout the entire site and take as many notes as possible. Things that don’t work, unrelated topics linking to each other, dead links… this part takes a bit more of a seasoned eye.
  2. Crawl site with sitebulb. Main thing we’re looking at here is the crawl depth and crawl map. If the depth goes beyond 5, and the map looks messy, we know architecture needs working on
  3. Check for errors and number of indexed pages on GSC. Many sites suffer from index management issues where tons of urls are added to the index but shouldn’t.
  4. Check for all linking domains on ahrefs and do a bulk analysis of those. Any sites not getting traffic or with very low DR we need to check to see if they’re legit or worth disavowing.
  5. Check mobile speed, accessibility and best practices through a Chrome audit. (F12 on chrome, >> Audits)
  6. Analyse the competitors in the space. How are the 3 strongest sites in the niche built? How does their content compare to yours? Their backlink profile?

After running those checks on both sites, we knew we had to go deep. And it’s paid off.

The moral of this story is that when you do take the time to sit down and completely overhaul a site, you are going to have a few months of patience on your hands, but the end results can be well worth it.

We also just wanted to give an update on some of our results, because they deserve celebrating.

Get in touch if you’re interested in having us help you buy or site, or partner up and have us completely turn a declining site’s fortune around.

Share with: