Monday, March 5, 2018

Declining Organic Traffic? How to Tell if it’s a Tracking or Optimization Issue

Posted by andrewchoco

Picture this scenario. You’re a new employee that has just been brought in to a struggling marketing department (or an agency brought on to help recover lost numbers). You get access to Google Analytics, and see something like this:

(Actual screenshot of the client I audited)

This can generate two types of emotional response: excitement or fear (or both). The steady decline in organic traffic excites you because you have so many tactics and ideas that you think can save this company from spiraling downward out of control. But there’s also the fear that these tactics wont be enough to correct the course.

Regardless of whether these new tactics would work or not, it’s important to understand the history of the account and determine not only what is happening, but why.

The company may have an idea of why the traffic is declining (i.e. competitors have come in and made ranking for keywords much harder, or they did a website redesign and have never recovered).

Essentially, this boils down to two things: 1) either you’re struggling with organic optimization, or 2) something was off with your tracking in Google Analytics, has since been corrected, and hasn’t been caught.

In this article, I’ll go over an audit I did for one of my clients to help determine if the decline we saw in organic traffic was due to actual poor SEO performance, an influx in competitors, tracking issues, or a combination of these things.

I’ll be breaking it down into five different areas of investigation:

  1. Keyword ranking differences from 2015–2017
    1. Did the keywords we were ranking for in 2015 change drastically in 2017? Did we lose rankings and therefore lose organic traffic?
  2. Top organic landing pages from 2015–2017
    1. Are the top ranking organic landing pages the same currently as they were in 2015? Are we missing any pages due to a website redesign?
  3. On-page metric
    1. Did something happen to the site speed / bounce rate / page views etc.
  4. SEMrush/Moz keyword, traffic, and domain authority data
    1. Looking at the SEMrush organic traffic cost metric as well as Moz metrics like Domain Authority and competitors.
  5. Goal completions
    1. Did our conversion numbers stay consistent throughout the traffic drop? Or did the conversions drop in correlation with the traffic drop?

By the end of this post, my goal is that you’ll be able to replicate this audit to determine exactly what’s causing your organic traffic decline and how to get back on the right track.

Let’s dive in!

Keyword ranking differences from 2015–2017

This was my initial starting point for my audit. I started with this specifically because the most obvious answer, for a decline in traffic is a decline in keyword rankings.

I wanted to look at what keywords we were ranking for in 2015 to see if we significantly dropped in the rankings or if the search volume had dropped. If the company you’re auditing has had a long-running Moz account, start by looking at the keyword rankings from the initial start of the decline, compared to current keyword rankings.

I exported keyword data from both SEMrush and Moz, and looked specifically at the ranking changes of core keywords.

March was a particularly strong month across the board, so I narrowed it down and exported the keyword rankings in:

  • March 2015
  • March 2016
  • March 2017
  • December 2017 (so I could get the most current rankings)

Once the keywords were exported, I went in and highlighted in red the keywords that we were ranking for in 2015 (and driving traffic from) that we were no longer ranking for in 2017. I also highlighted in yellow the keywords we were ranking for in 2015 that were still ranking in 2017.

2015 keywords:

2017 keywords:

(Brand-related queries and URLs are blurred out for anonymity)

One thing that immediately stood out: in 2015, this company was ranking for five keywords, including the word “free.” They have since changed their offering, so it made sense that in 2017, we weren’t ranking for those keywords.

After removing the free queries, we pulled the “core” keywords to look at their differences.

March 2015 core keywords:

  • Appointment scheduling software: position 9
  • Online appointment scheduling: position 11
  • Online appointment scheduling: position 9
  • Online scheduling software: position 9
  • Online scheduler: position 9
  • Online scheduling: position 13

December 2017 core keywords:

  • Appointment scheduler: position 11
  • Appointment scheduling software: position 10
  • Online schedule: position 6
  • Online appointment scheduler: position 11
  • Online appointment scheduling: position 12
  • Online scheduling software: position 12
  • Online scheduling tool: position 10
  • Online scheduling: position 15
  • SaaS appointment scheduling: position 2

There were no particular red flags here. While some of the keywords have moved down 1–2 spots, we had new ones jump up. These small changes in movement didn’t explain the nearly 30–40% drop in organic traffic. I checked this off my list and moved on to organic landing pages.

Top organic landing page changes

Since the dive into keyword rankings didn’t provide the answer for the decline in traffic, the next thing I looked at were the organic landing pages. I knew this client had switched over CMS systems in early 2017, and had done a few small redesign projects the past three years.

After exporting our organic landing pages for 2015, 2016, and 2017, we compared the top ten (by organic sessions) and got the following results.

2015 top organic landing pages:

2016 top organic landing pages:

2017 top organic landing pages:

Because of their redesign, you can see that the subfolders changed between 2015/2016 to 2017. What really got my attention, however, is the /get-started page. In 2015/2016, the Get Started page accounted for nearly 16% of all organic traffic. In 2017, the Get Started page was nowhere to be found.

If you run into this problem and notice there are pages missing from your current top organic pages, a great way to uncover why is to use the Wayback Machine. It's a great tool that allows you to see what a web page looked like in the past.

When we looked at the /get-started URL in the Wayback Machine, we noticed something pretty interesting:

In 2015, their /get-started page also acted as their login page. When people were searching on Google for “[Company Name] login,” this page was ranking, bringing in a significant amount of organic traffic.

Their current setup sends logins to a subdomain that doesn’t have a GA code (as it’s strictly used as a portal to the actual application).

That helped explain some of the organic traffic loss, but knowing that this client had gone through a few website redesigns, I wanted to make sure that all redirects were done properly. Regardless of whether or not your traffic has changed, if you’ve recently done a website redesign where you’re changing URLs, it’s smart to look at your top organic landing pages from before the redesign and double check to make sure they’re redirecting to the correct pages.

While this helped explain some of the traffic loss, the next thing we looked at was the on-page metrics to see if we could spot any obvious tracking issues.

Comparing on-page engagement metrics

Looking at the keyword rankings and organic landing pages provided a little bit of insight into the organic traffic loss, but it was nothing definitive. Because of this, I moved to the on-page metrics for further clarity. As a disclaimer, when I talk about on-page metrics, I’m talking about bounce rate, page views, average page views per session, and time on site.

Looking at the same top organic pages, I compared the on-page engagement metrics.

2015 on-page metrics:

2016 on-page metrics:

2017 on-page metrics:

While the overall engagement metrics changed slightly, the biggest and most interesting discrepancy I saw was in the bounce rates for the home page and Get Started page.

According to a number of different studies (like this one, this one, or even this one), the average bounce rate for a B2B site is around 40–60%. Seeing the home page with a bounce rate under 20% was definitely a red flag.

This led me to look into some other metrics as well. I compared key metrics between 2015 and 2017, and was utterly confused by the findings:

Looking at the organic sessions (overall), we saw a decrease of around 80,000 sessions, or 27.93%.

Looking at the organic users (overall) we saw a similar number, with a decrease of around 38,000 users, or 25%.

When we looked at page views, however, we saw a much more drastic drop:

For the entire site, we saw a 50% decrease in pageviews, or a decrease of nearly 400,000 page views.

This didn’t make much sense, because even if we had those extra 38,000 users, and each user averaged roughly 2.49 pages per session (looking above), that would only account for, at most, 100,000 more page views. This left 300,000 page views unaccounted for.

This led me to believe that there was definitely some sort of tracking issue. The high number of page views and low bounce rate made me suspect that some users were being double counted.

However, to confirm these assumptions, I took a look at some external data sources.

Using SEMrush and Moz data to exclude user error

If you have a feeling that your tracking was messed up in previous years, a good way to confirm or deny this hypothesis is to check external sources like Moz and SEMrush.

Unfortunately, this particular client was fairly new, so as a result, their Moz campaign data wasn’t around during the high organic traffic times in 2015. However, if it was, a good place to start would be looking at the search visibility metric (as long as the primary keywords have stayed the same). If this metric has changed drastically over the years, it’s a good indicator that your organic rankings have slipped quite a bit.

Another good thing to look at is domain authority and core page authority. If your site has had a few redesigns, moved URLs, or anything like that, it’s important to make sure that the domain authority has carried over. It’s also important to look at the page authorities of your core pages. If these are much lower than when they were before the organic traffic slide, there’s a good chance your redirects weren’t done properly, and the page authority isn’t being carried over through those new domains.

If, like me, you don’t have Moz data that dates back far enough, a good thing to check is the organic traffic cost in SEMrush.

Organic traffic cost can change because of a few reasons:

  1. Your site is ranking for more valuable keywords, making the organic traffic cost rise.
  2. More competitors have entered the space, making the keywords you were ranking for more expensive to bid on.

Usually it’s a combination of both of these.

If our organic traffic really was steadily decreasing for the past 2 years, we’d likely see a similar trendline looking at our organic traffic cost. However, that’s not what we saw.

In March of 2015, the organic traffic cost of my client’s site was $14,300.

In March of 2016, the organic traffic cost was $22,200

In December of 2017, the organic traffic cost spiked all the way up to $69,200. According to SEMrush, we also saw increases in keywords and traffic.

Looking at all of this external data re-affirmed the assumption that something must have been off with our tracking.

However, as a final check, I went back to internal metrics to see if the conversion data had decreased at a similar rate as the organic traffic.

Analyzing and comparing conversion metrics

This seemed like a natural final step into uncovering the mystery in this traffic drop. After all, it’s not organic traffic that's going to profit your business (although it’s a key component). The big revenue driver is goal completions and form fills.

This was a fairly simple procedure. I went into Google Analytics to compare goal completion numbers and goal completion conversion rates over the past three years.

If your company is like my client’s, there’s a good chance you’re taking advantage of the maximum 20 goal completions that can be simultaneously tracked in Analytics. However, to make things easier and more consistent (since goal completions can change), I looked at only buyer intent conversions. In this case it was Enterprise, Business, and Personal edition form fills, as well as Contact Us form fills.

If you’re doing this on your own site, I would recommend doing the same thing. Gated content goal completions usually have a natural shelf life, and this natural slowdown in goal completions can skew the data. I’d look at the most important conversion on your site (usually a contact us or a demo form) and go strictly off those numbers.

For my client, you can see those goal completion numbers below:

Goal completion name

2015

2016

2017

Contact Us

579

525

478

Individual Edition

3,372

2,621

3,420

Business Edition

1,147

1,437

1,473

Enterprise Edition

1,178

1,053

502

Total

6,276

5,636

5,873

Conversion rates:

Goal completion name

2015

2016

2017

Contact Us

0.22%

0.22%

0.23%

Individual Edition

1.30%

1.09%

1.83%

Business Edition

0.46%

0.60%

0.76%

Enterprise Edition

0.46%

0.44%

0.29%

Average

0.61%

0.58%

0.77%

This was pretty interesting. Although there was clearly fluctuation in the goal completions and conversion rates, there were no differences that made sense with our nearly 40,000 user drop from 2015 to 2016 to 2017.

All of these findings further confirmed that we were chasing an inaccurate goal. In fact, we spent the first three months working together to try and get back a 40% loss that, quite frankly, was never even there in the first place.

Tying everything together and final thoughts

For this particular case, we had to go down all five of these roads in order to reach the conclusion that we did: Our tracking was off in the past.

However, this may not be the case for your company or your clients. You may start by looking at keyword rankings, and realize that you’re no longer ranking on the first page for ten of your core keywords. If that’s the case, you quickly discovered your issue, and your game plan should be investing in your core pages to help get them ranking again for these core keywords.

If your goal completions are way down (by a similar percentage as your traffic), that’s also a good clue that your declining traffic numbers are correct.

If you’ve looked at all of these metrics and still can’t seem to figure out the reasoning for the decrease and you’re blindly trying tactics and struggling to crawl your way back up, this is a great checklist to go through to confirm the ominous question of tracking issue or optimization issue.

If you’re having a similar issue as me, I’m hoping this post helps you get to the root of the problem quickly, and gets you one step closer to create realistic organic traffic goals for the future!


Sign up for The Moz Top 10, a semimonthly mailer updating you on the top ten hottest pieces of SEO news, tips, and rad links uncovered by the Moz team. Think of it as your exclusive digest of stuff you don't have time to hunt down but want to read!

No comments:

Post a Comment