A couple of months ago I watched as search traffic to this site declined rapidly. At it’s worst the 2 week drop was about 40% There were several things I could have done, but in the end I chose what turned out to be the best seo tactic of all…nothing.
It’s easy to panic when things suddenly take a turn for the worst. However, panic is never the best solution to anything as it only leads to poor decision making. While I certainly wasn’t happy with what was going on, I never panicked.
Instead I analyzed and researched, and ultimately chose to wait before doing anything rash. Doing nothing turned out to be the right choice as my search traffic has not only come back, but increased by about 15%, all from me changing not a thing.
2010 was a pretty good year for me in terms of traffic growth into this site. It hasn’t all been search traffic, but search makes up anywhere from 35–45% of the traffic here depending on the month. For most of the year traffic had steadily grown, punctuated by an occasional large step as I found new sources to market the site.
The week of Thanksgiving, traffic dropped as you might expect. I didn’t think much of the drop and expected it would return the following week and then perhaps see a similar dip as we approached Christmas and then New Year’s.
While traffic did come back, it was a little less than I expected based on the weeks leading into Thanksgiving. Traffic then took a nosedive a little earlier than I would have thought to see any decline. I could pinpoint the drop to December 14th.
Over the course of about a month (Dec 14 – Jan 13) traffic from search engines (mainly Google) was down 20% over the previous month.
Some of this was expected, but again it happened earlier than I thought and the drop was also greater than expected. What’s worse is that once the holidays ended traffic didn’t seem to come back to the same level where it had been. It was about 10% less than where it had been prior to Thanksgiving.
While in the middle of the traffic loss I had to decide what to do.
Research and Analysis
Instead of panicking I went with 2 other courses of action.
- Search the web to see if anyone else was reporting similar issues and even more to see if anyone offered a cause and a solution
- Dig a little deeper into the numbers to compare to previous years and see if the data could reveal more about the possible cause of the traffic loss
The first didn’t lead to much. A few people reported traffic loss and one or two started seeing the loss about the same date I was seeing it. However there was nothing to indicate a possible cause and no solution.
Through Analytics I found I was mainly losing traffic from long tail keywords, which I found troubling. I wondered if perhaps Google had made an algorithmic change that was no longer favoring this site on many long tail phrases.
Earlier in the year Google had made such a change, now commonly referred to as their May Day Update. Lot’s of sites lost long tail traffic at that time. This site wasn’t one of them. In fact the May Day Update started sending more traffic my way. It didn’t seem logical to think that 7 months later I was finally being affected by that change.
Looking at the numbers
For comparison here are traffic numbers in terms of visits and the associated number of keyword phrases sending that traffic between mid October and mid February, the last 2 years.
|Oct 14 – Nov 13||8,953||5,092|
|Nov 14 – Dec 13||9,402||5,118|
|Dec 14 – Jan 13||8,326||4,724|
|Jan 14 – Feb 13||12,396||6,464|
|Oct 14 – Nov 13||23,978||13,475|
|Nov 14 – Dec 13||22,863||12,807|
|Dec 14 – Jan 13||18,639||10,221|
|Jan 14 – Feb 13||27,849||15,320|
In both years numbers are up significantly after the middle of January. Last year the dip in December was marginal. This year it was quite significant.
Also note that while I’m showing the numbers from mid January to mid February here, those numbers didn’t exist at the time of this analysis. Note too that last year’s January increase can be explained by some high quality links I had gained during December.
While neither searching nor analysis shed much light, I did have some thoughts on what the problem might be.
- Normal loss of traffic during the holidays—I had expected less search traffic around the holidays, just not as much as I was seeing.
- Search loss was tied to referral loss—Referral traffic had also dipped, though not quite as much as a few sources that had been sending consistent traffic stopped. Mentions on Twitter and Facebook were also down some. I thought perhaps there was a connection.
- A significant algorithmic change favoring sites that weren’t mine—Since I didn’t find much when searching online I assumed this didn’t happen.
- A loss of indexed pages—Since long tail traffic was down i thought perhaps a number of pages might have been removed from the index or maybe there was some kind of duplicate content issue. I couldn’t find any evidence of this.
In the end I found nothing conclusive or even promising as a possible cause and solution. Given that anything I would have done would have been based on a guess, I felt taking a wait and see approach was the best option.
I planned on more research as more data came in.
Traffic Comes Back
Waiting was the right choice as you can see in the data above. It took a little longer than expected, but the lost traffic came back. In fact it’s been steadily rising ever since and is now up about 15% from pre-Thanksgiving levels.
Once again let me remind you that I did nothing to bring the traffic back.
Naturally I’ve done things to market this site, including some things to increase search traffic, but I haven’t changed anything significantly from what I was doing back in the fall. Just continuing with what I had been doing.
My guess is that in the end what I observed was normal holiday traffic loss for a site like this one. The links gained last year might have masked a similar drop. I’ll have to pay attention next year.
What I Might Have Done
Had I been in panic mode I might have made some rash decisions. I might have
- Begun mass rewriting of page titles and content across the site
- Restructured the site content
- Attempted to fix an imaginary duplicate content issue
- Spammed the web for quick links
The problem with any of the above is that as my waiting clearly showed there was nothing wrong with the site. At least nothing wrong causing this problem. I’m sure there’s plenty of things wrong with the site not tied to this problem.
Making changes could have done more harm than good. I hope you’ll agree that spamming the web is never a good option, but even the other things I could have done weren’t necessarily good options either.
I would have been trying to fix a problem that didn’t exist. Fortunately I chose not to.
Every business has its ups and downs and that includes how much search traffic is flowing into a site. While no one enjoys experiencing the downs the worst thing you can do is panic when they happen.
What I mainly want you to take from this post is the lesson to never panic. Sometimes through no fault of your own things will take a turn for the worse. If you panic when it happens you could make things far worse and cause the problems you hope to solve.
When things go bad do try to figure out why. See if the same bad things have happened to others and take a deeper look into any changes you might have made. If you’re sure there’s a problem ask others to take a look as well.
However if you don’t find anything conclusive don’t rush to take action for the sake of doing something. Sometimes the best strategy or tactic is to do nothing.
Download a free sample from my book, Design Fundamentals.
Maybe the reason traffic declined is that your blog post quality declined and nobody wanted to read bad blog posts, it’s generally the number one factor is declining blog traffic.
Would you say your blog posts from the period where your traffic was at its highest are of the same quality now?
Was most of your blogs high traffic because of one or two key posts?
You need to ask yourself these questions.
I don’t think the quality of the post dropped off during the time in question, though perhaps I’m not the best judge. You’re welcome to read the posts between mid December and mid December and decide.
However we are talking search traffic and search engines aren’t all that good at being able to judge quality.
Search traffic here flows into several hundred posts. The most popular posts were as popular as always. I didn’t notice any significant drop in traffic to them.
I think in the end most of the loss was just due to people searching for different things around the holidays. More searches for gift items and less for information. I think I noticed it more this year, because I get more search traffic this year than last and also because last year the extra links I had earned around the time masked the same kind of drop off.
I’ll see what happens when the holidays roll around again.
Loved this Steven – thanks for writing. You are right, frequently doing NOTHING is the smartest thing of all. You can cause more problems that you then have to solve.
Thanks Adam. I think some do tend to panic when something goes wrong and want to rush to change, but making changes when you can’t identify the problem doesn’t make sense. You can just as easily make things worse as better.
I had to look twice at your charts and stats as they are near identical to mine and several of my clients, all of which match exactly what you experienced over the mentioned weeks, months.
I think, you perhaps nailed the issue on the head with associating the end user trends as being more product/gift search rather than information search during the holiday season, it would certainly make sense, and fit with the analytic data i have been looking at.
Funny coincidence that our stats are so similar.
It seems natural that during the holidays people are going to search different. I would imagine ecommerce sites typically see a little extra traffic and informational site see a little less.
I didn’t think the drop would start as soon as it did or last as long based on last year, but now I’m thinking the same thing did happen last year, but was masked by some marketing I had done the month before.
Something to keep an eye on for next year.
Don’t you have traffic stats from last year? Your content is not seasonal, but perhaps part of the audience has other preoccupations during that particular period of time. Good aproach anyway. I sometimes have people telling me to “do more SEO” in order to increase traffic, without identifying the reason why there’s not enough traffic…
Your articles are still interesting, I don’t think you need to worry about that. 🙂 As a non-designer I particularly appreciate your clear and simple (yet not simplistic) exposés on design and information architecture. And I can’t wait to read your chapter in the new Smashing book. Keep up the good work!
I do have statistics for every year this site’s been live. I compared the last two years in a table in the post.
I agree the content here isn’t seasonal. I expected that there would be less traffic here around the holidays as it’s naturally on people’s minds and I assume they search more holiday related things at that time.
The temporary drop in traffic isn’t something I worry about. This post is more to show how not worrying is a good thing. I think other in my situation might have made changes they didn’t need to make to fix a temporary problem.
Wait and see is sometimes the best approach to take with seo. You shouldn’t jump to act until you have all the information.
This is a great collection of information. Thanks for the help.
I am new to SEO and promoting a website frustrates me big time. I didn’t know what to do, until I landed on this page – which is awesome. I thought to myself that panicking won’t do me any good and you have helped me affirm it, so thank you.
Thanks Karl. Panicking is usually never a good thing. There are definitely times where there is a problem that needs a quick response, but you generally want to at least make sure you first know you have a problem needing a quick fix before acting. Otherwise your making blind changes that could cause more problems than they aim to solve.
Declining traffic is an emotional punch in the face. Every day of increasing traffic just “that’s how it works” but a few days of dwindling traffic can be devastating.
You showed a lot of discipline waiting it out.
True. No one wants to see their traffic go down, myself included. I think people have a tendency to panic though because of that emotional punch. They want their traffic back yesterday and so start to make changes.
Sometimes you have to be patient and figure out what’s really going on or you can make things worse. Search traffic goes up and down for a variety of reasons. A temporary decline isn’t always an indication of a real problem.