Everything you need to know about SEO, delivered every Thursday.
Yet More Tips For Diagnosing & Fixing Panda Problems
Here we are, five months after Panda and the only publicly confirmed Panda recovery stories are those that followed the Panda 2.3 update. This is notable because this particular update was one in which Google took actions to help restore some sites that were unfairly hit by the earlier versions of Panda.
Google told Barry Schwartz about the update: “this update incorporates some new signals that help differentiate between higher- and lower-quality sites. As a result, some sites are ranking higher after this most recent update”.
So what is the formula for Panda success? It starts with understanding what the underlying causes are.
Restating Google’s Goals
I view Google (and Bing) as having two major objectives with their search results:
- Find the fastest possible answer to the searcher’s question. It is important to internalize this. Can the user see what they want in a sub-second scan of your page? Or at the very least, can they tell that they will be able to get what they want, even if it requires a bit more effort, in less than a second? Consider this a working definition of a quality page.
- Sites with the highest level of engagement. Put very simply, this equates to a site that can be trusted. Search engines like sites they can trust. Why? Because they are not likely to spam their customers, and are more likely to provide fast answers to user’s questions.
There are tons of tools that Google has to measure engagement and trust. One of the most important ones, in my opinion, is the user’s click interaction with search results. If the position you are in normally gets a 10% click through rate for a given search phrase and you are getting 5% CTR, that is a strong signal to move your ranking down.
If users who click on your result have a higher than normal tendency to hit the back button and click on a different result – this is also a bad signal.
In my example image, the user clicked on the first result, did not like it, came back tried the second, did not like it, and so forth. This is exactly what search engines want to avoid.
In their ideal world, each of these SERPs has different takes on the problem of making french toast, and a title and description that clearly calls out what that that difference is. So if the answer to the user’s question is to be found by clicking on the 3rd result, they can easily figure that out right in the SERPs.
Engagement is a different matter, and there are tons of available signals. Here are some examples:
- Average Time on Site
- Page Views Per Visitor
- Repeat Visitors
- Brand Name Searches
- Social Media Mentions / +1s / Likes / Tweets
- Traffic from Non-Search Sources
- Pages Printed
- Scroll Bar Usage
Which signals the search engines use do not matter at this point. I know we all like to know what they are, and I am no exception in that regard. But, if you are looking to future proof your SEO, you need to start with the assumption that the particular signals that are in use are going to change.
So even if you can determine the precise signals that matter at the moment, you are well advised to solve the bigger problems of providing fast answers and high engagement.
9 Panda Diagnostic Steps
The place to start is with figuring out what is going on with the user experience on your site. Here are some key steps in that process:
- Ask yourself some tough questions. Did you put a lot of editorial effort into the content on your site? If not, you will probably need to address that. Poor quality content will be noticed by users and they will vote with their clicks. For e-commerce sites that use manufacturer supplied descriptions, this may mean some tough challenges ahead. If you are competing with Ebay, Amazon, Macy’s, or other major brands, and you are selling the same items with no added value you have some big issues to address.
- Now for the second tough question. Did you have providing the user the fastest possible response to their question in mind when you designed the site? If not, you will have to think about addressing that too.
- Compare your engagement metrics with those of your competitors. You can look at signals such as time on site, repeat visitors and page views per visitor with services such as Compete.com and others like them. Yes, the data accuracy is crude at best, but you are looking for directional indicators not a calculation required to land a spaceship on the moon. Compare your brand search volume using the Google Adwords Keyword Tool. If these metrics seriously lag your competitor, this may be telling you that you have work to do.
- Examine your incoming search queries for more information. What do they tell you about what the users coming to your site are looking for? Are you getting the right visitors? Do you need to adjust your titles and the content to make sure you are getting people to your site that you are going to be able to help?
- Survey the visitors to your site. Ask them what they are looking for. Give them a multiple choice list of logical answers, and then an “Other” that they can check where they can manually enter in anything else that they feel better describes what they want. This is a great way to truly determine what they are looking for.
- Do some surveys of people who are not at your site. Mechanical Turk is a surprisingly good way to get this data. A survey there can give lots of interesting data, and if your traffic is low at the site might be the best way to quickly find out what people want when they come to a site like yours.
- Set up mouse movement tracking software on your site. See what users are looking at on the page. Do the majority of them go straight to the right column of your 3 column site? If so, then that content is closer to what they actually want then what is currently in the center of your page.
- See if you have pages on the site with a very high bounce rate as compared to other pages on the site. It may be that you have one section of the site that is hurting you. Unfortunately, it is probably rare that life will be that simple for you, but it is worth taking a look. If you see that this is the case, consider no-indexing those pages.
- Ask yourself if you are hitting the users with too many ads. Is a large portion of the above the fold real estate used up by advertising plus your own banner graphic if you have one? If so, plan on changing that.
A Couple of More Key Points
The above diagnostic steps are a great place to start. Next up are two specific areas that I want to highlight.
Review your content for uniqueness
Certainly duplicate content is not a good idea, such as the use of canned manufacturer descriptions on an e-commerce site, but it goes much deeper than that.
For example, if you write an article on “how to make french toast” you have lots of company (854,000 results come up for that search query in Google including the “”). There is no way that you are adding any new value.
The same is true if you write an article with basic mortgage advice. Unless you are a major brand, you won’t stand a chance competing against major brands unless you do something different enough that users will notice it and vote with their clicks, shares, +1s, and other similar signals.
This is largely because the trust users place in a brand will cause them to generate better engagement signals on their article about french toast than yours – unless yours is a lot better. Superior site design or some unique new angle might do the trick. However, some of those pesky brands may be investing in their website engagement too.
Even if you are a major brand, points of actual differentiation in the content and experience you provide will provide you with major advantages over your competition. Invest in it to give yourself an edge.
Include positive brand signals on your site
Not a major brand? Then these things are even more important. Your site needs to scream credibility and trustworthiness.
Following up on one of Google’s suggestions, survey site visitors, and groups of non-visitors too, and ask users if they would give a credit card to your site. Keep making tweaks and enhancements until a lot of them start saying yes.
As you can see, this approach is designed to attack the two basic Google Goals head on. This is not easy work, but it is the work that must be undertaken if you are going to get out from under a Panda setback. It will take time.
I have focused mostly on the initial diagnostic steps and not the follow up actions, but at the end of the day, the major items you will most likely have to work on are:
- Content Quality / Uniqueness of Content
- Site Design and Layout to help users find answers faster and improve engagement
- Creating a Positive Brand Experience – make the users comfortable with your site
All evidence points to this being a grinding process for most people. This is unfortunate, but it is best to start the hard work now.
If it does not make sense for you to take the long and thorough road, you may want to consider cash cowing the site, and taking what it can still give you and investing in other projects.
To all of you who are offended by the prior statement, my apologies, but I am not making the rules here, I am just trying to help with interpreting them the best I can.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.