The latest hurdle in the search industry is the possibility of 100% of search referral data to be hidden from Google organic search. Yowza. I’m not going to lie, that stings. But, as usual, marketers will not only find a way, but the savvy marketer will find a way to stand out from the pack.
Rand has already put up a whiteboard video that covers the approach to existing in a world without search term referral data. Rand’s video did a great job of quickly hitting why keyword data was important, where keyword data gaps will now exist, and how to fill those gaps with other data.
I wanted to take Rand’s advice and dig a little deeper into specific techniques. I’ll cover this in two parts: how to fill the gap in finding opportunity, and how to measure the economic value of your SEO efforts.
Finding opportunities
The first thing we’ll address is replicating the opportunities previously found in keyword data. In yester-year, we used to be able to look at keyword data, trend it, and say “Hey, this keyword is really picking up speed, what gives?” Then, we’d dig a little deeper and find that typically either some new content was now being found for a keyword that we might be able to push or that for some reason existing content was now experiencing an increase in searches due to a change in search trends. That luxury is gone and we’re going to have to find ways to replace it.
Here’s a few ways to help fill the void of missing keyword data.
- Monitoring rank data
- Filtering Google analytics by landing page and traffic medium
- Google webmaster tools reporting
- Keyword data from paid search campaigns
Rank data
Monitoring keyword ranks can help you get an understanding of when a keyword or cluster of keywords starts to make drastic progress. When it comes to looking for opportunity, you’re going to want to focus on bigger changes. Small fluctuations between SERPs is becoming decreasingly reliable with personalization, localization, etc.
I find rank tracking data much more reliable when used in conjunction with other techniques. The rank itself, as many have pointed out, is hard to quantify with a single number because of the factors listed above. However, it’s a good general gauge for where you stand and when used with other data can still be useful.
If your targeted term is tracked at a rank of 34, but you’re seeing natural traffic to that landing page, you can assume the bulk of it is not from the targeted head term. Now that we’ve lost the simplicity of keyword data in our analytics tools, we’ve got to rely on multiple data points to paint the picture.
Pages that perform well, are seeing some natural traffic, and have room for rank improvement can represent opportunity. Rely on advanced segments to help you find the page, then utilize your rank data to help you identify which pages have the most room for improvement.
Advanced filters in Google Analytics
Since we’re now unable to analyze keyword data, we’re going to use data at the page level that comes from organic search. Just remember that you are making an assumption when you do this: all the terms referring traffic to this page are of the same or similar value. The worst violator of this assumption is usually the brand term, so you’ll want to make sure you’re brand terms aren’t a big portion of the referring traffic here.
Here’s a sample filter for the marketing page of my agency’s site, also filtered for organic referrals only.
Use an advanced segment in Google Analytics to see the natural traffic to a given page.
Ok, so now we can use the advanced segments in Google Analytics to examine keyword clusters, but how do we use this to identify opportunities? Here are two good approaches to getting more actionable data.
1. Use a filter to look at a landing page report with only organic search referrals. Trend the traffic data over time, and look for any pages increasing sharply. There are a few scenarios that are usually associated with this increase, but each represents some opportunity.
The page could have started to rank for a good term. Use rank tracking data to verify an increase in ranking; if there’s still room for improvement, push the term a little more.
A keyword that the page ranks well for may have seen a sudden influx of interest. You can use Google trends to help find the keyword that has seen a breakout in impressions. For a scenario like this, you’ve got a resurgence of interest for a term. Very often, it’s because of news breaking on an old topic, which means you may want to ensure your content is still relevant to the topic at hand.
2. Use a filter to look at a landing page report with only organic search referrals; now look at economic value driven by these pages. Find pages with a high economic value per visit that aren’t likely to have brand terms. This will help you identify pages that are performing well for the search traffic they’re receiving, but could stand to see an increase in rank or traffic for their targeted terms.
Take a sampling of pages with strong performance and try to identify their keyword rankings and search volume for their targeted terms. At this point, you’ll have married up performance with potential. Take this list of pages with a high economic value that have potential for rank increases and you’ve got yourself a pretty solid to-do list.
Google Webmaster Tools
Along the same lines as identifying pages with sharp increases of traffic, we can use Google’s Webmaster Tools to identify search queries that have seen sharp increases in impressions.
Here’s a sample formula I used to identify opportunities. Export your search query report from Google Webmaster Tools and drop this formula in column J:
=IF(C2>1,IF(H2>4.9,”Opportunity Knocks!”,””),””)
This is just two IF statements that say, if the change in impressions is greater than 100% and the average position is greater than 5, mark it with “Opportunity Knocks!” The idea is to find upward moving keywords with potential to be pushed farther. Here’s a sample of that in action.
You’ll notice that the two terms that have been identified as an opportunity have sharp increases in impressions and are in the average position of 11. Both of these terms could see a nice increase in volume by making it to an average position on page 1 of the search results. In short, opportunity knocks.
Paid search keyword data
Now, I know this one is going to be somewhat controversial since this is probably “exactly what Google wanted” with an update like this. That said, in my opinion using paid search to test market is a tool that does not get enough play in SEO. Here’s a couple paid search approaches to SEO I like to use.
1. Check the value of head terms – SEO is all about driving economic value, not just traffic. So, when I get a client that’s relatively inexperienced in SEO & paid, I encourage testing the head terms they’d like to target in SEO in paid search first. Running a few campaigns using exact match keywords will help you get a feeling for what head terms drive the most economic value for your site.
2. Find mid and long tail keywords – another benefit to using paid search as a testing ground for SEO is the search query data that comes from running paid campaigns. Use phrase match to find close variations on your head term and use fully modified broad match keywords to help you find a wider range of queries.
Get a search query report and prune through it to look for mid and long tail inspiration. In addition to finding a whole mess of search query data, you’ll also start to better understand the intent of searchers for a given keyword. It can help you anticipate and avoid noise associated with keywords as well.
3. Use product listing ads to identify product page SEO opportunities – if you’re working for an eCommerce site this is an amazing source of data. Go to your product ads campaign and check the keywords tab; you can still see the search queries for this campaign even though you haven’t specified any keywords.
These are search queries that are landing directly on product pages. You’ve got built in search volume, conversion rates, etc. Look for top performers and check your rankings against those keywords and find opportunity.
4. Understand location specific volumes and conversion – use geo-targeting to simulate localization in organic search. You can find out exactly what the search volume for a given term in your area as well as determine its economic value to you. You’re going to get a far clearer picture than any keyword tool could provide. So, before you wind up pushing a term that doesn’t convert well, test out a few head terms with geo-targeting and find what works best for your site. Use this in combination with phrase and fully modified broad match keywords and you’ll wind up finding some geographic terms as well.
Connecting effort to economic value
Anyone that’s ever asked for a budget in SEO knows you have to connect your efforts to the bottom line or it’s not happening. You’ve got to prove that the time you and your team spend glued to your monitors is providing value or those paying your salary will assume you’re just trolling Buzzfeed for the 25 ways your life is like an 80s movie instead of working.
All right, so keyword data has disappeared and you can’t easily put a report in front of a CMO or a client that says “this keyword we targeted is driving this many conversions.” Bummer. It’s sad, I know, but we’ve got to move on. Here are a few methods for determining the real economic value of your efforts.
Monitor overall organic traffic and subtract brand increases
A method Rand mentions in his video is to essentially look at organic traffic trended over time, then use other sources to subtract any increases of brand search volume. This is a viable method to get an overall feel for how things are trending. This kind of approach could provide a good overall number that you could float to someone controlling the budget on your work.
Using branded paid search campaigns to monitor brand traffic increases is the most accurate way to determine when an upswing of organic traffic is brand related or not.
There are a few drawbacks of this approach though:
- There are other things that factor into a change in brand traffic. Competitors advertising on your terms, a higher paid share of the keywords, or even a set of keywords casting a wider net could all cause brand search volume changes. These factors make it difficult to determine exactly how much organic brand search volume will be affected.
- When drilling into economic value, how much is a product of the brand increases and how much is a product of the non-brand SEO efforts?
- Using ratios and top-level averages often hides the most valuable insights.
If you’re going to use this approach, just be aware of your assumptions. Don’t try to get exact data out of this, it’s not a method that can be used to say “Non-brand search is up 3.6% this month” I would use it over much longer time frames and only for larger gains.,
Trended increase in value from targeted landing pages
Using the same Google Analytics advanced segment we used in the finding opportunities section, we can use that to trend an increase in traffic as well as economic value. In this scenario, we’re still going to want to police the data for brand terms. There’s a few solid ways to determine the real value of your non-brand search traffic.
1. Use the sampling of Bing keyword data to determine how large of a factor brand keywords are. If you’re seeing not only a decent amount of brand queries pop up in your reports, you can bet that they’re appearing for Google as well. If you have a good enough sample of Bing keywords, how much economic value is coming from your brand terms? If it’s a large percentage, you’re going to have trouble determining the real value of the non-brand traffic to this page.
2. Export a search query report from Google webmaster tools. Use this formula in excel to search for your brand in any of the search queries.
=SEARCH(“Brand”,A2,1)
If there is a brand term in that query, the formula will shoot back a number. Otherwise it will spit out #Value!. Sort your search queries with this formula and examine those that return a number. Those are your brand search queries, so take a look and see if any are likely to be landing on the page in question. A quick Google search should let you know what page is appearing for that query.
If you’re finding a lot of volume for a brand keyword is landing on the page you’re examining, its going to be difficult to weed out the non-brand organic search value of that page.
3. Is there an increase in non-organic traffic to this page? Why non-organic? If there’s been some other kind of marketing push that would also increase brand traffic to this page, there’s probably a non-organic increase in traffic too. PR, events, direct mail, and other marketing efforts that tend to cause an increase in specific brand search volume will also cause an increase in traffic to a page that’s non-organic. Use a filter like the one below to make sure this isn’t the case:
4. Use paid search to determine economic value for targeted terms. If we’ve got a murky situation that we can’t seem to weed brand terms out of, but we still want to know the value of the non-brand terms we’re sending to the page, we can always rely on paid search data.
If it’s not economically feasible to keep the campaigns running to these targeted pages you can always send a sample to your targeted pages and rely on sample conversion rates and economic value.
When it comes to measuring the value based on paid search keywords, make sure you’re looking at search query reports and not just keywords. This will get you the granular data you want.
5. Failing convincing your company or your client to run paid search campaigns to measure value, you can always use the value of a paid click. If the average CPC for a targeted term is $3.50, you can assume any natural traffic you drive to the page can be valued similarly. But, in this scenario, remember that if the click was indeed worth the average CPC to that site, they’d have no problem running paid search ads in the first place 🙂
Now that we’ve worked out a method for measuring the value of you non-brand organic efforts, you can start to paint a picture of the value your efforts providing. Since this is a pretty granular look at things, you may want to start to cluster some pages together. Depending on who you’re presenting this to and what your goals are, that’s going to affect how you group these sets of pages.
You can still utilize the advanced segments in Google Analytics to cluster your targeted pages together. For instance, this regular expression below will report any data for organic visits landing on pages in the blog directory.
If I were presenting this to an eCommerce client that I told we were going to focus on his product page SEO and the long tail, I would group the results as shown to demonstrate the trended increase. The percentage increase in traffic is an easily digestible figure for a client, and so is a dollar amount that represents the incremental increase in value from our efforts.
We’ve taken a somewhat complex analysis, boiled it down to as few assumptions as possible and made it an easily digestible figure. Showing the incremental value as a dollar amount helps you establish an ROI on invested time and justify a budget in the future.
Wrap-up
It’s a tough transition and a daunting task for marketers, but there’s still plenty of data out there to take an analytical approach to SEO. I’ve outlined a few good techniques for existing in a world without keyword referral data, but I’d love to know, is there anything you’re doing to fill the keyword void that I’ve missed? Let me know in the comments.
