SEO without keywords

Life without keywords makes puppies sad

Pages, user intent,  and user paths matter now

SEOs were finally adjusting to life without ranking,  then late last summer the hummingbird algorithm update and total keyword [not provided] changes rolled out at the around the same time.  Every time there are Google changes, the SEO world freaks out, but these were two pretty big shifts at once.  But with time I’ve come to see this as a opportunity to  re-energize my thinking about metrics, SEO, and the goal of web sites.

A world without keywords

Digital marketing has so many different data points available, focusing on keywords was the simplest and easiest option.  But with so many variables, keywords never told the whole story. Now it’s time to work a little harder on SEO and metrics and create something better. And anyway, we cannot hack our way back to keyword data (believe me, I tried).  Keywords are truly dead (for now, at least).  The future is in refining our pages to answer the specific questions our visitors are trying to answer. And it just so happens that the Hummingbird update is very much centered on user intent and providing more relevant pages  — highly focused pages are the way to simplify your metrics and win in Google with this new algo update.

What is your page for

It’s not important to know all the keywords a page is ranking for, but it is important to already know what your page is about and how your visitors are using the page.  Questions to answer are:

  • Do you have visitors?
  • Where do these visitors come from?
  • Where do visitors go after they access your page?

Your metrics need to be as focused as your pages.  And honestly, those giant lists of “Top keywords” and “Top converting keyword” reports we digital marketing types were cranking out were never really that actionable or useful without context.  Good news! we don’t have to make any more of those keyword reports — they are now officially meaningless. More useful data can be found looking at the top pages overall to get a sense of what resonates with your audience.  We can infer the keywords just from that.  Most articles “rank” for the title of the content. But really, we should already know what our pages are for.

Hummingbird and user intent

Hummingbird deserves a complete post of its own, but I will echo Warren Lee’s  statement that  Hummingbird is about “concepts and not keywords” (Hummingbird In The Trenches: A Canary In The Coal Mine ).  Google is trying to answer complex questions more quickly using the data it has — the relationships between the keywords, the information Google has on the searcher, and the search history have far more influence on the SERPs than keywords on the page.  We can’t know how Google makes those leaps from intent to SERP.  The only thing we can control is making sure our pages answer specific questions and support specific goals.

Metrics without keywords

Avinash Kaushik has a great model here,  Search: Not Provided: What Remains, Keyword Data Options, the Future,  based on his “Acquisition, Behavior, Outcome” metrics.  He recommends sorting out traffic sources by channel so you can tell what methods are working to drive traffic. Avinash, as always, does a better job than I can of providing examples, explanations, and even templates for  Google Analytics, so  I recommend taking a look at his post to get some ideas.

My simplified view of a keyword-free SEO strategy:

  • Figure out what your page is for. Think in terms of the goal of the user. Build (or-rebuild) pages around goal-focused content to meet the user needs.  If people come to the page and take an action, then you are creating valuable content to you and your audience.
  • What metrics should you look at now?  The same ones you already should have available.  Visits/visitors/etc from:
    • Organic Search
    • Paid search
    • Direct visits
    • Social media
    • Other site referrals and whatever other channels you may using to promote your content
  • Then look at what people did before (if possible) and did next.  Did they take an action or did they bounce? What action did they take? Does it make sense based on the goals of the page?

Content should be built around intent and action
Let’s say you have a book about crafting with cat hair and you want to sell this book. Yes, this is really a thing.  What you need is a page that meets the user intent. The data you want to look at is the data that tells you  people are coming to this narrowly focused page and converting. Specifically, the question you need to answer is whether people accessing your page built around ordering a book on making cat toys with cat hair are taking an action (ordering the book).  To help with this, the page should be created to work much like the amazon.com order page for this item with very few options to do anything that is not directly related to learning more about and buying the book. Narrowly focusing the page meets the needs of your user, fulfills the requirements for optimizing for Hummingbird, and is going to make your metrics much easier now that keywords are no longer available.

Amazon screenshot cat crafts

Amazon screenshot focused cat crafting book sale page

How NOT to build a web page to sell a book about cat hair crafts

Because we could always check the keywords after the fact to see what was working, many digital marketers created pages that tried to do everything for everyone. Those days are over.  You should not have a page that describe why you like cats, how to sew, how to teach your cat to sew, etc.    A page like this is not going to work in Google because there is no specific user value, much less a call to action. This method of digital marketing is OVER:

Unfocused cat hair crafting book sale page

Unfocused cat hair crafting book sale page


Useful data for focused pages

Here is what your data on cat hair crafting book pages (or any goal-focused page) should look like:

  1. Visits/Visitors/etc segmented by channel.  Will tell you what is generally working.
  2. Top referrers tied to action taken.  The closest you can get to seeing user intent.
  3. Bounce/ Time spent on page/other related metric. Do people bounce every time they come from Organic search? If so, your page is not meeting the user intent.
  4. Conversion  and other action data. Do people take an action when they come to the page?  What action?  If this is the page that should compell the user to buy the book, then your key action is the clickthru to the page to buy the book.  If people are clicking any other link, this page is not working.

Still addicted to keywords?
If you cannot see giving up keywords yet and need to slowly wean yourself (or your colleagues/execs) off keywords here are some suggestions:

  • Paid data –You can use paid data to help inform your organic keyword ideas by using the Keyword planning tool or, if you are running Adwords, use your matched search query data .
  • Google Webmaster Tools data (GWT)  –You can also use your Google Webmaster Search Queries> Top queries report to see the aggregated view of the top keywords for ctr and visibility along with the aggregated view of the Top pages for those keywords.  The key word (ha!) here is aggregated.  This is not your old keyword data back from the dead, this is just a little extra information about Google results related to your site.
  • Legacy data –If your site has been around for a while and your offerings/products/topics have not changed much in the last 4 years, you can use your keywords from 2010 as a base for your work on user intent. But any keyword data from October 2011 or beyond will not be usable.
  • Google Trends — This may be the most useful tool for keywords because it requires you to already know what your content is about but gives you ideas of how the interest for that keyword is changing (or not) over time.

But really, keywords are dead. And it is a big shift to make from a keyword-focused strategy to a content-focused strategy, but Google has changed the rules of the game and we don’t have much of a choice. It’s time to embrace that change and refocus our efforts on refocusing our pages.  It will pay off in the long-term for our marketing goals and the goals of our visitors.

is encryption stripping out referrer data?

Evil 1990s kitten says check the Mac Plus for your lost data

Evil 1990s kitten says check the Mac Plus for your lost data

Is encryption causing you to lose search referrer info in your Web metrics?

The short answer is yes, but it has always been hard to match all referrals back to specific URLs. And for once, the problem is not just Google.

Encryption is already a  problem

As most everyone who looks at data knows, analytics programs are definitely are not capturing all keyword data, since Google began encrypting keywords in 2011 . That encrypted keyword data is what is being reported as “(not provided)” in Google Analytics reporting. But Google has not encrypted data to the point where you cannot see the URL, only the keyword is being stripped out.

You may be familiar with a URL referrer called “None” or “Direct” or “No referrer” in your data. This referrer could come from apps, offline, and non-browser referrals to your site from anywhere (including, but not limited to Google properties). It is not the keyword encryption from Google and it is not a flaw in any other metrics software that is causing “None” referrers. The “None” referrer just isn’t recordable as a URL, because most Web metrics tools only track browser referrals.

However, there is another issue with encryption that may cause you to lose URL referral data. And as encryption becomes more broadly used you may start seeing more “None” referrals in your metrics. The way it works is when you are on a secure site (https) and then click on a link that is not secure (http) and the URL data is lost in the transfer from https to http.

Google has a solution for SSL search, but mobile and Safari searches may still be stripping out URL info
When Google decided to encrypt all signed in searches they fixed that particular problem with the https to http transition with Google SSL search because they really didn’t want us thinking that Google stopped sending any traffic to us, they just didn’t want to share the keywords with us.

But the problem still exists and complete loss of referrer info is a problem in some mobile searches using Google and Safari browsers. More info on this can be found at SEL: Safari Shifts To Google Secure Search in iOS 6, Causing Search Referrer Data To Disappear

So that is real problem — a loss of referrer URL data because of encryption — it is not a Google search problem though, except possibly for mobile and Safari. The unfortunate thing is I have always seen “None” referrer data in all the metrics I have monitored over the years and it is impossible to know how big a problem it could be when the source of None referrers is not trackable.  But if you start seeing an unexplainable increase in “None” referrers you may want to investigate whether your site or your audience behavior is changing in such a way that encryption could be a factor.

Manual Rank Checking in the era of Google Personalization

I’ve never been a proponent of manual rank checking (I’ve also called this “canned ranking checks”) as a reliable KPI for Search Marketing success, but with the “upgrades” to Google results in the form of personalization, my recommendation is that all manual rank checking be discontinued.

As Conrad Saam noted in SearchEngineLand last month in Excuse Me While I Have A Ranking Report Rant , ranking reports “convey progress while hiding failure. They distract from business goals and promote the misallocation of resources.” Personalization is one of the major factors as to why ranking reports are essentially a waste of time.

At SEMPDX’s SearchFest 2011 last week, Chris Sherman stated that personalization of results in Google is now the default. There is no way to completely turn off personalization and 60% of the results in SERPs are based on your secret Google “profile” based on your location, your search history, etc. You can’t get a “clean” ranking report anymore, so I recommend retiring this KPI.

Not showing up in SERPs is always a concern, but unless you can exactly recreate the experience of your potential audience in your own search results, your real and actionable data is in referrals to your site from Google. This is basic SEO, the goal of Search marketing is not #1 results in the SERPs; the goal is driving traffic from search engines to your site and gaining an audience of users with optimized content relevant to their searches.