You are here: Home / Archives for Bill Hunt

Why Do You Still Use Separate Paid and Organic Search Landing Pages?

July 6, 2012 by

Big companies tend to do things in big ways and one of the biggest ways you can do search marketing is to have separate landing pages for paid and organic search. On the surface it makes sense. Paid and organic landing pages are designed to do different things. But what if we look below the surface?

Below the surface, we start to see something very different. We start to see something changing over time from the old cozy days when separating landing pages did seem to make sense. Let’s go back to remember why we separated paid and organic landing pages:

  • Separate pages allows separate optimization. Why constrain your paid search landing pages with all the tweaks you must put your organic landing pages through? If you can get some higher conversions out of the paid pages
  • Different internal teams work on paid and organic. In large companies, it’s common for the SEO and PPC people to be completely separate teams–sometimes even in very separate groups, such as the SEO people in IT and the PPC people in marketing. Why force them to work together when they can move faster separately?
  • Different external teams do, too. As organic and paid search have become more specialized, it’s common for search consultancies or agencies to have separate teams, too. Sometimes large companies even give their SEO work to one agency and their PPC work to another. Why make them work together when you can be more flexible separately?

These were all good reasons a few years ago, and they aren’t dumb reasons even now. But we’ve seen changes that point to cooperation being fruitful between paid and organic landing pages.

As the years go by, PPC landing pages are more likely to benefit from some of the same work that SEO landing pages have always required. The rise of Quality Score as a PPC ranking factor means that landing pages that are SEO-optimized are far more likely to already have high PPC quality scores, without the PPC team doing and work at all.

And the benefits go both ways. How many SEO types pore over their landing page conversion rates? Not enough. They are too often happy just looking at rankings and traffic without focusing on ringing the cash register. But PPC people have no such luxury–they obsess over conversion rates, constantly optimizing landing pages and the subsequent buying path to move the needle 0.1%.

So what would happen if you forced the landing pages together? Now, it’s understood that there might be many PPC landing pages for very deep keywords that can’t be squeezed into the regular site navigation and must be ignored for SEO purposes. But what of the rest? Working together might be a challenge, but optimizing at least the top pages together might result in higher organic search conversion rates and higher PPC quality scores, plus a reduction in costs by eliminating duplicate landing pages. Working together won’t be nirvana either, but isn’t it at least worth a try?

Filed Under: PLP Monitoring

Are You Still Treating SEO and PPC Keyword Research as Silos?

June 29, 2012 by

Large companies do something that small companies often can’t do—they specialize. But those specialties can sometimes work against them, because they attack problems in such separate ways that they never find the advantages that accrue from working together. SEO and PPC are often handled by different teams in large enterprises, which makes sense in some ways. One area where it makes little sense, however, is keyword research.

You might have many reasons for having separate teams for SEO and PPC, but this silo approach has implications for keyword research that should trouble you. Although it is true that keywords that work for paid search might not work for organic, and vice versa, a large piece of keyword research has nothing to do with paid and organic—it is about market segmentation and about searcher intent.

Some of you might be objecting at this point. Yes, it’s true that there is a limit to the number of keywords that you can optimize with organic search. At a certain point, you can’t keep adding pages to the navigation of your Web site. And you need to make decisions about which synonyms you want to emphasize. And, yes, you need to make hard decisions about which long tail keywords you need to leave behind in SEO. But all of the rest of your keywords can be managed jointly by the paid and organic search teams.

The most advanced search marketing programs recognize that keyword research should be a joint effort between the SEO and PPC teams. Most of those keywords should be shared between both teams, with an additional set of deeper keywords that should be part of paid search campaigns only. Unfortunately, few companies do it that way.  Our own survey of Keyword Management best practices shows only one percent of companies have looked at paid and organic search together, which is exactly what you must do for keyword research.

While specialization is needed, especially in large companies, too much specialization in keyword research is counterproductive. In search, keywords are your market segments, and your prospects and customers signal what they are interested in based on the words that they use.

Take an example of a keyword that converts very well in your paid campaign but badly for organic search. With a silo approach, your paid team will pour resources into that word while your organic team might conclude that it should not be focused on that keyword—which likely be completely wrong. Also, you could be spending a lot of money on clicks for paid search since they are important but not even be on the list for the SEO team.

The searchers using that keyword are the same people for your paid and organic campaign. If the paid search ad works, it is worth some time to explore how to improve the organic performance to achieve the magical 1 + 1 = 3 that all the research shows is possible.

When your PPC and SEO teams collaborate on keyword research, you uncover these opportunities far more often than when they are left to their own devices. There are certainly places for specialization in digital marketing, but search keyword research isn’t one of them.

We have taken a lot of the thinking and Excel wizardry off the table by integrating these common questions into our Keyword Opportunity Models.  If you are ready for more information or a demo contact us to get started.

Filed Under: Co-Optimization

5 Reasons Search Keyword Data is a Mess

May 21, 2012 by

This post should go on my personal site since it is more of a rant than an educational post. Over the past few weeks we have ran number of pilots with the keyword management tool. During this process I have had my believe validated tie and time again since the data we receive is a complete mess.

This chaos of data is one of the key reasons have had resistance to this tool by some of the pilot participants. These reasons essentially sum up as “our data is a mess and we are too tired, frustrated and in some specific cases, actually too scared to try and fix it.”

1. Siloed Paid and Organic Activities (and data)

We expect this since only a few large companies have actually integrated their paid and organic teams into a central team and based on our survey, only about 1% of companies have ever integrated paid and organic data. These are the good “1 percenters” the companies that are realizing the value of the collaboration of paid and organic. My previous post on “Search Marketing Data gathering” and illustrates how bad the situation is and why a tool like ours can help.

But even worse are companies where they are physically or geographically separated by business siloes. I met with companies that still have no central relationship for search. Paid search is in advertising or media and SEO is in marketing or IT and it is a shame that they often never talk let alone share data.

2. Agency IP or Process

Agencies trying to protect their process or IP’s In once case the paid agency refused to share any data with the organic agency since thy felt the organic agency would gain valuable IP on their proprietary way of managing paid search. In this specific situation, “valuable IP” was code for how poorly it was being managed. By using my tool they were able to share the performance data and see the overlap and opportunities that would have never been considered under the previous close of privacy.

Another challenge is the agencies often shield the client from certain data. Sometimes hiding bad data in the good (showing the top 10 best performers but hiding the other million underperforming keywords. Other times it is more of a good thing where they are trying to keep the client out of the weeds by giving them mountains of data to sift though.

3. No way to effectively manage the data

Our survey shows that many used Excel to manage their keywords. While a great tool, Excel is inefficient for managing large-scale keyword data for the average person. Most people are on laptops that are lightweight and don’t have the CPU nor memory to manage this data. Our team all work on Mac Pro’s with 32 to 64 gig of memory that allows us to manage this data.

For example, one of our mid-sized clients just hit the 500,000 keyword mark and wanted to sort through the words off-line so we did an export for them. This file with their various classifications (12 per keyword) and latest period of paid and organic data exported as a 750 MB file. Trying to sort and create pivot tables is a challenge with files this large. They quickly gave up trying to do this online and leverage the segmentation functionality of our application to review their data.

We also run into version control and back up problems in Excel that can result in a lot of wasted time. Recently a client came to us after spending a few weeks on keyword research and organization only to have it lost when a colleague saved the original file into the Dropbox folder overwriting the updated version – yes they should have renamed the file once they made the change.

4. Our Web Analytics tools are too complex

This is typically the biggest challenge – it is just too hard to get a large volume of data from some of these applications. It is interesting and sad how complex some companies have made their data analysis. Managing enterprise analytics data is a massive problem – collecting it across hundreds of servers and then masking it up is no easy task.

5. We like it this way

This is the scariest of all since it will not allow for change. I find that in some cases on the client side they actually want to be in the dark and often not want to know the truth that lies in their data. We have had a few companies abandon pilots since the data scared them. One company had a negative ROI on nearly 30,000 keywords amounting to nearly $180k – how is that possible with all of the advanced bid and ROI management tools available? Nevertheless it was the reality and they did not want senior management to know about it.

Whatever the reason your search data is a mess it does not have to be that way. Contact us for a demo of our Keyword Manage Suite and see how easy it can be to go from chaos to opportunity.

Filed Under: Keyword Management

Managing your High Rank Low Click Keywords

May 20, 2012 by

Back in October we announced a variation of this function and how to “Maximize your Top Ranking Keyword Performance” since then we have enhanced this functionality significantly to make it easier to identify opportunities and show the missed opportunity.

The key factor of this analysis is these words are already ranking well – in position 1, 2 or 3 so the hard work is done. Once identified we just need to determine why our highly relevant page to the topic is not getting clicked by searchers. Using this sample account we can illustrate the value of this sort of analysis:

In this example for the phrase “car rental” there is over 28,000 searches a week for this phrase and they are currently ranking #1 so we expect them to be able to get at least 5% of the clicks. Note, this demo site currently has a 34% click rate on non-branded #1 ranking terms so 5% is quite conservative. Unfortunately, they only received 164 clicks or .58% share of the clicks missing out on an expected 1,242 that they should have received. Given their current revenue yield, had they received those visits, they would have increased their weekly revenue by nearly $10k. These are the types of quick fixes that management likes to see since they have a direct impact to the bottom line.

So what do we do with this knowledge? The course of action in this case is to look at their listing in Google’s search results and see if it is as compelling as the other top ranking keywords. If the snippet is bad adjust the page copy and meta description to improve it. Sometimes the wrong page is ranking such as a PDF or a support page and the description from that page does not encourage clicks.

We might also look at our paid search data to see if we are getting clicks in paid search – we can click the keyword and go into keyword details.

Based on our integration of paid and organic data in our application, we can see they have the keyword active both in broad and exact match campaigns. The paid is not doing much better getting only 35 clicks (about 1/3) of those they received from organic. We can therefore assume, our PPC message is not very compelling either resulting in a negative ROI in paid search.

You should repeat this process looking at least at the top 10 keywords in the default view to help you identify some quick fixes that can bring some quick traffic increases. If you would like to identify your underperforming organic keywords contact us for a demo.

Filed Under: Keyword Management

Search Marketing Data Gathering Challenges

May 14, 2012 by

Over the past few months we have been doing a number of pilots with our Keyword Management Suite and the biggest challenge has been to gather data. We have actually spent the better part of the last four builds to enhance the data import, cleaning and filtering functions to deal with the chaos of search data.

We expected these problems early on since only a few large companies have actually integrated their paid and organic teams into a central team. These are the people that are realizing the value of the collaboration of paid and organic. The following will illustrate how bad the situation is and why a tool like ours can help. The following is an example of the process to get the data from a single company that both the company and their agencies have won awards for their search marketing programs.

April 2nd kicked off a pilot with a large company that spends double-digit millions on paid search. They realized they were not getting the collaborative benefits from their search program and wanted to understand how fractured it really is.

April 2nd received the list of the companies “most important keywords” from Global Search Marketing Manager – 1,120 keywords

April 4th received list of “most important keywords” from SEO team – 98 keywords

April 6th received export from their Enterprise SEO Management tool that is tracking performance and auditing the tool – 298 keywords

April 18th received paid search data from PPC Management tool – 187,664 keywords (note 2 weeks to get paid data)

April 22nd received Omniture organic performance export – 397,832 keywords drove visits to the site from organic search for the week we are testing. Note 3 weeks to get data from Omniture which prides itself on how easy it is to pull data.

Interesting facts:

A 1,012 word difference between the list the Search Manager says is important and what is being tracked in SEO tools

Only .029% of the words used in Paid Search are important (all words equally budgeted)

Only 1 of the Top 20 keywords with highest Cost Per Click was on important list, active in SEO or ranked on page 1

4,887 of the keywords in the PPC program are currently ranking in the top 3 positions of Google which makes them prime candidates to test co-optimization to identify collaboration or cannibalization.

The collective team was surprised at some of these number and have since been working together to get them reconciled and start benefiting from the collaborative value of paid and organic search.

How big of a mess is your data? Ask for a trial of our tool to find out.

Filed Under: Keyword Management

SES Interview – Advanced Keyword Modeling

April 4, 2012 by

Interview with Tom McCarthy from LinkDex reporting for WebMaster Radio on keyword modeling and our Keyword Management Platform.  Have a listen

 

Filed Under: General

Current State of Keyword Management

February 5, 2012 by

 

They current state of Keyword Data Management has not improved much over the years.   Back Azimuth has recently completed a survey of various sized companies about how they manage their greatest search marketing asset – keywords.   Some of the results are what you would expect but others are pretty interesting.  If you want to add your answers to our Keyword Management Survey go here https://www.surveymonkey.com/s/2658JTX

Sample Size:  This survey has responses from 112 different respondents from October 2011 to Present.

Surveyed Companies:  They range from small eCommerce companies to a lot of Fortune 100 multinationals.

Data Presentation:  While using pie charts and percents is not the most accurate way to represent this data as “research” I did it this way since this research is pretty revealing of the sad state of keyword management and did not want to give away too much data that would motivate potential competitors.

 

 

Question 1:  How Many Keywords are you Currently Managing?

The majority of respondents are managing more than 1,000 keywords with 54.5% managing between 1,000 and 100,000 and 27.3% managing more than 1 million keywords.

Question 2:  What Keeps you from Managing More keywords?

While not expected but not really surprising, the biggest reasons people don’t manage more keywords is that they don’t have the budgets.

  • 63.6%  Don’t have the budget to manage more keywords
  • 45.5%  Don’t have the time to manage more keywords
  • 36.4%  Don’t have a system to organize and manage keywords
  • 18.2%  Don’t have the time to find more keywords
  • 18.2%  Found all their keywords and don’t need any more
  • 9.1%  Current keyword management solution is complex and hard to use
  • 9.1%  Limited by my vendor’s capabilities

 

 Question 3:  If it were easier to manage more keywords and performance how many would you like to manage?

Again, a little surprising but we need to match up this response to company size since that may be the limiting factor.  Smaller companies may max out at 100 keywords that represent their offer to companies like WalMart or Travelocity who need over 1 million phrases just to capture their massive product inventories.

 Question 4:  What are the most time consuming and challenging aspects of keyword management?

Given that only a few respondents using anything other than Excel to manage their words there is no surprise that “organizing words” was the biggest challenge.  Closely following was reviewing keywords in paid and organic programs.  Most programs are managed separately from each other and it is rare that I find a company that even knows which words are in the other program.   Two other responses jumped out, not knowing the business or performance value of keywords was important to search marketers.  Without knowing the value and incremental performance it is hard to make a business case for more funding for search marketing activities.

  • 63.6%  Organizing keywords
  • 45.5%  Reviewing keyword currently used in paid and organic search
  • 36.4%  Assigning keyword ownership across business units
  • 36.4%  Finding and adding new keywords
  • 36.4%  Not knowing the business value of keywords
  • 36.4%  Not knowing the performance potential of keywords

 

 Question 5:  Do you currently aggregate your Paid and Organic Search Data into a single view for analysis?

Another surprise answer.  I have conducted and informal surveys at Search conferences over the past few years and the most responses for integration was 5 out of over 100 attendees so this number was a but surprising.  I am encourages that nearly 50% that are not currently aggregating and analyzing are interested in doing so.

 

  • 46.2%   Yes, currently aggregate for analysis
  • 46.2%  No, do not currently aggregate but would like to
  •    7.7%  No, do not currently aggregate for analysis

 Question 6:  Do you analyze the collaboration or cannibalization impact of Paid and Organic Search data?

Following Question 5, that less than 50% aggregate the data it follows the same number of respondents actually do the analysis.

 

  • 46.2%   Currently analyze the collaboration or cannibalization of paid and organic search
  • 38.5%  Do not currently analyze the collaboration or cannibalization of paid and organic search – but want to
  • 15.4%  Do not currently analyze the collaboration or cannibalization of paid and organic search

 

Question 7:  Do you currently segment your keyword phrases into categories or segments?

Again, a little surprising but we need to match up this response to company size since that may be the limiting factor.  Smaller companies may max out at 100k worth of products where someone like WalMart or Travelocity needs over 1 million just to capture their massive inventories.

 

  • 100%   Segment keywords by brand and non-brand
  • 54.5%  Segment keywords by purchase or buy cycles
  • 54.5%  Segment keywords by categories or other logical segments
  • 45.5%  Segment keywords by searcher intent
  • 36.4%  Segment keywords by business units
  • 18.2%  Segment keywords by audience types or persona
  • 0%       Segment keywords into question formats  (i.e. How to clean marble counters?)

 

 

Question 8:  What should a keyword management tool have that would impress your boss to increase your budget?

This question was to help me understand the features that a user would want in a keyword management tool to help them demonstrate the value of Search Marketing to their boss.   The responses were not what I expected and all over the place.

 

  • 83.3%   Identify issues and opportunities more quickly
  • 66.7%   Lower or better control over PPC spend
  • 58.3%   Better coordination between business units and teams
  • 33.3%   Quicker access to keyword lists and associated data
  • 16.7%   Other

The other responses were the ability to tie to Salesforce to demonstrate lead value and for an agency to demonstrate their value to the business.

 

Filed Under: Keyword Management

Benefits of Preferred Landing Page Management

December 29, 2011 by

Lets start by defining what a PLP is… PLP is an acronym for “Preferred Landing Page” which is the URL which represents the page that you believe is the best page to be presented to a searcher based on the implied intent.  Well duh, isn’t that why we optimize specific pages?    The reality is that while we do optimize specific pages to rank better the reality is most ranking tools only tell you that you have “a” page ranking but not if it is the best page or the page you have been slaving over.  In most cases the agency nor the client don’t really care as long as they can show “a page” is ranking well.

I have been advocating the use of PLP Monitoring or PLP Management as some call it now, since I started managing the SEO Program at IBM in 1998. I went into detail about the process at PubCon Vegas in 2004 and was greeted with skepticism.   In the session, I has shown where we identified a huge problem at IBM where very often IBM Research pages would often out rank our brand and product pages for a product or service.    The research pages often beat us since their pages were keyword rich, had links and did not contain Flash or other creative elements that can negatively impact rankings.

At first, it did not really matter who ranked just that “we ranked.”  Once we broke out the reporting by Business Unit rather than a master IBM report it started to matter which Business Unit and more imprtantly which page was ranking since we now held the BU Manager accountable for both rank and traffic increases.

The solution as simple, to identify the page with the best chance of converting the visitor to each keyword by Business Unit and mapping it to the keyword phrase.  While it seemed simple on paper it was painful to implement.  This let to a very robust Keyword Arbitration and Owner Management program around keywords across business units we can discuss in more detail later.    We needed to understand the overlaps and the current state so we developed a process where we would run simple rank reports them pull them into a Microsoft Access Database to do a match to see if the “right page” was the page actually ranking.  If it was, then it would pass and if not it would fail.

For those that failed, we would evaluate the higher ranking page and decide if it were the better page even if it was for a different business unit.  We had a challenge when RFID technology became hot since we had 26 business areas in
IBM that all wanted to rank for RFID and of course since it is new technology IBM Research often raked the best which did not lead to leads or revenue.   Our approach was to develop a cluster analysis of the phrase RFID and then parsed out words to the approriate business units.   RFID hardware words went to hardware, RFID software went to software and the generic category phrase RFID was assigned to a cross-BU content page where we could monitor the clicks to the different interest areas and could update the message and banners on the pages based on what the searcher really wanted when searching on a generic category phrase.

Leveraging Back Azimuth’s VOCDMS suite to Manage Preferred Landing Pages

So, how can you start reaping the benefits of managing PLP’s at scale?  In our VOCDMS Keyword Management Solution we take it to the next level by looking at PLP Management with 4 different types of analysis to help you identify opportunities and prioritize our efforts.

Analysis 1:  No PLP in Top 10

This analysis shows us which of our most important keywords do not have their optimal page ranking.  We focus on keywords that are in the Tier 1 classification and find all words were the PLP is not currently ranking in the top 10 positions.

Success Example:

A large travel company was ranking #1 for the keyword phrase ‘Easter Holiday Deals” and even had a 50% click through to the organic page.   Unfortunately, there was a 98% bounce rate and no conversions.  We detected this and found it was last years Easter holiday deals page and it contained no offers.  After a quick 301 redirect the current page was ranking #1 and within 10 days they had received $60,000 in bookings that would have been lost had they not detected the less than optimal page ranking.

Analysis 2:  PLP Ranking 11 – 15

This analysis we focus on keywords that are in the Tier 1 classification and find all words were the PLP is is currently ranking in the 11 to 15 potions of the search engine.  This is the top of page 2.   The idea is these are low hanging fruit that often needs minor adjustments to get them to page 1.  This is a great report to use to find quick opportunities for improvement.

Success Example:

A multinational software manufacture pulled this report and put their SEO team to work on a few of these pages and wer able to get 12 of them onto the first page just by adjusting title tags and getting some links from their partner blogs.  These pages generated over 35,000 in incremental traffic over the next quarter and and a significant increase in sales for these products.  These were words that were not on the “to do” list of the SEO team and by identifying them and quickly fixing them were able to show short term gains to their management team which bought them the trust to do this on a large scale.

Analysis 3: Non-PLP Ranking Top 10

This report focuses on identifying keywords where some page other than the PLP is ranking in the Top 10 positions.      This is a great report to use to find pages that are ranking that should not be ranking or often better PLP’s.

Success Example:

A great example of this report happended in two parts.   One thing I learned at IBM was that long after we stopped marketing and selling products our customers still use them.  For a large computer manufacturer I went to Wikipedia and grabbed a list of about 50 products they had previously sold and added them to the tool.   We used the Missed Opportunity Model to identify that there were over 1.2 million searches a month for these 50 product names.   We identified there were 3 reasons for search ing on the prioduct name.  The first, they needed support, second they wanted parts like power cords or docking stations and third, they wanted to upgrade to a new version of this model.  Based on the opportunity analysis they created 50 pages  (one for each product name) which offered the three options and loaded them.  We added them as PLP’s in the tool and were able to monitor their performance.  In about 2 weeks these pages replaced their Support PDF’s and the page ranking and traffic started to increase as well as revenue.  At the end of 6 months they had generated about $400k in incremental revenue from parts, accessories and upgrades due to these pages.   These words were added to paid seaerch as well as a team set up to develop pages for every product that was retired in teh past 5 years.

Analysis 4:  PLP in Top 10

Th analysis is the opposite of Analysis 1 where this time we should which of the keywords are in the top 10.  This analysis shows us which of our most important keywords which do their have their optimal page ranking but can also see if there is another page ranking higher.

No success examples for this report.  This is just the “warm fuzzy” report that show what is working and included just because users wanted to be able to show what was working.

Opportunity Modeling and Business Case Development

We are working on a new set of filters in the tool now to monitor improvement once the PLP ranks as well as models around Predictive SEO.  We already can show the “Optimization Impact Value” which is if we can get into the first page of more importantly the top 5 positions what is the expected lift in traffic and conversions when that happens.   We are monitoring the traffic and conversions from top ranked pages with and without PLP’s being the ranking pages.   We are finding that there is often a 30% increase in clicks and a 50+% increase in conversions when the PLP is the page that is ranking the highest and ultimately on the first page.  Since we have data for all words and pages we are able to extrapolate the success of single pages.

How to get Started using Back Azimuth’s PLP Management tool

It is simple to get started, we just need a list of keywords and the pages you want to rank well for these words.  If you don’t have a list we can generate a report that replicates the “keyword site:yourdomian.com” query syntax
for Google to return the list of pages Google things is the most relevant.   From this list you can choose which one to be the PLP.  You can always change it at any time.

Contact the Back Azimuth team today to schedule a demo and be on your way to search greatness.

Filed Under: PLP Monitoring

Maximizing your Top Ranking Keyword Performance

October 30, 2011 by

How many of your top ranked keywords are under performing?  If you are like most companies, your not tracking it so you may not know.

The goal of most SEO’s is to achieve top rankings and high conversions.  Unfortunately, once we get the top rankings little is done to maximize it or even make sure the correct page is ranking.

Many SEO’s have stopped doing ranking reports due to personalization or the inability to show any real performance improvement from it.  I can understand both of these which is why I developed a report in the application called “Top Ranking keywords that Suck”  – well, that is the working title until we can figure out what we want to call it.    We originally started with the tongue twister you see below “ Top 5 Ranking Keywords with Less than 5% share of clicks”

When I have done SEO for clients I wanted to focus on the immediate low hanging fruit that drives incremental revenue.  I have talked at conferences in the past where I found keywords that were ranking and simply fixing the snipped resulted in not only increased traffic and revenue but on one case a PPC click cost reduction of $26k in one month.

The following is the use case that I created for this function in the application.

Tool Use Case: Identify keywords that are currently ranking well (top 3 or 5 positions) but getting a low share of clicks (less than 5% of the total search volume or less than the average click rate for its position).   This would allow the Search Marketer to quickly identify underperforming words to review snippets for relevance as well as messages and offers.

In our first generation we simply called this out on the dashboard and allowed you to click in and see the details and do sorts on the data.  The default view is performance sorted by highest CPC but you can sort as below – by highest Google demand.

 

The new version of this report we are calling, as mentioned above, “Top Ranked Keywords that Suck” – in this example, this client has a fairly experienced and robust SEO team.  In our analysis across a few beta clients we are finding similar patterns:

In this example, the client has 1.4 million keywords in the database.  Of those keywords, 101 currently rank in the top 3 positions of Google but are getting less than a 5% click rate.  If they could increase the clicks from Google for keywords to 5% that would be in increase of 103,678 visits a week.

[Why 5% - we use a simple rationale for these.  10 organic and 10 paid = 20 chances of being clicked.  If a searcher clicks just 1 listing we have a 5% chance of being clicked.  No complex formula of paid and organic, national brand, message, universal search results – just simple probability. ]

Actionability of the Data

I have to say that whenever I do a prototype for a new function I am haunted by the voice of Avinash Kaushik in “don’t just Puke Data – offer Actionable Insights” – based on his “Actionable Dashboards” presentations.  this was no different.  How can someone process 101 results? We are still working on ways to just mine the uber nuggets of data – see “Crowdsourcing Questions” below.

In our tool we can sort these by Google Search Demand, Revenue Per Visit, or Conversion Rate or any other number to identify which of the 101 that we want to fix first.  Typically the user is choosing those that have the highest demand and the lowest click rate.   The typical user of the tool is trying to identify 5 or 10 keywords a week to try to improve.

Action 1:  Sort by highest Google Search Demand – this helps us find the keywords with the greatest opportunity overall.

Action 2:  Sort by revenue or Revenue Per Visit (RPV) to identify keywords that generate the most revenue per visit.  This is a good variable since our goal of this look is to find words not getting the share of visits.

The following are key insights when we do this:

.acme is a brand misspelling keyword and there are clearly not 695k searches for it but Google associates that to the brand name.  We are now dealing with this by filtering out keywords with keyword type = Brand Misspelling

Cheap Tickets – in our demo case, they are ranking #2 in Google and only getting 6,492 clicks or 2.6% share.  A 5% share would be 12,500 or 52% of our goal opportunity.  Even a 1% increase in traffic would

Note: In this case if we average the actual click rate of this client for keywords with an organic position of 1, 2 or 3 without paid search influence it is 4.32% so not far off the 5%.  We can assume the actual click rate would be higher if we could count those 103k clicks we should be getting at 5%

In the next generation of the application, for each keyword phrase we will return the snippets for the top 5 positions to review them and apply a “reason for non-performance”  In doing beta tests we started to understand why some of the words are ranking and others are not.

Mitigating Factors

As with everything in search, there are never absolutes.  In working through the unique percentages of clicks for a company like IBM we may need to have more robust scoring moving forward.   The following were interesting findings from an exercise with Lee Moore, Global Search & Syndication Manager from IBM.  He wants the ability to flag those not performing for one of the following reasons and re-weight the variables.

  1. Context of the listing – for example, There are x searchers for SOA and IBM, while ranking #4 is only getting about 1% of those searchers.  Looking closing we can assume it is due to “context.”
  • #1 ranking listing is “Society of Actuaries” which has been around a while
  • #2 is Wikipedia for the technical term Service Oriented Architecture
  • #3 is the popular TV show “Sons of Anarchy” which is abbreviated SOA and most of the cast and followers in social media refer to the show that way.
  • #4 IBM Service Oriented Architecture (SOA) solution
  • #5 is Oracle which is a competitive placement.

 

So we can assume a large share of  the “Searchers Intent” and clicks are related to “Actuaries” and Sons of Anarchy and not expected to go to IBM.  Since this is also generic, a site like Wikipedia may give a better “What is SOA” than IBM or Oracle if they are looking to the technical variation of the acronym.

  1. Message – Not related to this IBM case but if the query and “searchers intent” is price based and the price is out of line this will cause a reduction in clicks.  If the message is branding and they want to buy that may reduce clicks.
  2. Paid Search – If there is or is not a paid search ad for the company may impact click rate.  Many studies show that both a paid and organic lead to increased brand awareness or assumption they re bigger resulting in a click on one or both.
  3. Branded Keyword – branded keywords tend to get clicked more than non-branded.  For example, Dell Laptops vs. Laptop Computers.  In this case listing 1 and 3 are branded for “other” brands and not for Service Oriented Architecture.  It can be a battle of brands between IBM and Oracle.
  4. Gibberish Snippet – already addressed but added in for completeness, if the snippet is bad for IBM or any listing the potential of clicks drops exponentially.

Base on these factors, we need to factor in some sort of automatic or manual override of the percentage based on an analysis of the listing so that it is not counted against the company going forward.  We can also use the same analysis to help identify the need for and priority of corrective action.

Crowdsource Questions:  Post any suggestions in the comments section below:

  1. Should we restrict the rank to top 3 or top 5 positions?
  2. Is a search demand of 500 the right cap? If we lower it, how do we filter out those in mass?
  3. Since these are ranked high this will draw clicks from paid search.  Due to bad snippets there should be increased clicks on paid.  Not sure how to calculate yet but am thinking of a “Cost of Not Clicking calculation that would show that the delta in the 5% is made up with paid clicks.
  4. Should we offer facets to allow user to choose which position, search demand and which click rate?
  5. Should these filter variables be user editable in an admin area?
Filed Under: Search Performance

89% Uplift from Paid Search Clicks

June 8, 2007 by

The above is the headline of a nice shinny object that Google is dangling in front of marketers. One that is being used out of context and I am sure has resulted in significant money being pissed away in paid search. Now, I am not against paid search in any way – I think it is a great tool and works even better when it is in collaboration with organic listings. That is what I advocate – co-optimization. How do we make them work better together.

Over the past few weeks I have heard that 89% quote in four countries, at every conference and at least 20 times alone at SMX in New York. So where did it is come from?

This is essentially the findings of a research study released recently by Google employees titled “Incremental Clicks Impact of Search Advertising”  that said the following:

A meta-analysis of several hundred of these studies reveals that over 89% of the ads clicks are incremental, in the sense that the visits to the advertiser’s site would not have occurred without the ad campaigns.

Immediately this was translated by the market place as the following headlines in articles and blogs with the first being my favorite “doom and gloom” heading:

  • “Danger – STOP Paid Search Advertising & Lose Up To 89% Of Your Web Traffic!”
  • “89% lift when Paid Search is added to Organic Search”
  • “Studies show search ads drive 89% incremental traffic”
  • “Paid Search delivers 89% more traffic than organic SEO alone”
  • “Google Study: 89% Uplift from Paid Search Clicks”
  • “Google: Search Ads Drive 89% Incremental Traffic”
  • “Google Research Shows Paid Search Ads Get 89% More Traffic Than Organic Search Results”

Google went on to create an “idiot proof” Paid Search is Great video that showed that in some cases 98% of the traffic

For those of you that actually read Google’s study other than the half-assed paraphaseing blogs you might have noticed the “your mileage may vary clause” in the last paragraph of Section 3:

A low value for IAC may occur when the paid and organic results are both similar and in close proximity to each other on the search results page. This increases the likelihood of a user clicking on an organic result as opposed to a paid result.

Close proximity occurs when the ranking of the organic result is high, placing it near the paid results. Organic results triggered by branded search terms tend to have a higher ranking on average and this may lead to a low IAC value.

Matt Van Wagner scared me for a moment with the headline  in  a recent Search Engine Land article Google Study: PPC Ads Do NOT Cannibalize Your Organic Traffic  fortunately Matt was not another Google fanboy and strongly suggested that  people actually test the data.  As I mention on my personal blog, New Venture Announcement – Voice of Consumer Data Management System I have done a few surveys and found only a few people actually combining the data and doing anything with it.

Brad Geddes has been talking about this the longest and a recent post on his blog goes into the mechanics of doing the testing of paid vs. free clicks.  I had already added this specific testing into my tool and it starts to show some very interesting results.

One of the biggest reasons I found as to why people don’t do it is it is too hard to do for most.  In Brad’s post he simplifies it but what if you have a lot of keywords?  This is one of the key elements that I have built into my tool. I have only found a few companies that even know if they are ranking for key paid listing.

Below is a screen capture from my tool that shows that for the 20 most expensive words by Cost Per Click they did NOT rank on even on the first page. In this case, yes, Google’s study holds true – if you have no exposure in organic search then the only exposure you will get is from paid ads.

In this case they are paying $10.00 or more per click, their highest CPC and they are not ranking well.  We can’t even get to a collaboration scenario until we have the organic rankings.   This company was not aware of this problem since they were not looking at the data collectively.  Immediately after learning about this they went to work optimizing the pages to try to get these to rank better.  In a few cases, there were not important and they reduced their average CPC.    This is the opposite reason people use PPC – to make up for the shortcoming of their organic performance. Maybe they can redo the study and show what happens when they have organic rankings.

To help companies understand once they have an organic ranking and a paid search rankings what is happening.  I have built into the application a simple ROI calculator. For your PPC Loyalist and Co-Optimization Haters – yes there is no message context or any other variables other than the fact this word had a negative ROI.

In the example below, we have a keyword that everyone thought was performing acceptably well.  When we actually do some analysis we see that it has a negative ROI and is loosing the company $11,825 dollars in the current month the the organic term was generating $6 million.

To be fair, we can look at a positive ROI example where the paid and the organic have generated a positive ROI.

In this case organic still does out perform PPC but PPC has a positive ROI.  In further tests when this PPC ad was day parted to appear less frequently, Organic did not pick up the additional clicks.  This showed us that in this specific case, paid and organic were collaborative and having paid search resulted in incremental visits and clicks.

There are a few things you need to do and consider when looking at the analysis.

You don’t have to do all of your keywords.  You should decide if they are the brand name, branded product names or if they are general category or specific non-branded words you are looking at.

The tests you want to do are the following:

What happens when we have paid only?  This is a good test to do before you optimize content and do not have an organic position.

What happens when we have organic only? You can day part of pause the paid search for a period.  Most of the times a few days or a week is sufficient.

What happens when we have paid and organic? Once you rank well you can start the comparison.  This will tell you what is happening when they are both together.

We are NOT trying to eliminate paid search for all words.  Only those words there there is not an incremental lift if clicks.  If we turn off paid search and all or most of the clicks and conversions that went to paid increase the organic clicks and conversions then paid is NOT complimentary but cannibalistic.  If the clicks and conversions do not increase we can assume that they are collaborative and simply un-pause the paid ads until you can do a message test.

The point of this is just test it and see what is happening.  If you want to better understand our analysis tools send us an email and we would be glad to give you a demo.

Filed Under: Co-Optimization
Newer Posts»