Archive for the ‘MSN’ Category
Weekly Search Engine Market Share Update
Last week I published information about how the market shares for the search engines have evolved over the last few months. I’m going to attempt to post updates almost every week, offset randomly by other interesting trends such as browser share numbers. This week, I’ve put together a weekly update, as quite a few people emailed me about the evolving trends.
To highlight the trend, I’ll re-post the data from early September.
search engine market shares for search engine usage based on searchers located in the US
| Yahoo | Bing | Others | September 7 | 78.68% | 11.51% | 6.80% | 4.06% |
|---|---|---|---|---|
| September 14 | 78.35% | 11.13% | 6.50% | 4.02% |
| September 21 | 77.43% | 11.35% | 7.11% | 4.12% |
| September 28 | 77.65% | 10.80% | 7.27% | 4.28% |
| October 4 | 77.78% | 10.66% | 7.23% | 4.25% |
| October 12 | 77.78% | 10.66% | 7.21% | 4.36% |
| October 18 | 77.89% | 10.65% | 7.29% | 4.16% |
Again, this data represents search engine click through activity where the people initiating the searches are located in the U.S. At some point I’ll present information on global search activity.
Search Engine Market Share Update October 2009
As mentioned in my previous post about browser usage, I’m going to start posting more data, more often. This week I’m putting out search engine market share numbers. I haven’t posted a U.S. search engine market share update since March 2008, so I’m definitely overdue!
Rather than providing a simple set of monthly numbers, I’ve decided to provide weekly breakdowns using aggregate data from our suite of search marketing tools, which makes it possible you review a few months worth of data at a detailed level. I’ll try to update these every two weeks, with weekly breakdowns for now, to try and identify trends in very granular ways. Interesting high-level, and consistent growth for Bing, overall growth for Yahoo! as demonstrated in this chart. Remember, this is based on clickthrough activity.
Astute observers may note that these numbers differ from other reports. Our data is collected at the web site level, based on actual referrals received. While people may search any of these engines more often, but not click through to the results. We report on the activity we see hitting the network of sites we’re tracking, and we are very careful to report only on the search referral traffic, so this won’t include referral traffic from Yahoo mail or gmail, for example. We’ve also rolled up all the various types of Google search activity into “Google” and not separated out Google Image Search / Blog Search, etc.
U.S. weekly search engine market share breakdowns
As you can see, Yahoo! and Bing have been climbing, slowly but steadily, as a percentage of overall search engine referral activity. For more details, please view the weekly breakdown chart below.
| Yahoo | Bing | Others | ||
|---|---|---|---|---|
| April 5 | 82.74% | 9.49% | 4.99% | 2.77% |
| April 13 | 82.65 | 9.60% | 4.99% | 2.76% |
| April 20 | 82.03% | 9.85% | 5.30% | 2.82% |
| April 27 | 82.20% | 9.62% | 5.41% | 2.76% |
| May 4 | 81.26% | 9.55% | 5.22% | 3.97% |
| May 11 | 80.98% | 9.69% | 5.17% | 4.16% |
| May 18 | 79.94% | 10.45% | 5.32% | 4.29% |
| May 25 | 80.06% | 10.39% | 5.37% | 4.19% |
| June 1 | 79.80% | 10.44% | 5.39% | 4.37% |
| June 8 | 79.48% | 10.69% | 5.53% | 4.30% |
| June 15 | 79.41% | 10.54% | 5.79% | 4.26% |
| June 22 | 79.40% | 10.56% | 5.79% | 4.26% |
| June 29 | 79.52% | 10.37% | 5.62% | 4.48% |
| July 6 | 79.51% | 10.46% | 5.48% | 4.55% |
| July 13 | 79.69% | 10.31% | 5.56% | 4.44% |
| July 20 | 79.61% | 10.30% | 5.62% | 4.48% |
| July 27 | 78.21% | 11.18% | 6.18% | 4.43% |
| August 3 | 78.77% | 10.95% | 6.06% | 4.22% |
| August 10 | 77.92% | 11.54% | 6.36% | 4.18% |
| August 17 | 77.33% | 11.82% | 6.80% | 4.06% |
| August 24 | 78.11% | 11.51% | 6.27% | 4.11% |
| August 31 | 78.62% | 11.27% | 6.27% | 4.11% |
| September 7 | 78.68% | 11.51% | 6.80% | 4.06% |
| September 14 | 78.35% | 11.13% | 6.50% | 4.02% |
| September 21 | 77.43% | 11.35% | 7.11% | 4.12% |
| September 28 | 77.65% | 10.80% | 7.27% | 4.28% |
| October 4 | 77.78% | 10.66% | 7.23% | 4.25% |
| October 12 | 77.78% | 10.66% | 7.21% | 4.36% |
In context of all three engines, here’s a chart of the global search engine referral rate market share numbers which does show interesting trends over the last few months.
Search Engine referral rate activity report
Enquisite collects data from a network of web sites distributed globally. The data used in this reports represents web sites distributed globally, accessed by searchers located in the U.S.
Let me know if you would like a similar report for the UK, Canada or other areas, please.
Microsoft Bing Tracked
Someone asked me yesterday if we could report on Microsoft’s Bing search engine. They were concerned because their present analytics vendors, (plural) were not tracking it properly or at all.
I am happy to confirm that yes, we added bing.com upon its release, and you are able to see it recorded within your Enquisite Optimizer search analytics reports.
I was chatting with someone in Marketing from Bing.com last night at the SMX Advanced event in Seattle, and they were quite interested to find out we were already tracking their results, and I proposed that we publish a note next week, and subsequent follow-on posts about how their marketshare is evolving. She was excited to see the results. So, next week, I’ll start putting together data to post on Bing.com’s evolving marketshare.
Keep watching this space…
Building a New Business Begins
continued from part 3 Starting to Build the Campaign Platform
So, In June 2005, I set about separating Enquisite as an entity from Metamend. This was a crucial and critical step for me, and the company. I needed to establish a firewall between the companies for many reasons, not the least of which was to ensure that competing agencies would never have to fear that their data was accessible to potential competitors. My co-founder at Metamend, Todd Hooge, along with Glenn Convey, a very talented individual in his own right, took over operating that business. It took a few months, but I removed myself from all operational involvement, which was personally challenging. But by December 2005, it was done, and I was out.
By this time, I also had a complete outline for what would become the core elements of Enquisite Campaign, and the next generation of products still to come after it. Over the next few months, I proceeded to break down the functions into manageable pieces we could build as foundational elements for the platform. I received a lot of valuable guidance for each foundational block along the way, and added some key people to the company whom I still depend upon today. Key internal technical counsels for me were our Lead Developer, Rick Morris, and our VP for Technical Operations, Greg Caws. They foresaw many technical hurdles we would face long before we ran into them. Their advice made a dramatic difference every step of the way.
Building the foundation of Campaign was not trivial. Fundamentally, we needed five pillars to support the architecture for the system. We built the various pillars and released them as individual products. So while we’ve received great reviews and feedback for Pro (Now Enquisite™ Optimizer), PPC Assurance (now Enquisite™ Auditor) and our Links Report (within Optimizer), we’ve really looked at these as stepping stones. The last two pillars - the internally-facing “Collector” and a massively scalable, super-fast database - rounded out the foundation.
To build the opportunity analysis and the reporting functions for Campaign, we needed to deal with paid search traffic in a way no one before had. We needed to understand, and value, paid campaigns in relation to organic ones. This perspective, and the requirement to segment out paid search traffic in a new way, led to our PPC Assurance product. This application reveals click accuracy for PPC with amazing precision, enabling advertisers to credibly claim credits from Google, Yahoo! and others and potentially save thousands, even millions of dollars. Equally important, this first step allowed us to understand paid search in a different way from existing solutions.
As a bonus, the work on Enquisite Auditor led to our second product, Enquisite Pro. Earlier this year it was officially recognized by Yahoo!, when they began recommending it via their Traffic Quality Center. We built the early version of Pro - now renamed Optimizer for a purpose - to provide search engine positioning reports, and to segment traffic and campaigns in ways that had not done before. We also pushed the development team to devise a new way to collect data which resulted in the simply named “Enquisite™ Collector.” In case you haven’t noticed, we have very simple and functional product names. In this case, the Collector is an fantastically scalable next-generation data logging system that uses over 26,000 web servers distributed worldwide to collect log file data for our reports.
Continued in part 5….
Starting to Build the Campaign Platform
continued from part 2 The Genesis of the Enquisite Campaign Idea
To make Campaign work technically, I knew we needed at least two things off the bat: (1) we needed to capture some critical information which wasn’t presently available; and (2) due to the absolute scale and magnitude of data, we needed to capture it much more efficiently than present web-based logging systems did. The first conundrum was capturing that critical information, and it was during an animated discussion with my business partner at Metamend, Todd Hooge, that we hit on a means to gather it. This led to Enquisite Optimizer’s (formerly Enquisite Pro) search engine positioning reports.
At first, I didn’t know if the rank reporting process would work the way I needed it to. So we started testing, and I kept dreaming up ideas around a better business model. I didn’t share these ideas with Todd, as we were far away from capturing the data necessary to power a new reporting system. I also suspected that if I had shared these thoughts, he probably would have looked at me like I was nuts (again), and made some comment like “it would also be nice if cars could fly.” Plus, I was still thinking about all the things we would need to build for the “finished” system, including the nuts-and-bolts implementation. Anyhow, Todd helped me tremendously to break down that first, crucial, barrier around capturing the necessary data. It may seem trivial today, but it was a big deal then. Even today it’s still a big deal to people when they get an explanation of the entire data capture system.
Once we worked out the germ of the idea on how to capture the data, we were off. I hired a developer to turn the theory into practice. Simultaneously, I started writing up some of the ideas for patent submissions. By June of 2005, we were ready to try out a basic data capture and reporting system, and I had written a phonebook’s worth of documents to file for patent protection.
When I showed the initial reports to some of my search colleagues, they all said, “When can we have it?” It was at this point that I seriously started considering building out the entire suite for a larger purpose. While at first it was an interesting and meaningful project to help me build out my search marketing firm, based on the initial reaction to the first reports (which now sit within Enquisite Optimizer), I realized we could build a system for everyone, not just ourselves.
Continued in Part 4:
Search Engine Market Shares 2007
So I arrived at Search Engine Strategies New York today, and I was asked by a couple of people about search engine market shares. After pulling out the Ask numbers last week, I had all the data ready to go for the other engines. Remember, this data reflects the search referral data we’re seeing across the entire network of sites that Enquisite is tracking, so thousands of sites’ data contributed to these numbers. When I actually graphed the data, it looked quite interesting.
I had to break the data into two parts. In this first graph we see Yahoo have its customary summer spike, which generally seems to relate to the end of school. During the summer months students spend less time online, but when they go online it’s to fetch mail and the like. During this period, Yahoo! generally goes up in market share, as most students appear to use Yahoo Mail. Normally, we also see Google drop during this period.
What’s interesting is that MSN is slowly but surely gaining traction, and moving up. It’s gone from 2.9% in January 2007 to just over 5% at the end of January 2008. Still small, but almost 100% growth, and anyone in business know’s 100% growth does matter.
Meanwhile however, Yahoo’s actually losing market share, and at a greater rate than MSN’s growing.
Now take a look at what happens when we add Google to the mix.
Google’s actually over 80% of all search referral traffic we’re seeing across our network of sites. In fact, the data I’m looking at for March has Google reaching 83% of all search referrals we’re seeing. This data is culled from well over 250 million referrals in the last year.
So, is search getting more competitive? Not really. Is Microsoft buying Yahoo going to make much of a dent in Google’s lead? Nope. But (as Rand pointed out) if you look at their combined reach in the display ad business that’s a different matter.
Getting Refunds from Yahoo! and Google for PPC Campaign Errors
A lot of people focus in on how to get refunds from Google or Yahoo for Click Fraud issues. Google doesn’t always call it click fraud, they often call them invalid clicks, and when they catch “invalid clicks” they pro-actively discount your bill accordingly. They don’t catch everything, but they do try hard. Yahoo! does the same thing, but less obviously. They don’t actually show you how many invalid clicks you’ve received, they just don’t appear to bill you for every click.
In both cases there are defined processes for requesting refunds or more commonly, credits.
Completing the documentation to request a refund isn’t simple, trivial, or a speedy process, (unless you have PPC Assurance where it’s a one click process). In fact, it’s quite complicated. Rather than confusing matters by outlining processes for both Google and Yahoo!, I’ll focus on Google. They’re the 800lb gorilla which everyone cares about.
In Google’s case, to file a request for credits for clickthroughs you believe you were improperly billed for, you need to identify all the original referrals, which means figuring out which entry in your log file is the original referral, and isolating the unique Google Click ID (gclid). You then need to document everything possible about that click, as in the course of an investigation, Google’s team might ask you for a lot of data. Be prepared. They are just trying to be thorough.
One issue you’ll face is how to claim what. The obvious documentation on the web deals with “Invalid Clicks” Unfortunately, invalid clicks don’t always mean the same thing to you as they do to Google. Not all invalid clicks are Click Fraud. To you an invalid click might be a referral for an incorrect keyword match. These do happen, but you’re unlikely to notice them in a large campaign, as too many terms are flying across your screen. This type of mistake actually gets handled by a different department at Google. Challenging to navigate, that’s for sure.
It’s not that Google actually sets out to make it difficult to claim back a refund, or to get a credit for mistakes. Simply put, Google’ a big organization with responsibilities for different issues assigned to different groups. They are trying to be as efficient as possible, but these efficiencies don’t necessarily make processes simpler for you, or your clients. They simply need to be thorough.
Is it worth your while to manually track down all the errors? It depends on your cost per click, and your volume. Is it worth doing so automatically? Definitely. At a cost of 1% of campaign spend, knowing what’s going on, when things go wrong, and how to deal with them is invaluable. Knowing you can recoup more than that means the ROI is pretty simple to work out.
Ways to Minimize Click Fraud
I was reading yet another article the other day which referred to Click Forensic’s Click Fraud Index, and was particularly interested in their threat map, pictured below.
I’m not going to bother dealing with the numbers quoted, or commenting on which countries are more threatening, but if their threat map is real, what an easy problem this is for you to start dealing with, at least as far as Google, and Yahoo are concerned. That’s right, using their information, it’s relatively easy to minimize your exposure to Click Fraud, and to make your campaign much more effective at the same time.
How about that? Advice which won’t cost you a penny, but will save you a bundle in your PPC campaigns. The best part of it is, you’ll not just limit your exposure to Click Fraud, but you’ll also increase campaign ROI in innumerable ways.
Here’s how you go about it. If you are a retailer selling only in specific countries, why aren’t you simply geo-targeting those countries? If you only sell, or want to reach customers in the U.S., why would your campaign not have geographic parameters? Simple, isn’t it? But you can’t just simply choose U.S. only in your geo targeting. If you do that you’re simply limiting yourself to people accessing .com, .net, and .org sites. People in India (fiery red hot problem spot according to the Click Fraud Index), and Canada (a much bigger problem than the U.S., with only 10% the population), who use Google.com, or go to read businessweek.com will still see your ads. Why?
Well, if all you do is select “U.S. only” then you’re limiting your ads to anyone in the world using the U.S. default engines. Same thing if you’re running a UK only campaign, people in Argentina (another hot spot!) looking for information about the Falklands (err… Malvinas) on google.co.uk, or theregister.co.uk will see Google ads set to “UK Only.”
The good news is that it’s easy to keep those nasty Argentinian click fraudsters away from your ads, (actually I know and like quite a few Argentinians, and their wines!), so you never need to worry about them causing you grief.
The solution really is simple, go into your campaign settings, and instead of choosing “U.S. only” choose each of the 50 States, plus D.C. individually. Now you’ve just limited your campaign to people located only within the 50 States & D.C. Much better, isn’t it?
So what happens now when Google (or Yahoo or anyone else) serves out your ad to a viewer in Argentina, India, or Canada? Well, now you can go back to them, and file a claim for incorrect billings. These clickthroughs should now be labeled as “invalid”, and you should not be responsible for them. After all, if you rent a billboard in Las Vegas, and the company instead erects it in San Jose, you wouldn’t logically be required to pay would you? Same principal should apply here as well.
So, it’s really pretty easy to cut back on your exposure to potential click fraud, isn’t it?
One-Click PPC Refund Service and Click Fraud
In my earlier post over the weekend, I used an analogy to explain what PPC Assurance is. Today, I’d like to expand on that thought and explain some of the extras included in (what we think is) our revolutionary paid-search advertising product. One of our user’s favorite features is the one click refund service for PPC (pay per click) advertisers.
PPC Assurance is more than the next step in the evolution of advertising verification services. Unlike the click fraud companies which garner lots of attention by selling fear, PPC Assurance provides you with resolution.
The product name “PPC Assurance” explains a major aspect of our service: Assurance; know what’s happening, and what you can do about it.
Present click fraud services are not assurance services. They attempt to discern whether or not the pattern of behavior of a visitor “seemed” normal. Some of them let you define the parameters of fraud. This is a poor methodology, because it allows for lots of “false positives.” These false positives are a big issue, because you’re essentially making false claims. They are a disservice to you, and to your ad network.
PPC Assurance does not use any subjective measurement techniques. In fact, we’ve ascertained that most of the real fraud that does occur arrives at your site without triggering JavaScript loggers and is thus invisible to most of the click fraud companies out there.
Realizing this, we consulted with the ad networks themselves and figured out a means to coordinate resolving this issue. Expect a major announcement on that front in January.
Here’s how PPC Assurance works. PPC Assurance examines your account configuration, and compares it to the actual traffic you’ve received. Using a simple and easy to understand graphical report, we identify what was good and what wasn’t, (charges for undesired / not good traffic). We also offer you a one-click refund claim submission for the errant clicks.
Many click fraud prevention companies count mistakes in campaign execution as click fraud. I’ve heard innumerable people say, “If my ad runs in China, and I only wanted traffic from Chicago, it must be fraud.” In reality, it’s not fraud. It’s a mistake. But it’s one you should not pay for, and we help ensure you won’t ever have to pay for such mistake clicks again. We’re the only company which provides this one click refund service.
As a search marketer, or even an independent businessperson who is running a PPC campaign, you don’t want to spend time monitoring and verifying every single click that you receive. You would like to know that you’re only paying for the traffic you wanted, and not for the traffic you didn’t. You don’t have time to manage this process but proper business practices require that you do so. PPC Assurance is designed for you.
Small business owners can’t afford to spend the time to verify your traffic. Until now, they have trusted the networks blindly, and wondering why their ROI is fluctuating. Articles in Businessweek, Forbes, Fortune and a myriad of other news sources discuss the impact of click fraud, making small business owners wonder about their own exposure. Unfortunately, “mistakes” - such as an ad targeting New York appearing in Delhi is really not click fraud, although it’s been classified as such. Now, you can minimize your exposure, and get your money back, without having to spend valuable work time monitoring the problem.
Search marketers are pressed for time with multiple accounts all needing support, advice, management, updates and reporting. The reporting tools used up until now cannot answer the simple questions relating to invalid click activity.
“I see traffic coming in from China, but my campaign was supposed to be in California: how did this happen, and did I pay for it??”
Now you’ll know, AND you’ll know if you paid for it or not. If you did end up paying for clicks you hadn’t asked for, press the “submit claim” button, and get the ad network to refund you for those clicks.
Having taken a unique approach to click-stream analysis, we are proud of our PPC Assurance product because it is the only PPC click protection service out there that fully strives to understand your traffic stream from a search technologists’ perspective. It is also the only product with an instant refund reporting option. It’s that simple.
SES Chicago Notes
Alright, I’m way behind. The Search Engine Strategies conference in Chicago ended ten days ago, and I haven’t yet posted anything.
One of the things I wanted to comment on was SES Chicago. SES Chicago is the smallest of the big 3 SES shows of the year. Each SES has its own flavor. SES NY in March seems to lead attendance, and there’s heavy participation from ad agencies and advertisers. SES San Jose in August draws heavily on the SEO and SEM practitioner crowd. Chicago seems to be drawing a more corporate crowd, and is the only one which seems to draw almost exclusively from the Midwest. The other two draw a more national crowd. All three shows are extremely worthwhile.The conference itself seemed to be enjoying about the same attendance as the previous year. I was told that registrations were off by less than 1% over the previous year. A lot of us were concerned that with WebMasterWorld running in Vegas at the same time as SES, there would be a significant drop in attendance. After all, only a few people would choose snowy Chicago over sunny Vegas, or so the thinking went. I actually like Chicago and snow! This year I accidentally stumbled into a German style Christmas market at the corner of Dearborn and Washington. It’s worth checking out.
In fact, when I spoke to people who had been to Vegas (I flew into Vegas Thursday afternoon right after the SES show, and just in time for the MSN party at Ghost Bar), I was surprised to hear that they were disappointed in attendance. Others told me attendance was on par with previous years, and they were happy. The crowd for this show would be more akin to an SMX Advanced conference. The emphasis is on practitioners. Many of the opinion makers in the search industry appeared in Vegas, and will appear at SMX Advanced. Almost all the other ones who weren’t there were in Chicago.
As to the sessions, I spoke on two panels. Search Marketers on Click Fraud, and User Behavior, Personalization & Universal Search. Having done the latter panel in San Jose in August, the second time around was a lot easier. What made it challenging was that presentation time was reduced to 5 minutes, with a longer Q&A session. While the shorter presentation time made presenters get more to the point, there was a lot to discuss. Each presentation was unique, and I believe a lot of useful information was shared.
Based on questions asked, and comments I received afterwards, I believe the audience got a lot of value from both the presentations and the discussion. Greg Jarboe of SEO-PR did an excellent job moderating. Knowing that each presenter brought something different to the table, he ensured that questions were answered by everyone, so that the audience got a well rounded perspective on issues. He also posed a couple of questions to presenters, which assisted everyone in highlighting points of interest.
The second session I presented at was on Click Fraud. This became a two part session for the conference. I the past there was one panel, with click fraud specialists, marketers and the engines themselves on the same panel. I strongly prefer the new format. Tom Cuthbert from Click Forensics apparently does not, as he actively complained from the stage that he preferred to sit on the panel with the engines. I disagree. The “discussions which occurred when the engines and the click fraud specialists were on the same panel were often not overly productive.
Marketers want solutions. They know they are trying to deal with serious issues, and want to learn information and strategies for dealing with the problems. By splitting the click fraud sessions into two parts, SES is doing something very positive for attendees. The search engine’s / ad networks session allowed marketers to learn about what the engines are doing to attempt to combat the problem, and to ask them specific questions about specific issues. The engines also provided tips as to how they suggest you as a marketer can help combat click fraud.
Immediately after the search engines on click fraud session came search marketers on click fraud. I believe that all presenters attended both sessions, so really, all SES did was allow for twice as much time, and gave marketers an opportunity to focus on the issue from distinct perspectives. The session was moderated by Jeff Rohrs, and each presenter attempted to focus on different issues around click fraud. In my case I focused on campaign issues which often get labeled as click fraud, but really are cases of the ad networks serving out ads improperly.
As you spread your ads out across the content networks, the incidence of mistakes increases. Detecting and providing you a means to recover the costs associated to these undesired clicks are what PPC Assurance focuses on. Click through traffic which does not match the terms and conditions of your contract is undesired traffic. Auditing and verifying your PPC traffic is what we do, and resolving campaign issues through our unique one click refund claim submission is what sets us apart.
I attended a few other sessions in Chicago. The quality was excellent.
Anyone involved search marketing who attends an SES conference will receive tremendous value for the experience. Even after all these years, every time I show up I learn more. It’s not just a great environment for people who know what they are doing to gather and exchange tidbits. There’s opportunities for anyone at every level to get educated. If you’re a decisionmaker, and want to understand the marketplace, there’s sessions which are right for you. If you are the practitioner who deals with the nuts and bolts, then there are sessions for you as well. It doesn’t matter what level you are at, there’s always something for you at an SES Conference.