Maine Munchies Ad

Monday, July 30, 2007

AdWords Experiment Update #1: Democratic Debate Search Trends

Part of the reason for my CNN YouTube Presidential Debate AdWords Experiment is to answer this question, posed in a comment to the Democratic Candidates Miss a YouTube Debate Opportunity post:
But the question remains, how many people seriously search for the debate after it is over?
Recall that the Democratic Debate was on 7/23 and note the Google Hot Trends data for that day for the keyword phrase democratic debate:

democratic debate search trends

Now, a week later, that search phrase doesn't register on Google's Hot Trends service. So, I'm running an AdWords experiment to measure the search impressions in an ad campaign to see if that trend continues. It does appear that people do continue to search for information about the debate. I started the test ad campaign yesterday ~6pm. For the 6 hours from then until midnight, here's the impression data from the Google AdWords campaign for the top few searches:

democratic debate adwords impressions

I should point out that the content network is turned off for this campaign and the budget is set sufficiently high so that the impression data from this ad campaign reflects all searches on Google and its search network (which could include some parked domains). The geographic region for the AdWords campaign is limited to the United States and Canada so this search volume is lower than all searches across the globe. I'll provide more details on the keyword list once the advertising test has concluded. Note that these top searches are all exact matches. For a variety of reasons, I tend to use all 3 match types in a single ad group.

Now, 430 searches in 6 hours (~72 searches per hour) is not a massive amount. However, that was on a weekend and does not count all the variations of searches that include that phrase. The day of the debate, though, and the following morning, there must have been thousands of searches on that precise phrase. Why didn't the candidates advertise then? Why are most still not following the keyword search trends and buying these keywords via AdWords? Here's a recent snapshot of the AdWords sponsored links for a democratic debate search (that link includes "adtest=on" so as not to influence the AdWords impressions count):

democratic debate google ads
When I grabbed that screenshot, there was actually a YouTube ad across the top in the blue - no yellow. So, only 4 ads but only 1 of those is from a presidential candidate, Barack Obama. Well done, then, by the Obama campaign team. Where are the others? I wonder how successful the Obama AdWords campaign is considering the lack of competition it faces.

I'll have more data soon, when Monday is over. That should be interesting - a full day's worth of AdWords data from what's often the biggest search volume day of the week. Hopefully, I've managed to make my ads compelling enough to have a decent quality score but not compelling enough to generate too many clicks. More on that in the next update...

Tags (made w/ TagBuildr + TagTrends): , , ,

Sunday, July 29, 2007

CNN YouTube Presidential Debate AdWords Experiment

In Democratic Candidates Miss a YouTube Debate Opportunity, I argued that the presidential candidates should be watching current search trends and buying pay per click advertising via Google AdWords. I've decided to buy some keyword ads myself, as an experiment to answer this question left in the comments on that post:
But the question remains, how many people seriously search for the debate after it is over?

I doubt the search engines will release the numbers to tell the popularity of the search terms such as you mention -- though if it was significant, you'd think they'd have an advertising rep. contact the campaigns to tell them x-number of people are searching for information on the campaign/debate.
Google, via Hot Trends, does in fact provide some of this information, albeit in a qualitative and not quantitative form, which is also heavily filtered. (I use a tool I'm testing called TagTrends to help me scan the Google hot search trends.) Consider these hot search trend results from the day of the CNN YouTube Democratic Debate and the next day:

youtube debate

youtube debates

For the Google search youtube debate, there was a spike in keyword search volume on the day of the debate. The next day, though, people continued to search for those keywords. One way to see if that trend continues is to go ahead and buy those keywords via AdWords and to then track the keyword ad impressions. That's what I'm going to do. I plan to also use this as a case study to explain some AdWords strategies I implement on behalf of my clients. Subscribe to Apogee Weblog if you want to find out how this AdWords experiment unfolds.

Now, since I'll be spending money during the course of this experiment, I might as well promote the candidates I'm currently considering voting for. As an independent voter, I'll link to the campaigns of two candidates from each party.

Democratic Candidates:
Bill Richardson
Barack Obama

Republican Candidates:
Rudy Giuliani
Ron Paul

Apogee Tags (NOT Technorati Tags): , , , , , , , , , , ,

Wednesday, July 25, 2007

YouTube Debate +1 for John Edwards Campaign

Quick followup to my post (about the CNN YouTube Debates), Democratic Candidates Miss a YouTube Debate Opportunity. Today, I see the John Edwards campaign is running a Google AdWords pay per click ad for the youtube debate search:

youtube debate google ad

So, score 1 for the people managing John Edwards Internet marketing campaign. They're clearly following the trends. Here's the landing page for the ad which includes a YouTube video about hair. Will any of the other Democratic candidates catch up? Since there will be another YouTube debate, the keyword search trends I mentioned in my last post should still hold for awhile. Other candidates would want to be visible for these searches. The most effective way to do that right now is to simply buy those keywords via Google AdWords.

Tags (byo w/ TagBuildr or TagMuse or TagTrends): , , , , ,

Monday, July 23, 2007

Democratic Candidates Miss a YouTube Debate Opportunity

Checking the search trends after tonight's CNN YouTube Debate, it appears as though the Democratic candidates are missing an important opportunity to engage the American public. Consider the pay per click ads running alongside this Google search for cnn debate:

cnn debate search ads

Score 1 for the Republicans - specifically the campaigns of Mitt Romney and John McCain. For these other top searches, I'm not seeing any pay per click advertising: youtube debate, democratic debate, cnn youtube, democratic candidates. Yes, I realize some of these searches are ephemeral. Still, why not buy these keywords, if only for a few hours? Influence those searching after the debate is over. If I had Internet marketing money to spend for a candidate, I'd be watching the search trends like a hawk, even if my candidate was a dove. ;-)

Tags (made w/ TagBuildr with some help from TagTrends): , , , , , , ,

Can Facebook Help Business 2.0?

Interesting to see Facebook used as a platform to help Business 2.0 magazine. There's a Facebook group called I read Business 2.0 - and I want to keep reading. I became aware of the group when I noticed it mentioned on a handful of the blogs I read on a regular basis, including:
I've been a subscriber since my Red Herring subscription was switched over to Business 2.0. At the time I was a bit annoyed, but I've found Business 2.0 to be an excellent magazine and useful for generating ideas. In fact, I've referenced Business 2.0 a number of times from this blog:

May 25 2007
Favorite Bloggers in Business 2.0 Magazine June Issue

May 22 2007
Business 2.0 on CNBC: Domainers Like Ham Profit from Spam

May 02 2007
Trail Mix for Business Travel

Mar 27 2007
Domain Traffic Spinning Straw Into Gold?

Aug 26 2006
Introducing Ads by Apogee

Mar 15 2006
Business 2.0 Energy Food Article

So, I hope they don't cancel Business 2.0. Join the Facebook group, if you want to register your "vote" for the continuation of Business 2.0 in print.

Tags (byo w/ TagBuildr): ,

Monday, July 16, 2007

Purple Flowers Are Not Pink Despite What Google Might Think

In my last post about the new Google AdWords search query performance report, I mentioned that advertisers are going to start noticing problems with broad match keywords. I've been warning readers of this blog, since last year, to be careful with AdWords expanded matching. It's important to recognize that there is no such thing as broad match. Let me say that again: There is no such thing as broad match! Broad match is actually expanded match. This help page from Google about the new search query performance report serves as a good example:
Ad group A - [purple flowers]
Ad group B - "purple flowers"
Ad group C - purple flowers

When the user query is pink flowers, only Ad group C (purple flowers) can be shown for this query. The Search Query Match Type column will say BROAD.
IOW, a broad match of (purple flowers) can trigger an ad for a search on pink flowers. On the one hand, then, Google is putting the burden on advertisers to create rigorously relevant ads with custom landing pages in order to achieve a high quality score. On the other hand, Google is relaxing broad match so ads show up for keywords that aren't highly relevant.

purple flowersIf I've crafted an ad group around a "purple flowers" theme and am driving those clicks to a custom landing page (like these purple flowers), I'd be pretty unhappy to pay for clicks for pink flowers. Expanded matching is contrary to quality score. Is Google putting its own financial gain ahead of its advertisers? Or, is this a case of Google getting too big and the team coding the expanded broad match algorithm isn't working closely with those working on quality score changes? Whatever the reason, the expanded matching algorithm undermines changes advertisers have implemented in reaction to the quality score(s) algorithm.

So, how do you deal with broad matches implemented as expanded matches? I've found it useful to set discrete bids for all three match types. This is contrary to Google's advice:
There's no definitive Quality Score benefit to be gained from adding the same keyword multiple times with different match types to your account.

When choosing the appropriate match type for a keyword, we typically recommend following a 'broad-to-narrow' strategy:
  1. Start by using the broad matching option for a new keyword.
  2. Monitor which keyword variations are triggering your ads with a Search Query Performance Report.
  3. Refine the keyword.
Google does not recommend using all three match types in an ad group. They recommend starting with broad match only. This can be disastrous for a small business. I highly recommend using all of the match types (including negative keywords) in the keyword list for a single ad group. Think of it this way:
Google AdWords Match Types
Exact match is a subset of phrase match which is a subset of broad match, which is actually implemented using expanded matching. How do you apply this knowledge to your Google AdWords account, though? Set higher bids for exact than phrase and then lower bids for broad than phrase. Think of it this way:
  1. With exact match, you know exactly what keywords you're buying
  2. With phrase match, you have a pretty good idea what you're buying.
  3. With broad (expanded) match, you have absolutely no idea what you're buying!
Set bids accordingly. For instance, using the purple flowers ad group theme, that ad group's keyword list might look something like this:

[purple flowers] ** 0.44
"purple flowers" ** 0.21
purple flowers ** 0.07

You still derive the benefit of broad matches without risking too much of your budget paying for untargeted traffic. Track the actual keyword searches that the broad and phrase matches trigger and evolve the keyword list over time. Know what you're buying. Caveat emptor!

BTW, if you've found this Google AdWords tip useful, consider subscribing to Apogee Weblog. I write a fair amount about Google AdWords strategies. Also, try these 11 Tips for AdWords. Not 10. That list goes to 11.

Apogee Tags (made w/ TagBuildr): , , ,

Friday, July 13, 2007

Why More Companies Will Now Notice Google AdWords Flaws

I've noticed some problems with the new Google AdWords Search Query Performance report. What this new report will really bring to light, though, are flaws with Google's expanded matching implementation. Consider a post today (on Marketing Pilgrim) which poses the question: Google AdWords Using Non-Selected (and often Non-Targeted) Keywords? I thought most search marketers were well aware of the fact that all broad matches are actually implemented as expanded matches. Those that don't will start to notice this as they run search query performance reports. Here's a quote from that post:
Today we were doing some experimentation at my firm, and we came across something very bizarre with Google AdWords. We were trying out Google’s new search query reports, which allow you to see the actual search queries searchers used to find your ad. This differs from the keyword reports in Google AdWords in that these reports show the actual queries versus the keywords that you establish in your ad groups.
Readers of this blog have known about these problems for well over a year now. Consider the Google AdWords Budget Tip from March 2006 which mentions: "Google's expanded broad matching has gotten out of hand lately and can lead to ..." Seems like a good time to highlight posts since then that explain problems with AdWords broad match and how to deal with those problems:

December 2006 - Be Careful with AdWords Expanded Matching

March 2007 - AdWords Flaw Could Cost Small Business Millions

March 2007 - Solution to Google AdWords City Targeting Flaw

May 2007 - Top 5 Ways Ignorant Advertisers Lose Money to Google via AdWords

It's important to understand what you're buying when you enter keywords into the Google AdWords system. I hope these posts save people money. Caveat emptor!

Tags (made w/ TagBuildr because search marketing firms don't use Technorati Tags, right?): , , , ,

Monday, July 09, 2007

Postini to be Acquired by Google (not a search engine)

postini on tagtrendsI noticed Postini today via TagTrends. Hopping over to Techmeme, the big news is that Google plans to acquire Postini. More details via the Google blog and TechCrunch. Basics from the press release:
Google Inc. (NASDAQ: GOOG) announced today that it has signed a definitive agreement to acquire Postini, a global leader in on-demand communications security and compliance solutions serving more than 35,000 businesses and 10 million users worldwide. Postini's services -- which include message security, archiving, encryption, and policy enforcement -- can be used to protect a company's email, instant messaging, and other web-based communications. Under the terms of the agreement, Google will acquire Postini for $625 million in cash, subject to working capital and other adjustments, and Postini will become a wholly-owned subsidiary of Google.
The pace of acquisitions (and hirings) at Google is phenomenal. Do they really buy a company a week? As Google moves beyond being simply a search engine, what is it becoming?

Tags (byo w/ TagBuildr): , ,

Tuesday, July 03, 2007

Is Google NOT Serious About Webspam?

Interesting to see Matt Cutts (head of webspam at Google) respond to Thomas Claburn's article on InformationWeek: Is Google's Spam Fight a Sham. Claburn asserts that a post on the Google webmaster central blog about creating startpages is a recipe for spam creation. That's a bit of a stretch but this point is constructive:
And therein lies the problem: It's hard to judge intent. Google admits as much by asking its users to submit spam pages so they can be removed from its index. If Google could do so algorithmically, it wouldn't have to ask for help.
Claburn then moves beyond startpages (that could be misconstrued as spammy, doorway pages) and claims that Google profits from spam on parked domains:
AdSense for Domains treats domain names like search keywords for the purpose of placing ads on the parked domain... Google, I know, would defend these ads as providing useful information to Web searchers. If you ask me, it's a spam service. Google doesn't like people manipulating its index, but it seems to be okay with Web pages posing as real content.
That's a bit of a stretch to use the term "spam service" to describe AdSense for Domains. However, I have seen cases of spam traffic from AdSense for Domains, so I can see why he would make this assertion. Google isn't helping themselves by refusing (so far) to detail traffic on the AdWords side that originates from AdSense for Domains (on the content network but parked domain traffic also exists on the search network).

Matt Cutts seemed irritated with Claburn's assertion that Google can't be serious about webspam since it runs an ad service for parked domains. It's important to separate Google's natural search program from its paid ads program. Cutts works on the former and has no control over the latter. Still, I'd like to see him involved in this meeting. He seems genuine in his efforts to combat spam, although the issue of snitching on paid links is rather silly (but that's a topic for a future post). I think Google IS serious about fighting webspam in their natural search results but IS NOT serious about fighting webspam on the rest of their system that does not exist on the core Google search property. Think of it this way:
  1. Google = Natural Search + Paid Ads
  2. where Paid Ads = Search Ads + Content Ads
  3. and where Search Ads = Search Engine Ads + Parked Domain Ads
  4. and where Content Ads = Contextual Ads + Parked Domain Ads
From what I've seen, Google is only serious about spam in their natural search results. Apparently, it's a battle they're losing (and that's why they're resorting to paid link reporting). Now, why don't I think they're serious about fighting spam beyond Google's core search property? In my post about Clickbot.A I noticed this peculiar quote from Google:
It is important to note that in a Clickbot.A-type attack, top-tier search engines would not pay miscreants directly. Instead, they would pay syndicated search engines a share of revenue, and syndicated search engines would, in turn, pay a share of their revenue to doorway sites that posed as sub-syndicated search engines or referral accounts set up by the bot operator.
See the problem with Google's stance? Since they don't pay "miscreants" directly, they claim to not be culpable. IOW, it's not up to them to police their syndicated network and deal with these "miscreants" (partners who are creating spam or engaging in click fraud). They don't seem serious about dealing with webspam to me. As for Cutts, he does seem serious. However, since 99% of Google's revenue is derived from paid ads, and more and more of those paid ads are not displayed alongside Google's natural search results, it doesn't matter how serious he is. It's the AdWords and AdSense teams at Google who need to address this issue. Not Cutts.

Tags (these are not paid links made w/ TagBuildr): , , , , , , ,

Google is NOT a Search Engine. What is it?

Google is no longer a search engine. What is it? Consider these very recent developments:
  1. Google buys GrandCentral Communications, a company that provides services for managing voice communications
  2. Google dabbles in health care
  3. Google plays ad agency
What is Google?

Tags (byo w/ TagBuildr): , ,

Sunday, July 01, 2007

Sicko: Google Health Advertising Blog

Skimming Techmeme this morning, I stumbled upon a new Google blog: Google Health Advertising Blog, News and Notes from Google's Health Advertising Team. It's the "Does negative press make you Sicko?" post causing a bit of an uproar (see Google vs Michael Moore on TechCrunch). The post itself is a bit peculiar:
The healthcare industry is no stranger to negative press. A drug may be a blockbuster one day and tolled as a public health concern the next. News reporters may focus on Pharma’s annual sales and its executives’ salaries while failing to share R&D costs. Or, as is often common, the media may use an isolated, heartbreaking, or sensationalist story to paint a picture of healthcare as a whole. With all the coverage, it’s a shame no one focuses on the industry’s numerous prescription programs, charity services, and philanthropy efforts.

Many of our clients face these issues; companies come to us hoping we can help them better manage their reputations through “Get the Facts” or issue management campaigns. Your brand or corporate site may already have these informational assets, but can users easily find them?

We can place text ads, video ads, and rich media ads in paid search results or in relevant websites within our ever-expanding content network. Whatever the problem, Google can act as a platform for educating the public and promoting your message. We help you connect your company’s assets while helping users find the information they seek.

If you’re interested in learning more about issue management campaigns or about how we can help your company better connect its assets online, email us. We’d love to hear from you! Setting up these campaigns is easy and we’re happy to share best practices.
sicko search adsI don't think Google anticipated people outside of the healthcare industry would read this blog. Note the term "issue management campaigns" used in the post. I think this confirms my Google's Advertising Hubris Reaches Apogee post from a few months back. If you're using any of the "free" services from Google like conversion tracking or analytics, I think it's worth asking yourself how Google is using this data. Are they using your data when they build campaigns for your competitors? Are you sure they're not?

Looking at the current Google search results for sicko, I see that a handful of companies have, indeed, created "issue management campaigns" using Google AdWords. I wonder if these companies thought of this on their own or if they were prompted by Google's health advertising team.

I'll be watching TagTrends over the next few days. I wouldn't be surprised to see either "sicko" or "michael moore" on the list.

Tags (byo w/ TagBuildr): , , , , ,