Monthly Archives: April 2011

Heroview – Analyzing Ad Strategy With Tom Demers

As Co-founder/Director of MeasuredSEM and former Director of Marketing with Wordstream, Tom Demers (@TomDemers) is one of those PPC experts that many PPC marketers want to talk to.  That’s why we chose him to participate in our first-ever live #Heroview – real-time discussions with industry experts via Twitter. There are many SEM experts we at Hanapin Marketing want to meet and learn more about, so we thought you, our readers, would want to hear their advice too! Our Heroviews will be available on Twitter (@PPCHero) and here on our blogs every month. Our interview with Tom Demers, below, is about PPC ad strategy, ad copy, and ad metrics. It’s amazing how much insight Tom was able to cram into 140 characters!

Thank you to everyone who participated on Twitter this past week. Stay tuned for more Heroviews coming your way soon!

April 22, 2011


PPC Hero: Hello everyone and welcome to our first live interview! Today, we’re featuring Tom Demers to see what his take is on ad copy/testing.

Tom: Thanks for having me guys.

PPC Hero: So without further ado, lets get started! How are you today Tom?

Tom: Great, thanks guys! Very sunny here in Boston :)

PPC Hero: To get started, tell me a little about yourself. How long have you been working with PPC?

Tom: I’ve been working in search for over five years – in-house, agency side, independent consulting and I’m currently doing PPC & SEO consulting full time at

PPC Hero: Impressive. Given the theme of our interview today, how would you rate your experience writing ad copy and testing?

Tom: Pretty extensive, particularly since I’ve started doing some work with the folks BoostCTR– an ad writing network. They write ads for a variety of clients and have a lot of different copywriters, so have a lot of ad tests lately.

PPC Hero: So what is your typical thought process when going about testing different ads?

Tom: I try to think about the audience, and then look at what the data is saying – do I need more conversions, more clicks, or?

PPC Hero: How many variations of an ad do you prefer to have running at any given time?

Tom: Depends on the volume of the ad group – ideally 4-5, but for lower volume groups I run two to get to stat significance faster.

PPC Hero: How should the ad copy for branding campaigns differ from ad copy that’s meant to bring in conversions?

Tom: First, you know: mention the brand :), and also be more disruptive. You want to be remembered.

PPC Hero: What are the most important metrics YOU take into consideration when evaluating ad copy?

Tom: For me it’s: ROI first, conversion volume and costs second, CTR third.

PPC Hero: So what are some red flags that an ad is doing poorly? Perhaps something not as obvious as poor CTR, etc.

Tom: Consistently low quality scores within a tight ad group are a bad sign, and ads that perform worse than their peers. This is a great peer calculation worksheet:

PPC Hero: Should your ad copy strategy differ between Google, Bing and other third tier engines?

Tom: My approach is typically similar to begin with, but then you’re reacting to different data, so your iterations should be unique.

PPC Hero: How should your ad copy strategy differ between the search network and the display network?

Tom: Completely! Display network ad copy should be disruptive & seductive because of the different level of intent. On the other hand, Search network copy should be relevant & satisfying.

PPC Hero: What is your ad copy advice for PPC accounts with low traffic, in terms of testing ads and writing content?

Tom: Test fewer ads & use tested approaches. If you have a high traffic part of your account, try to transfer successful lessons from there into lower volume portions of your account.

PPC Hero: From your experience, what are the biggest mistakes PPC ad-writers tend to make?

Tom: Ignoring the search query/targeting options and leaving off calls to action: be relevant and tell people what to do next!

PPC Hero: Are there any particular types of ad copy experiments more people should be trying?

Tom: I really like things like symbols ( You can also try a JARRING, challenging headline as a test too.

PPC Hero: Interesting.. Recently you’ve been working with ad text crowdsourcing. What is the most surprising thing you’ve found so far?

Tom: The most surprising thing is the consistency with which small tweaks have double to triple digit impacts on CTR & Conversions. There are lots of good examples here:

PPC Hero: Something to look into. Well, that wraps up this month’s #heroview. Are there any questions the audience might have for Tom?

  • @JARooney8423: For display network ad strategy, do you mean surprising/distracting disruptive, or inciting disruptive? Can you go too far?
  • Tom: Yeah depends on the objective – for direct response & attention I like disruptive but depends on the client too.
  • @JARooney8423: Interesting! Do you find that you don’t need a call to action if your content is really eye-catching all by itself?
  • Tom: I think a CTA never hurts but particularly on Facebook I think a lot of times the image does the work.
  • @SEM_PPC_MattV: Where do you see the PPC industry going? (I’m being intentionally un-specific here).
  • Tom: More display/content, more creative-based (focus on ads and landing pages) & dramatically better tracking.

Tom: Thanks guys! This was fun – cool format and great questions!

PPC Hero: Thanks again for your time, Tom. Some very good insights!

Tom: Anytime! Stay dry!


Twitter Facebook LinkedIn Google+ Email Print More

Our Guide To Google Analytics IQ Testing: C is for Cookie

This is the first installment of a collaborative series between PPC Hero and SEO Boy to help readers prepare for the Google Analytics Individual Qualification (IQ) Test. This series covers the most important aspects of the IQ Test and will prepare you to earn a high score. Our writers in this series have all scored at least 90% on the test and will provide their personal insights for success.

cookie monster pictureCookie Monster has said on numerous occasions that C is for cookie and that’s good enough for him. But as Internet marketers, we know that cookies are so much more than that! As a matter of fact, they are one of the most important concepts covered in the Google Analytics IQ Test and you have to understand them if you want to pass the test. Fortunately, cookies are also one of the easiest concepts in the test to learn about and understand.

In order to pass the Google Analytics IQ Test, you need to score at least an 80%. There are a total of 70 questions on the exam, so you need to correctly answer at least 56 questions. Based on what I’ve personally observed, you can expect that about 15% of any given exam will pertain to cookies, so it’s worth spending some extra time studying them. I’ve organized the rest of this post into two sections: what you need to know about cookies and what you can expect on the exam.

What You Need To Know About Cookies

Google Analytics generates data about users who visit your site through pieces of text exchanged between browsers and servers called cookies. They are established when someone enters your site and analytics data is generated when they click off of your site.  Google Analytics cookies are first-party cookies, which means that they are established by your website. This allows the data generated by cookies to only be viewable by the website owner.

There are four cookies that Google Analytics establishes, _utma, _utmb, _utmc, and _utmz. (If you establish custom variables to define additional segments of data, you will use the custom variable cookie, _utmv as well) Let’s look at these cookies individually. Keep in mind that the most important things to remember for the test are the names of the cookies, how long they last, and what triggers their expiration.

_utma: This cookie identifies unique visitors by assigning them a unique ID upon their first visit to your site. It will last 2 years so long as the user does not delete it. At the 2-year mark, it will expire.

_utmb: This cookie generates data about a visitor’s session. It establishes or updates a session every time a page of a site is visited. If it is not updated, it expires by default after 30 minutes.

_utmc: This cookie works hand in hand with _utmb. It expires only when a user quits a browser.

_utmz: This cookie determines what brought a user to your site. This cookie tracks things like PPC ads, organic search, and referring links.  It also tracks page navigation so that you can see how visitors move through your site. This cookie expires after 6 months unless it is updated. Updates occur when a page view initiates.

Each cookie contains unique code. For example, the _utma cookie looks like this:

_utma= 17334963. 432099211. 1115376676. 1115376676. 1115376676. 1

The first number set is the domain hash and is followed by the random unique ID, the time of the initial visit, the beginning of the previous session, the beginning of the current session, and the session counter.  It’s important to note the time of the initial visit can be the same as the beginning of the previous session and/or the beginning of the current session if the visitor is there for the first time or they deleted the _utma cookie in the past.  These times are set in UNIX, which is seconds since January 1, 1970.

The _utmz cookie is organized slightly differently and allows you to generate a wonderful code to track your PPC activity. It begins with the domain has, but is followed by a timestamp instead of the code contained in the _utma cookie. It also contains a session number, a campaign number, and source values for things like PPC tracking (which you use in a PPC destination URL). It will look something like this:

_utmz= 17334963. 1115376676. 2. 3.utmcsr=google|utmccn=(campaign)|utmcmd=ppc|utmctr=keyword

If you have additional questions about cookies, feel free to comment on this post or visit Google’s page on cookies in Analytics.

What to Expect on the Exam

Once you understand the ins and outs of cookies, you’re technically ready for questions dealing with them on the IQ test. Sometimes it’s helpful to understand what kinds of questions may be asked about particular topics. Every IQ test question I’ve ever seen about cookies has been a multiple-choice question dealing with one of four things; the names of the cookies, when they expire, what they track, and how they are organized.

Names of Cookies:

You can expect a question or two about the names of the cookies. A common one I’ve seen is, “What are the four kinds of cookies used by Google Analytics?” The answers are, of course, _utma, _utmb, _utmc, and _utmz, but occasionally you’ll see answers like _utmd and _utme.


In my experience, common questions about the expiration of cookies have dealt with multiple cookies at once. For instance, a question you might see asks which will not expire if there is a new page view while they are active? The answer to this is _utmb and _utmz.

What Cookies Track:

The most common cookie questions address what each cookie tracks. For example one question might ask “The _utma cookie tracks what aspect of a visitor’s activity?” Answers could include what brings a user to a particular page on your site and when a user quits their browser, but the correct answer is that it assigns visitors a unique ID.

How Cookies are Organized:

I’ve only ever seen questions about the different components of the _utma cookie, but I recommend reviewing the slides on the Conversion University site to understand each component of each cookie. I’d also suggest having this page open when you take the exam.

In the next few days, some of my colleagues will discuss other key aspects of preparing for the Analytics exam. Stay tuned to PPC Hero or SEO Boy for these helpful tips and as with any of the concepts we discuss on our blogs, leave us comments if you have any questions. Good luck!

Make sure to check out the rest of the series:

Part 2: E-commerce Tracking

Part 3: Event Tracking & Virtual Page Views

Part 4: Regular Expression (Regex) Character Guide


Twitter Facebook LinkedIn Google+ Email Print More

PPC Hero By Night…Easter Bunny By Day?

PPC Hero is really trying to get into the Easter spirit! But…brown just isn’t his best color. And it looks like SEO Boy agrees – PPC Hero should stick to his usual costume! Happy Easter from all of us at PPC Hero and SEO Boy!


Twitter Facebook LinkedIn Google+ Email Print More

News Update: AdWords’ New Geotargeting Settings Cut Advertisers Some Slack

Good news! Managing your geo-targeted campaigns in Google just got a lot easier!. Remember the days of concatenating geo-modifiers to long lists of keywords to nab searchers not physically located in your target market? Well, as of this week AdWords will now let you target people geographically in three ways: physical location, search intent or a combination of both, which means your days of long concatenated lists are over (or at least as far as geo-targeting is concerned).

How, you ask? Targeting your selected locations by search intent means the locations selected in your campaign settings will be automatically applied to search queries. For example, your ads would show for anyone who is searching “hotels in Chicago” regardless of their physical location. This alleviates you from having to plug in keywords for all relevant location names. Conversely, you can also exclude traffic by search intent, which could be incredibly useful in large metropolitan areas where services really are borough specific.

Implementing this new method of geo-targeting for your campaigns is easy.  If you’re in Google AdWords, click into one of your campaigns, and then choose the “Settings” Tab. Under “Locations and Languages” click “Advanced location options.”

Next to “Targeting method” you can choose how you want to filter searchers geographically (location, search intent, or both). From there, you’ll just select your locations as you would have previously.

Not only does this change save you time, but it also makes your targeting effort more exact. You won’t have to try to come up with all the different phrase variations people might use in their searches for location-based services to target them. Plus, without having to manually create lists of keywords for this type of targeting, your keyword lists will be easier to manage and track since you’ll only be dealing with your core terms.

All in all, this is a great update that will allow you to more easily target people nationally searching for services within your region. And I’m sure you’ll be able to find something worthwhile to do with all the time you were spending on geo-modified keywords. Feel free to share your thoughts with us in the comments below!


Twitter Facebook LinkedIn Google+ Email Print More

Help PPC Hero Save the Earth!

We can always count on PPC Hero to help! Today he’s supporting the Earth, and so should we! You don’t have to be a superhero to help save the planet. Get out there and plant a tree, or pick up trash in your neighborhood, commit to driving your car less and taking public transportation instead, or join in on a community Earth Day activity. Let’s join PPC Hero in being good stewards of our Earth and its resources! (Check out what SEO Boy’s doing to help!).


Twitter Facebook LinkedIn Google+ Email Print More