PPC Blog | Paid Search Marketing | Sem Geek - Making Sense of Paid Search Marketing

What Search Marketing Analysts Could Learn From Sports Analysts

On the way to work one morning, I was listening to the Philadelphia local Sports Radio Station, WIP SportsRadio, that detailed individual statistics don't really matter as much as WINS and LOSSES. In fact, some of the radio hosts were animatedly opposed to statistics as a false representation of what they see on the field. They also talked about how statistics do not represent the athletes true skill set or mindset during the game. For example, you could be a Home Run King, but also the leader in strikeouts and be the worst 1st baseman in the field and lose games for the team.

Here is a great quote from a MLB Hall of Fame Pitcher, Tom Seaver: "If you dwell on statistics, you get shortsighted. If you aim for consistency, the numbers will be there at the end."

So that radio talk show got me thinking about how we can relate that mentality to Search Marketing and Web Analytics in general. Here is my view on how they are the same.

Similarities of a Baseball Analyst and a Search Marketing Strategist:

Typical Baseball Analyst: Most of them have played baseball for years, worked their way up the ranks and have the uncanny ability to provide insights about the game, what's going on in the players heads and what the audience is interested in hearing better than someone who just reports on what is happening.

Search Marketing Strategist: (In my opinion) Most SEM's have spent years "doing the actual work", providing cutting edge best practices, understand web analytics, understanding the client's business goals and objectives, dealing with multiple clients, presenting high level plans, etc... Over time, that person (should) be promoted to a higher level where they develop the strategic plan, organize a team around them and make the client happy and give them reports and ideas they want to hear.

Too Much Data Syndrome: I have often blogged about how there is too much data in analytics and too many evangelicals and analytics companies are pushing more and more nifty bells and whistles and putting a high price on them. However, with the sinking economy, increased CPCs and many companies going "in house" for the PPC Marketing, how can one afford the time to spend 5 hours comparing apples to oranges when all that matters is:

1. What is working?= Wins
2. What is not working? = Losses

In conclusion: According to Wikipedia.com, Baseball Statistics were not fully understood until 1951 and over time, (allegedly) more and more Sports Analysts use statistics as more of information vehicle than what is really happening in the "Winning" Column. I am not saying statistics are a bad thing, they are a crucial and certain metrics are more important than others and have and will always be a necessity to help make intelligent decisions.

However, the purpose of this post is to highlight on the fact that if you starting judging your clients or business success on a single statistic, you may not realize that your statistically best home run hitter often drops fly balls in the outfield and may cost you the World Series.

Google Glitch On Hiding Cost Data From Placement Performance Reports?

Now, I am sure this is just a glitch and a few loyal readers confirmed for me that it might of been just a rare glitch (Thank You), but tell me why Google has has these "bugs" when making it difficult to see how much their placement/site targeting partners are costing us when we run a Placement Performance Report. It appears that when you utilize the new conversion tracking option and select "ACTION NAME" in your report, it deactivates all other metrics relating to costs and sales/leads values.

Obviously, we can get this costs data while in the campaign & Adgroup level management area of Adwords, but I must ask the following questions:

Why is this data not available to us in a report format?

1. Is Google doing this on purpose?
2. Are they waiting for their recent Yahoo Search Partnership to roll out something new and exciting?
3. Are they simply trying to hide us from the "Macro-level" analysis that could hinder the continuous growth of ad spend on content and placement networks, since the Search Network CPCs are out of control and advertisers are migrating to these networks?

Here's a description of the report:

Placement Performance report
Create a report showing statistics (Cost is a statistic right?) for ads that appeared on specific domains or URLs in the content network. Note that complete statistics for the current day aren't available until the next day at 3:00 p.m. PT (Does Cost pertain to complete? You cannot even choose Cost as an option)

In conclusion
I know this post is rather short and it was intended that way because there's not much you can say about Google's continuous mind games with regard to data. In a blog post I wrote some months back, I talked about how Google makes it very difficult to identify the "actual" parked domains, for which they refer to as DOMAIN ADS. Reason for the importance, is because they base their assumption on what is working, instead of the advertisers "real"experiences. I just find it interesting that Google has a nifty way of hiding data when it comes to $$$ and their network partners for which your ads are being run on. Maybe with the recent Yahoo deal, they will shed some light on that.

Why Too Much Data Can Kill A PPC Client

Do you think the "traditional PPC client" has the time or interest to read through 50 different reports, and find it as interesting as you the Search Geek that got all excited that changing a keyword to "phrase" match to [exact match] caused the CPC to lower by 2 cents? Nah, I don't think so. Is it that important to create a 30 page Powerpoint presentation to tell them that they saw a 5% increase in their overall CTR% because it was over the holiday weekend and impressions were down. Wrong again! Let discuss why this stuff does not matter.

If I were a client on the receiving end of a 1-2 hour weekly and/or monthly PPC status call, I would dread it, because I would not want to hear the PPC Guru or the Account Manager on the other end telling me every stink'n  little change in the traffic metrics different from the previous week or month. The first thing I want to know is the following:

Ok, I put $10,000 in the Cash Register this month, so how much did I get back?

That is the only thing I really care about because I gotta tell the VP, CEO, President that we are not losing money, so that I, myself can sleep at night.

Trying to Sound Smart To A Client? They Don't Care!
They can honestly care less about how smart or good you were to figure out the latest Algorithm or idea to create a 2 month Ad/Creative test where the only change in the Ads is a SINGLE WORD. They would also care less about how you broke out some adgroups to increase relevancy which will keep your keywords from going Inactive and forcing $5-$10 CPCs. Sounding smart can only do one thing, Make you feel better about yourself or to ensure your own boss that you are worthy of your next paycheck.

However, If you tell them that you created a new campaign from a new line of products which is delivering a 500% ROI, that is where you have something interesting to say. It's all about talking combining the "Weeds" with the "High Level" business conversation. Personally, in the past I have been at this crossroads where I had a difficult time switching from "In the Weeds Boy" to a polished high level and strategic presenter. But you always have to keep in mind that RESULTS is what matter most.

Too much Data, Too Much Overhead:
Not only can too much data confuse a customer or give them a MIDOL moment, it can also take a toll on you or your analytics team who is responsible to preparing these extensive reports that typically get thrown in the trash or deleted from the email a couple days later. Why overdue it when you could be doing something else, like say improve your client's PPC campaign.

In my opinion, the only slides that truly matter to a PPC-Only client are the following:

• Performance Metrics (orders, leads, CPA, Revenue, ROAS%, ROI%, Conversion Rate, AOV)
• Monthly Traffic (overall: clicks, new vs. returning, impressions)
• Monthly Engine Breakdown
• Month to Month Comparison (all metrics)
• Top Products (margin, Revenue)
  
• Next 1-4 month projected plan (seasonal, offerings, testing, etc..)

> Slides that Matter to some PPC-Only clients:

• Top 25 keywords & ads (revenue & traffic only)
• Events that made an impact (offline, online, seasonal calendar, etc...)
• Funnel Report (if conversion enhancements have been made or for benchmarking)

Information that MAY not be important to a PPC-Only client are:

• Avg CTR%
• Avg Time on Site (horrible metric these days with Tabbed Browsers in the background)
• Bounce Rate
• Entry Pages
• Exit Pages
• Path Analysis
• Browser type, screen resolution, time zones, etc...

In Conclusion:
It is obvious that Analytics is extremely important for measuring success of the client's investment of both the search engines, as well as your PPC knowledge and experience. We have to always realize that PPC is a business that is "cut-throat", has a high turnover rate and is all about results and making money for the client.

However, even with these realities, it is also a very rewarding and exciting industry to be apart of. To be a great PPC marketer, you not only have to work in the trenches and perfect the craft of being proactive, you also start thinking strategically about the client's business goals and through experience and confidence, you will realize that the metrics that may be important to the "Geek" in yourself, doesn't matter a whole hell of a lot to the client.

IndexTools & Yahoo: You Must Accept The Agreement Or Goodbye Customer

One of the first things I worried about with the Yahoo purchase of Indextools, was the impact it would have on the agencies who have many of their own clients under this umbrella. So, in the spirit of sharing what I know, below is word for word what Indextools had sent me and I am showing to you. If you are an existing or past customer of Indextools, below is some important transition information that you will need to know.

Clients & Partner to Accept new Yahoo Agreement:
In case you have not received In the coming days, we’ll require our partners and clients to accept a new Yahoo! agreement. We will make this agreement available to you soon for your review and let you know when your acceptance is required. The only issue so far with this is the next sentence which basically says "either you are with us or against us" It is important to note that our clients and partners must accept this agreement to continue using the service.

If you accept, it's FREE
Yahoo! currently intends to provide the service FREE of charge to clients and partners who accept the Yahoo! agreement. Further to this, it should be noted that Yahoo! does not intend to add any new partnerships or direct clients in the short to midterm, while we prepare for the next roll-out wave. This means that both our partners and clients will be in a unique position to either provide a service or use a service to which the rest of the market will not have ready or immediate access.

Only Indextools can Add Accounts of Projects
This is an interesting section to read. To avoid a deluge of accounts being added to the system, we have temporarily suspended the ability for you to add accounts/projects to the system. Your account manager has full access to do this and I would ask you contact them directly if you wish to add an account or project to the system. Please understand we have a lot going on at present and be mindful of requests to add 10s and/or 100s of accounts/projects to the system at this time.

Ongoing Support:
Our support capabilities will remain unaffected so that we can continue to serve both our partners and clients alike over the coming weeks.

Conclusion:
C'mon, this was eventually going to happen anyway, especially with the shift in company buyouts to control the market share of the Internet properties as well as advertising expenditures. Indextools offers a great solution for a much cheaper price than Omniture, Atlas, etc... Perhaps, this is Yahoo's attempt to out-due Google for their Google Analytics purchase of Urchin back in the day. As long as they continue to enhance their product and make it available to everyone at a lower cost, it's a win-win for everyone.

Why Monetizing Every Success Event Is Key To PPC Performance

The typical performance metric that is used to quantify a successful Paid Search Campaign is ROAS%, ROI% or Lead Generation with a defined CPL or CPA. Sounds straightforward and fair right? NOPE! Suppose you have a PPC campaign that has an actual ROAS of 100%, but the benchmark is 500%. Plus, the client (with the intent to improve overall conversions) decides to throw in an #800 on the site, extend and complicate the online form and/or increase the price or shipping of products without telling you the PPC Marketer. Guess what happens to the ROAS%? Guess who get's the call at 3am in the morning (I know a little over-dramatic for those political types, but can you say Roto-Rooter??)

Everything mentioned above will affect your PPC Performance. Ever notice that the more you spend in PPC, the Bookmark/Type-ins and the Brand Campaigns keeps getting stronger and better while the others barely keep themselves afloat? All of these scenarios are taking away from all of the keyword research, conversion analysis and Ad testing that you have done and sold the client on. So, you need to protect your efforts and find a way to set the expectation with the client from the beginning and find ways to monetize all of the consumer interacted events on the website which can lead to some sort of a conversion.

Other KPIs that deserve more attention:
Other Success Events traditionally not a part of the typical KPI mix that clients rarely want to include in the PPC efforts are (1) Email subscriptions, (2) Newsletter Subscriptions, (3) Live Chats, (4) Telephone Sales, (5) Bookmark Type-ins and (6) SEO traffic. Forgive me if I left anything out.

So how do you monetize these events?
Well, it may not be as simple as you would expect and everything is riding on the participation by the client to provide you an average monthly rate of these events. For example, if the avg monthly email sign-up=25, or # of live chats is 100/month, then you need let the campaign run for the first month, collect the data from the client and start the predictive process. Also, you will need collaborate with the client on how much each of the success events are worth to them. Is an email worth $1.00, Chat worth $2.00, telephone order AOV (Average over value)... you get my drift.

Why does monetizing these events matter?
Depending on your original discussion with the client during the scoping and proposal phase of the process, you need set the "reality" of the online world to them and explain that in paid search they are getting much more that an ROAS% metric. They are getting traditional increases in all success events and that there is a lifetime value associated with every impression or click of a paid search ad (textual or display).

Difficult Argument for Retail or Finance, but not for Political, Non-Profits and Charities:
When dealing with retail, there are so many factors that come into play, the biggest being competition. You have to deal with shipping rates, inventory issues, and simply complexities of the shopping experience. However, when dealing with 501c3 Charities and Politics, the conversion metrics are simple to identify yet difficult to monetize efficiently, such as :

• Donations
• Volunteer Sign-ups
• Email Newsletters
• Campaign Update emails
• Impressions per thousand (eyeballs)
• Bookmark Type-ins
• Returning Visitors
• New Visitors
• Viral participation (linking, etc...)

Plus, in order to capitalize on everything that your online marketing efforts are doing, I am seeing a trend where in order to communicate "online marketing" to political campaign managers and/or charity fundraising directors it needs to be compared Print and TV. The reason is obvious. They have been comfortable for so long with traditional advertising, it's imperative to to compare online metrics such as impressions to basic local and national circulation data (ie. eyeballs), demographics, messaging and expected response rates based upon the advertising spend.

In our online world, this is actually a very significant opportunity, yet still no one gets it. So to counteract this behavior, putting a price on impressions, emails, volunteers, donations, newsletter sign-ups, organic traffic, Viral marketing, etc... is a must part of the negation (aka. proposal).

Once you can get past the hours of discussions and persuasion to put more money online, their ability to see the big picture and have web analytics available for every metric they need, the strategy that was very recently seen as an unrealized Pewter statue is now golden.

With Looming Recession, Will High Priced Web Analytics Be A Thing of The Past?

Not to sound like a pessimist with a recession looming, but with all of the signals of a significant cut in advertising spend as well as resulting layoffs in the Newspaper Industry, see news article on Philly Newspapers announce layoffs from WPVI Channel 6abc.com, companies are going to eventually look to cutting back on online expenses, and instead of cutting PPC spend, they should re-evaluate whether the high cost analytics companies such as Omniture, WebTrends, Atlas and others are really worth keeping around. Let's discuss....

We need to watch the other Media Channels:
I think that with major newspapers laying off employees due to a continuous drop in Ad Revenues is a significant downward trend that all online marketers need to pay attention to. It's just a matter of time until these changes will start effecting the online world. It is inevitable. Now, with that said, I would think that cutting $2k - $10k/month in just analytics/reporting costs would be the first to go. Furthermore, I am not recommending getting rid of analytics, In fact, I am suggesting shopping for a lower cost solution that will give you want you need and hence enable you to either keep the extra cash or put that money back in your online ad spend. PPC spend is only going to get higher, so now is the time to start thinking about how and where your budget should be allocated.

The Benefit of going with a Lower Cost Analytics Solution:
Let's face it, do you really need 1,000 ways to look at data, or find (1) or (2) good ways that make sense to you and your business. Why come out of pocket for fancy "bells and whistles" that you may not pay much attention to when you can focus on the reports that give you what really  matters to your business such as ROI%, ROAS%, CPA and basic attribution and traffic data of all other "tracked" online media vehicles.

Web Analytics is here to stay, just not the "over priced" ones:
With the increased market share of PPC spend in the world, tracking the performance of every penny will never go away. However, the magnitude of high priced dashboards, bid automation, and fancy rules sets, as well as manual "dicing and massaging" of data maybe on the fringe of  non-existence. I know this is a strong statement to make and will have some negative responses to this, but this is in effect all about simplifying the data and lessening the monthly cost of getting overloaded with too much data that honestly doesn't make an impact on your business.

Here's the Bottom line:
Knowing what is profitable vs. not profitable is more important than how many lbs. of Hypo Allergenic Dog Food was sold on a Thursday vs. Dog Collars on a Friday. Get the point! In my "tunneled view", Analytics is about enabling us to make decisions that are much more intelligent than just "winging it". I get the sense from some analytics companies who pride themselves on putting on "too good to be true" powerpoint presentations where they want to look "smarter" than anyone else. Problem. Acting smart DOES NOT make it important.

When shopping for a new car, are you going to buy one that comes with a LCD TV, a toaster and a microwave. Of course not. The car makers could accommodate it for you (why?), but it's gonna cost you extra and all that could matter to you is gas mileage, safety and four wheel drive. After writing this analogy, I often think of the Simpson's Episode "Oh Brother, Where Art Thou?" when Homer is given the job to create the car of his dreams. In the end, that "dream car" which had all of these "bells & whistles" actually ruined his brother's company. I know, bad example, but I think it proves the point.

Adwords Impression Share Exact Match: Fact or Fiction?

In my previous post, I talked about the importance and/or irrelevance of Google's impression Share Metric and now I am discussing how the Impression Share Exact Match metric, which is frankly more of a brain-freeze than an interesting brain teaser. After reading Google's description of this report, (say like 3 times) I still cannot make any clear assumptions. In fact, when I ran a report with both Impression Share and Exact Match Impression Share, almost 95% of all of the campaigns show a higher Exact match impression share percentage as compared to the standard impression share. So, if I have interpreted this correctly, there are less opportunities to spend more on Exact match keywords and more to spend on Broad Match. So, when you put the two together, it means that Google wants its advertisers to spend more on generic keyword coverage, which conveniently for them,  has a much higher CPC and search volume potential than exact match. However, none of this should be a shock to anyone. Why? It's just more "gold" in Google's vault. (hence my photo of Mike Myers as Goldmember in Austin Powers in GoldMember. Hee hee!

So what is this telling me?

Here's Google interpretation: According to Google, "Impression Share Exact Match reports the impression share of your campaigns as if your keywords were set to Exact Match. Since keyword match type can affect your impression share, Impression Share Exact Match can help you determine your share of voice for just the specific keywords you are targeting without requiring you to make adjustments to your campaigns.

For example, one of the reasons why you may have low impression share is because you’re not showing on broad match keywords. Impression Share Exact Match tells you how your keywords would perform if they were all exact matched, so that you can determine whether or not match type is a cause for low impression share. This lets you rule out match type and focus on other possible factors when you are working to improve your impression share. "

Here's my interpretation:
Google wants its advertisers to believe that there is less opportunity on highly targeted long tail exact matches and more opportunity to increase CPCs and expand more broad/generalized keywords in order to increase market coverage and hence improve business goals. Funny phenomena, since Exact Match is supposed to filter out the bad traffic and drive more qualified leads to your site while keeping your costs down in order to stay in the game. How is it that with every new bell and whistle coming out of Google is suggesting you spend more and more on broad match, and less on what matters to you most. Cost effective performance.

Again, This feature is available as a column option under ‘Level of Detail’ when you run an Account or Campaign Performance report. Data is available for search campaigns only.

~ “No great marketing decisions have ever been made on quantitative data” - John Scully

Query Eyed for the Straight Keyword Results

Have you ever taken a look at the Google Search Query Report? Notice anything a little weird in the report? Well, myself and other colleagues have and it's a little annoying to say the least. Even though Google has far better reporting features than those of Yahoo, MSN, , however when they offer something to it's users that is not delivering the expectations that they claim, it's almost like they don't want you to know. In this case, identifying all of the actual "Search Queries" that people are typing in based on your Broad or Phrase Match, would be extremely valuable in terms of driving the most qualified traffic for the $$$. How? We would simply take all of the queries, make them [exact match] or "phrase match" and hence only drive traffic to the terms that people are typing in. You may ask yourself these two (2) questions.

1. Is this the best scenario for driving traffic. Not necessarily.
2. Is this the best tactic for improving ROAS? Well, it's definitely a step in the right direction.

So what does Google say about this. You can read Google's interpretation of this feature Google Search Query FAQs. But what is most annoying with this article is that their example of a Search Query Report is not matching up reality (ie. withholding data in actual "real-life" accounts.)  In other words, they show you this below as an example of the report (see below):

Well, as you can see Google is not being as straightforward with this reporting feature and the lingering question is why not. Why are they not disclosing the actual keywords that are being searched on. What's the big secret? I mean, there are many other Analytics packages do offer this and are relatively easy to get. You have to pay for them, but it's out there.

So in conclusion, Google claims they provide the actual search queries to it's advertisers, however I have seen that they are not disclosing them. Perhaps, this is a technical glitch and will be fixed over time. However, we cannot ignore the possibility that they are intentionally keeping this from their advertisers in order to continuing to drive up their stock price and keep us all guessing?

What do you think?

Web Analytics: Are We Losing Focus On What Really Matters?

Why is it becoming more and more difficult to go from "Point A to Point B" these days with regard to Web Analytics. Which is more important to you? Identifying what is working and what not is working? or whether you have sold 2 more blue widgets than red widgets and that those blue widget purchases came during a 1 hour time span served from 2 different search engines with totally different Ad/Creatives. Are you getting the point?

Well, unless you have an infinite amount of free time (which I doubt) or if you have (12) twelve interns working around the clock looking at this stuff in a high powered expensive analytics package, I would think it's more productive to identify the following:

• Why did they buy?
• Why didn't they buy?
• How can I make more people buy?

My point with this article is that we have been consumed and taught (of course with fancy presentations and lots of chotckys) to take advantage of all of these "Top of the Line yet expensive bells and whistles that will tell you the world and predict everything about visitor that you could ever want to know. But why waste hours focusing on the intricate little details and more of the "why". At the end of the day, understanding the “why” is what actually moves the revenue dial. Also, how can we afford to get so deep in the weeds, when we have so many variables affecting the performance of Paid Search.

Here are just a few: 

• Site Performance
• Product Inventory
• Competitive Bidding
• Offers/Promotions
• Product Pricing
• Customer Service

Trust me, that is enough to qualify for a MIDOL moment.

Lack of Simplicity:

In all of this, we have lost Simplicity. The simplifying of highly complex reporting and analysis that tells an over-arching, high level story. We need to focus on the obvious and fix the big things first, then focus on the smaller things that require the deeper analysis and the "kudos" from your bosses and colleagues.

Analytics – Giving you more than you need

Analytics companies strive on providing you more than what you could ever want so they can get that edge on the competition. But again, DO YOU really need everything? Here's my real-life analogy (I hope the master of analogies is reading this article - you know who you are!)

You can buy a Lexus because it has wipers on it's headlights. But does that really matter? Are these wipers affecting your ability to get from Point-A to Point-B? Ummn, I don't think so.

SEM & Analytics - The Love vs. Hate Relationship Continues

I figured a little personal anecdote would be best for this post. At my wedding, my brother-in-law gave a speech in which he said to 250 guests. "In this world, most people see either a glass of water that is (1) Half Empty or (2) Half Full. And then there's Greg who sees Chocolate Milk." Well, if your wondering where I am going with this, I am trying to make the point that analytics is playing a major part in the decreasing interest of outsourcing paid search. Think about it, Search Marketers, Directors and VP's of marketing are now realizing that they are limited in what they can afford except for their brand or trademarked names. Everything else that is non-brand is now identified as unprofitable for them based on the KPIs (ROAS%, CPO, CPA, etc..)  So now, paid search which evolved from a traffic PWYG (Pay what you get) model, to a quality score and metric based model where web analytics is (for the most part) is revealing more wounds than band-aids of what is profitable.

From my experience, I have seen a trend in almost every client where their Brand Name is one of only a few campaigns that are actually turning into a profit based upon their margins or set ROAS%. Furthermore, it is this exact scenario that is forcing most companies to focus more on full blown SEO (Social Media, Blogs, Universal Search) to drive qualified traffic at no cost (except for the search agency fees if applicable)  rather than SEM.

But is it really just a cost/performance problem? Nope, it gets much deeper than that. Its more of a "SEM worthiness" business model problem. It really depends on the industry they are in, competition, affiliates, resellers, and most importantly product margins. Unless you have killer margins in an opportunistic niche market, SEM & Analytics can be a double edged sword (I prefer Lightsaber) and it's a battle that needs to be fought.