(Text and URLs blurred to protect the innocent).Tags for This Post: Friday, Manage, caption, Agency, Fun, campaigns, Text, Copy
During the course of evalutions of client PPC accounts today we came across an example which so perfectly illustrates the importance of testing that I just had to share it.
This client had been managing their own PPC campaigns before they came to us. Up to that point, they had spent thousands of dollars with not much to show for it (except, as the client so aptly put it, “a whole lot of learning experience”).
We took over management of this account about 6 months ago. One of the primary tactics we employ with every account we touch is constant split-testing of ads and landing pages. The table below (sanitized to prevent outing the client) shows the ads we split-tested in a certain campaign in reverse-chronological order.
At the top is the most recent (and active) ad, the ad just below it is the one it beat to become the new champion. That ad beat the one below it, and so on.
At the bottom of the table is the ad the client had written themselves.
The important items to note are the 4-fold increase in click-through-rate, and the relative stability of the conversion rate.
The current top ad shows a 9% conversion rate, but when I grabbed the snapshot the conversion data had not yet updated, so 2 days worth of conversions are absent from this figure. The actual conversion rate of this ad is 11.89%
This particular client tracks leads as conversions. Our research over the past 2 quarters has shown that a lead is worth about $11.74 in profit after subtracting the COGS and Advertising expense.
As you can see from the table below, if the new ad had been running for the entire 6 months, it would have produced $4,837 in profit (an additonal $3,862).
What if we had stopped at the ad that produced a 3.74% CTR? The client would still have reaped an additional $3,000 (almost).
And this is just one AdGroup, and not even the most-profitable or highest-volume group. This kind of testing goes on in all of their campaigns.
In some cases, yes. This client still would have made $974 in profit over these 6 months. However, it sure makes it easier to smile when you know you’re doing nearly 4 times better than you would have.Tags for This Post: learning experience, whole lot, conversion rate, reverse chronological order, relative stability, cogs, conversion data
An interesting thread regarding SEO companies developed last week on Google Webmaster Central which I feel is worth commenting on.
The raw thread is here:
I was first alerted to the thread from this post at SERoundtable: Don’t Force Google’s Hand, Especially If You’re Sketchy
A potential client of an SEO firm called BetterPlacements posted on GWC asking if any one had heard of this company. The firm had offered to “guarantee” a #1 ranking within a 50 mile radius of the client company’s offices. They also claimed to have a partnership relationship with Google, with exclusive access to this new program. The potential client, rightfully skeptical of the claim, posted on the forum asking for advice.
After some responses by Google Rep John Mu stating that no SEO company can “guarantee” rankings and that Google does not partner with SEO companies (and pointing to official Google guidelines stating the same), the user ‘bruben25′ joined the conversation, ostensibly to play devil’s advocate concerning the wording of the guarantee.
After several more (increasingly hostile) responses from bruben25, Google Spam Team leader Matt Cutts responded, asking bruben25 if he had any affiliation with BetterPlacements. bruben25 denied any direct involvement with the company as a client, employee or executive member. Cutts then laid his cards on the table: the business filings of BetterPlacements in Ohio, with the managing agent’s name listed as ‘B. Ruben’ and pointed out that the contact email for the company was listed as email@example.com, and his handle also contained a ’25′.
I felt there were a couple of items I could add to the discussion surrounding this company’s (possibly temporary) public implosion:
As a potential client of an SEO firm you need to realize that there is no difference between an SEO business and any other business. You need to look at the situation, and not just what the company is offering. One of the first things the potential client in this case posted was the company’s contact info:
5763 Talmadge Rd Ste B1 – B2, Toledo, OH 43623
You can see from Google Maps that the address listed appears to be an office building of some kind: http://umktg.com/sgsd so score one point for BetterPlacements.
Be wary of anyone who works from home unless they are explicitly billing themselves as an individual consultant or contractor. Office leases show commitment to being in business for the longer term.
The domain betterplacement.info is the first warning sign. .info sites are much cheaper to come by than .coms, .nets, or .orgs. Unless the .info is a highly desirable keyword combination (like seo.info) there’s no other reason to select a .info over a more traditional TLD.
.infos can indicate a high rate of burn or turnover on domains as a company changes its identity. They are also popular among scammers outside of the US, where US$8 can be quite a difference in budget (.infos are typically $2-$3 per year in registration fees, with .coms being closer to the $8-$9 range).
The second warning sign is the gmail.com address. I work with a few professionals who use their gmail.com address successfully as their business email, however, I am always wary when I see a free email account on a business card.
The first reason is that, like the .info domain or using a home address, it doesn’t indicate a commitment to longevity. Again, this doesn’t mean that someone is un-trustowrthy or un-dedicated (I used a home address when Untouchable first started out), but in light of claiming to be “Google’s partner” it should raise red flags. Google has $33 billion dollars in cash assets. If they really found an interesting company in Ohio with a .info domain and a Gmail contact address, working out of a small office building, why wouldn’t they buy them outright instead of “partnering”?
The second reason is more specific to the industry. You may know that Google offers their “apps” to companies to be white-labeled. Untouchable Marketing’s email system runs on the gmail platform, and although the setup might be beyond the everyday user, anyone who can set up a domain and hosting should be able to run a gmail interface for their domain emails. The excuse “We use gmail because we like the spam filter and interface,” doesn’t fly. The setup takes 20 minutes and you can have the interface and spam features of Gmail, and the branding power of sending and receiving at your own domain.
If you ever receive offers from companies like this, remember that due diligence is the same in any industry. How would you feel if someone told you they were an authorized Toyota dealer, and had a top-secret new Prius that got 100mpg and went 0-60 in 2.3 seconds, but when you arrived for the test drive they just had a few pickups parked in their front yard?
Look at the business behind the promise, evaluate the claims, and ask for references or case studies. Any legitimate company should be able to provide any of these in no time. Asking a forum for their opinion might help (As in this case) but imagine if the Googlers had not been there monitoring the conversation? bruben25 may have convinced the client that his company is in fact a partner of Google.Tags for This Post: mile radius, webmaster central, partnership relationship, seo firm, cards on the table, matt cutts, business filings, devil s advocate, managing agent
A recent comment in my Facebook feed reminded me to check trending topics on Google:
“If I had to wikipedia “G6″ to find out what it means, does that mean I’m too old to listen to the song?”
This song has been everywhere recently, and I can’t even count the number of times I’ve heard someone ask “What’s a G6?” or “Do they mean the Pontiac?”
I was interested in learning a couple things from this case-study in motion: 1. How far does this phenomenon extend into the general, search-engine-suing public? and 2. How has Pontiac performed in maintaining their association with the product?
(In case you’re not familiar, the Facebook comment is in reference to a recently-released song that is getting a lot of airtime. Although the chorus contains quite a few references to other brands and bands, the most focused-on line is “Feelin’ so fly like a G6″. In this case, ‘G6′ refers to the Gulfstream G650 private jet, which might be considered ‘fly’ in a few different ways. Clever, no?)
Firing up the Google Trends machine, we can see the following chart for the search ‘g6′:
The first thing to note is that the search volume in the most recent 2-3 months has been an order of magnitude greater than the previous, “background” search volume. So the “What is a G6?” question extends pretty far into the mainstream, I would guess. To get some perspective, let’s compare this term with a “known quantity”: Audi’s R8.
As you can see, searches for ‘r8′ have been fairly stable over the last few years, with just a slight peak around the time the car was actually released. In fact, search volume on both terms was almost identical at the song’s introduction, with the difference probably explained by the fact that the R8 is more widely-known internationally than the Pontiac G6 (sorry Pontiac).
So, if you want your product to get 10x the searches it does now, you need a quizzical reference in a hip-hop song. But what has Pontiac done to convince the searching public that the song is actually about their sedan, and not a $60 million dollar private jet?
The first result for a search of ‘g6′ (aside from the music videos for the song) is a Wikipedia page, disambiguating the term ‘G6′. The sitelinks for the page provide links for the Gulfstream G650 and Pontiac G6 right next to each other. Score one for Pontiac, though the pictures on the Wiki-page aren’t exactly flattering.
The next link is a CNN article discussing the song’s impact on Gulfstream’s image. Of note here is that the author (or web editor) chose to make the title of the page “Fly like a G6″. The perfect phrase to capture searchers looking for more information on this confusing lyric. And when you’re CNN, you can rank for a lot of terms. But so could Pontiac, I wonder where they are…
#3 blogs.citypages.com… “Like a G6: What does it mean?” <-- Great title to get clickthroughs.
#4 Image search: all Pontiac cars, score another for the General (or "Chief" in this case). Unfortunately, one is from "pontiacg6info.com" an obvious Made-For-Adsense play, one is from Flickr, and two are from auto-magazines. No official sites in the bunch.
#5 UrbanDictionary.com "G6" entry. No help there. (Although the entertainment value of these pages is high, you get insight into the rationalization process behind song-listener's interpretations, for example: 'G6 can describe any person, place, or activity as long as it meets the requirements of being fly, no lame usages will be tolerated.')
#6... Finally! Pontiac.com with the stellar title: "Pontiac G6 - Four-Door Sports Sedan | G6 Coupe - Sports Coupe | G6 Convertible - Hardtop Convertible | Discontinued | Pontiac"
Ouch. Not only is someone over at Pontiac actually trying to rank for "four-door sports sedan" (which does not have enough volume to show up in Google's keyword volume estimations, by the way. Google suggests you search "four door porsche" instead), but their title also contains "hardtop convertible discontinued".
On second thought, that might just be a static part of Pontiac.com's title. I've been a Pontiac fan since my first car (a Pontiac Sunbird), so I know that they are famous for discontinuing anything that gets popular (the GTO, the Firebird, the Fiero...), typically in favor of ridiculousness like the Aztek, Trans Sport, or the Vibe. By the way Pontiac, if you're reading: I drive a VW now.
Since the title is so full of other garbage, why not just throw "Fly like a G6" in there? You own the trademark (I assume) so it's not like someone can sue you over the meaning of the song's lyrics. At best you might convince a few of the millions searching out there that hip-hop stars think your cars are "fly". At worst a few marketers will think you're trying to rank your site for something other than no-traffic, generic terms and news results about your discontinuations.
It would certainly make me feel better as a shareholder.*
Some takeaways here: there are many ways to take advantage of a social-media phenom in your space, even if it’s not about your product.
1. Get the major properties on your side: Wikipedia, Amazon, UrbanDictionary… etc. Make sure you have a mention somewhere on all of them.
2. Adjust your title-tags, don’t just set and forget. Web pages were designed to be living, evolving documents. Re-evaluate them constantly for opportunities. Google and Bing like sites that change regularly anyway.
3. If you’re in an industry with a high-amount of image search volume (cars, celebrities, yachts, jets, flowers, desks, tattoos… the list is huge) get your images tagged and into Google Image Search. The new interface shows your content behind the images, so if you can provide a compelling reason for people to close the lightbox, they will.
*American taxpayers owned over 60% of GM (Pontiac’s parent company) until this week.Tags for This Post: order of magnitude, search volume, google, background search, fact search
I’ve been carrying around this post for quite a while, waiting to have time to give it the attention it fully deserves. The catalyst for writing it now was an article on the New York Times about how Google wants to become your “answer engine.”
(I can’t seem to find the article that I read now, but it probably doesn’t matter since it’s been a few days and it’s probably behind the paywall now anyway.)
The jist of the article is that Google wants to refine their search to the point that a user can type in their idea of what they want, and Google can anticipate their actual need.
The most simplistic example of this is asking Google to convert between units of measure. Typing “30 dollars to pounds” will get you the current exchange rate. The next step is that when you type just “dollars to pounds” Google can anticipate that you are seeking a conversion rate (even though you didn’t use the keyword “conversion rate”) and give it to you.
The final step is that they give you the conversion right in the search suggestions, so you don’t even need to finish submitting a query.
Google wants to apply this logic to everything they can. Searching for “flight time from … to …”? They can anticipate that you might be looking for a flight and serve you a table of the upcoming flights to purchase.
So what’s the problem? The problem is the 99% rule. Just because something works 99% of the time, doesn’t mean it turns out just fine that other 1% of the time. Google is not infallible.
And my favorite (current) example of this searching for vegan recipes online. I tried out vegan-ism earlier this year for three months. I had to relearn how to cook quite a few things, and searching Google for recipes became a challenge.
In Google’s opinion “vegan” and “vegetarian” are interchangeable words. But while they may be similar, take the search query “vegan lasagna recipe“. There are fundamental differences between a vegan lasagna and a vegetarian lasagna (i.e. cheese and possibly cream).
But looking at the Google results, it’s pretty clear how Google views this query. Notice how “vegetarian” is bolded in several of the results? That indicates that the results matches your query!
(Something not shown in the image is that 4 out of 5 of the AdWords ads are vegan-only. It seems that Google’s advertisers are doing a better job of segmenting traffic than the search engine itself.)
But a site with a title like “Vegetarian Lasagnas – The Veggie Table – Vegetarian Recipes and Info” has nothing to do with veganism. In this case, the “answer” Google provides is nothing but a waste of time and potentially frustrating, especially when the user has been conditioned to see bolded results as “the answer” to their query.
The real issue is that Google knows that it doesn’t have enough “quality” results for “vegan lasagna” so they are bending the meaning of words to appear more relevant. In this case, a vegan might notice that the recipe calls for cheese before they cook the dish and dig in.
But how long before Google decides that “tesla roadster” is just another word for “lotus elise” and shows me where to put the gas in my car when I search “tesla roadster gas cap” (if you think that example is extreme then you should meet some of the people I know).
What about if Google decides that “Libertarian” means the same as “Republican”? Anyone’s blood boiling yet?Tags for This Post: time doesn, new york times, conversion rate, dollars to pounds, google
“Analytics says traffic is down! I thought SEO was supposed to INCREASE our traffic?”
UntouchableMarketing gets an email like the above (paraphrased) every once in a while. The first few times this happens it can be nerve-wracking, but I’m more jaded now and I know that 99% of the time there is nothing to worry about.
Google Analytics has attempted to boil down all the power the Urchin platform provides into digestible pieces for the lay web-user. The most prominent of these reductions is the dashboard view, showing the trailing 30-days performance of your website(s), in terms of visits, time on site, bounce rate, and completed goals.
Rounding out this summary is the “% Change” figure, a bolded number which is either red or green depending on if traffic, as compared to the preceding trailing-30, is down or up respectively.
Guess which number most users instinctively fixate on when they log in to analytics?
In most cases, when our clients witness the Big, Red, Scary Numbers, the explanation falls into one of three categories:
1. We recently added a filter to remove their office IP address, from which a large number of requests are made to the site daily. The filters don’t apply retroactively, so the previous month’s traffic looks greater in comparison.
2. The client had an article published on a popular site like Techcrunch or the New York Times which generated a large spike in traffic to their site for one or two days, which is counted toward the previous month’s total.
3. The client’s clients make use of pages on the client site to do things like: sell items from a store, register people for an event, or sign up for email/SMS updates. A client with a large user-base may have made use of the site for a big event, which has now passed.
Open analytics, click on ‘View Report’ from the dashboard. On the left side, click on ‘Traffic Sources’ then ‘Keywords’.
If you only want to check Organic Traffic, click on ‘Non-Paid’ just under the graph, to the left.
Change the date to a longer range and select “Monthly” for the graph aggregation.
Scroll to the bottom of the list of keywords and find the ‘Filter Keyword’ box. Change the first box to ‘Excluding’ and in the second box type in some words that you would consider “brand terms” e.g. your copany name, your website. Put them between parens and separate with pipes, like so:
Click ‘Go’ and the graph will change to show you all of the people who used non-branded keywords to search and find your site, in the selected months. The graph should be pointing up and to the right if your search marketing strategy is performing well.
Don’t be afraid of the Big Red Numbers, but make sure you find the right cause!Tags for This Post: dashboard view, bold numbers, digestible pieces, spike, techcrunch, scary numbers, google, web user
I was just reading a post by Aaron Wall (of SEOBook.com) about how successful people, especially in the Internet Marketing arena, seem to become jerks as they become successful.
It’s SEOBook, so obviously it’s a great post, but it got me thinking about another topic that comes up a lot: Untouchable Marketing’s Marketing.
It’s a pretty common joke around the UM offices that this company has never had to avail itself of its own services. Sure, we show up for searches like “denver web marketing consultant” and we’ve done the minimum due-diligence on optimizing our site, but the point is that exactly 0 of our current clients came to us via the contact form on our website.
So how did we get enough business to sustain the company? 100% referrals.
In Million Dollar Consulting, Alan Weiss talks about peer-to-peer referrals being the ‘Platinum Standard’ of marketing for a consulting firm, and I couldn’t agree more (he also mentions that you should carry a nice pen, which is something that I do not agree with, as the Pilot G2 is the finest pen ever invented).
The standard explanation for why referrals are great is that there is no better way to convince someone that you can deliver what you say, than to have his friends convince him for you.
However, I think referrals are great for another reason: screening.
In the world of business (as you can read in Aaron’s article referenced above) there are people who want the universe for a nickel. They will bargain, argue, and tear apart every single line item in a proposal. In the end, they aren’t happy with the results.
By getting a referral from a mutually-trusted acquaintance, you end up screening potential clients ahead of time. It’s unlikely that the small-minded “deal-makers” who want to nickel and dime you to death would be friends with your better clients. Their personalities are incompatible.
Personal referrals protect both the client and the consultant, and that is why I always recommend that you find a Web Marketing Consultant the same way you would find a good Lawyer or Accountant; because if they have resorted to spending money on advertising, then they need your business.
Keep in mind that the consulting world is very different from most other industries. If you have a repeatable process that would apply and show results for 90% of your clients (think software like Quickbooks, or an office cleaning service) then you really just want to get the word out to every applicable lead that you can. And that is why there is a vast market for what we do.Tags for This Post: alan weiss, marketing consultant, denver web, pilot g2, nickel and dime
The other day, as I perused the latest “nerd news” from news.ycombinator.com I came across a very interesting blog post that captured my attention. It was a summary of the properties of Benford’s Law.
By way of brief introduction, Frank Benford first became aware of this property while working at General Electric, when he noticed that pages of a book containing logarithmic tables showed much more wear on the page for the number 1, than they did for other numbers.
After considerable research, he found that any man-made data has a tendency for the leading digit of each number to be a 1 around 30% of the time, and a 9 only about 4% of the time. This is true with tax returns, baseball stats, and just about any other data you can name. This is very odd, because one would assume that the numbers 1-9 follow a normal distribution (and they do, when generated randomly).
This article stuck in my mind, and somewhere over the last couple days it connected with something else that had been bothering me.
By now, you probably know about the research that was done on the AOL Study Data that gave us a baseline for click-through rates on search results. This study showed that the number 1 result on a page gets clicked on about 35% of the time. The graph looks something like this:
Although this data has been very useful, what has always bugged me about it is that the study was only performed on the first 10 results, and so the only way to extrapolate a formula would be to graph the results, add a trendline, and use the equation of the trendline to predict other values greater than 10.
But take a look at this graph of numbers following Benford’s law (which has a known equation):
Eerily similar isn’t it? Putting the numbers side by side really makes it obvious that there is an interesting correlation between Benford’s law and SERPs behavior.
|Result #||% of total traffic||Benford’s|
First I took the equation for Benford’s Law and fed it the inputs of all numbers from 1 to 100. Unfortunately, beginning at #11, the numbers depart from what my gut feeling is on the actual click-through for pages beyond 1.
For example, the equation tells us we could expect a ranking at #11 to receive roughly a 3% CTR. Obviously, this seems high for the first result on the second page. But what if we assume that a similar number of people will click on the 10th result as will click on the ‘Next Page’ button?
Point of Clarification: The original algorithm provides for a 4.14% CTR on the 10th result. Assuming that another 4.14% of people will scroll down the page and click on ‘Next Page’ we use that percentage as the total amount for the next page.
By applying the same equation to the percentage of people who are predicted to click on #10, we see that roughly 1.14% of searchers will be predicted to click on #11. That seems reasonable, doesn’t it? I would say about 1 out of 100 times I will hit the second page of results and click on one of the listings.
Using this formula, it is predicted that searchers will basically not go past result #110, which is a pretty good prediction based on everything I’ve ever seen about search.
Of course, much of this is just arbitrary tinkering with numbers. But there is a great body of evidence that says that Benford’s law has applications to any human system of data.
Because search engines are merely aggregating data and applying a formula for ranking, it makes sense that it may be subject to some of the same underlying laws as the stock market, baseball stats, and other masses of human data.Tags for This Post: logarithmic tables, trendline, baseball stats, couple days, frank benford, normal distribution