Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

Heather has a mobile app she wants people to keep using once they’ve downloaded it. How can she make it more engaging?

  • Set up customize deep links
  • Use the bid strategy “Maximize engagement”
  • Add some large, memory intensive graphics
  • Customize the app for each format  (phone, tablet, computer)

irrelevantThat’s a great question from someplace other than the advanced search exam for AdWords certification, which is where it’s asked. While the fundamentals exam rightfully assesses knowledge of all the Googleverse’s channels, the search exam wrongly asks about the same non-search topics. Doesn’t Google stress the importance or relevance in absolutely everything? But not their certification exams?

Two others were recently spotted on the search exam that would be at home on either the fundamentals or advanced shopping tests:

Each of these are benefits you’d expect from shopping ads except

  • Free listings
  • More traffic and leads
  • Better-qualified leads
  • Easy of targeting without needing keywords

The format of a shopping as is different from that of a standard text ad in that it includes:

  • A product, image, title, and price
  • A product, image, title, price, and extension
  • A product, image, background color, and price
  • A product image, title, price, and merchant name

The question bank might use another scrubbing, with examples like these being re-purposed to beef-up the fundamentals.

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

Now that the seal has been broken and Artificial intelligence (AI) is an ingredient in paid search advertising, there will be several features we can expect it to improve in the coming years. Some are easier to improve than others, and the toughest one might be the automatic creation of ad copy for Dynamic Search Ads. Let’s propose a test that will signal the dawn of artificial intelligence’s domination over search engine marketing.

The Dynamic Search Ads Test

Boston University dog sweaterDynamic Search Ads (DSAs) are a keyword-less ad type. Google uses its organic search index to determine which term should trigger which ad for which product page. The moment our DSA is triggered, AdWords instantly generates new copy with a headline and final URL that links to the product page most matched to the user’s query. Within that moment, the headline is stitched-together dynamically with terms from both the query and the site content. This exists now, with limits.

There are keywords that Google could skim from product pages that it would have trouble matching to user queries. This is where AI might eventually become better than we mortals.  Imagine the titles of products in our online store are:

  • Boston Terrier cell phone case
  • Boston Terrier sweater
  • Boston University dog sweater

Maybe a human knows that a search for “Boston Terrier cell phone case” should return ads for cell phone cases that feature the likenesses of Boston Terriers, but AdWords might not be able to tell… yet. It could wrongly serve ads for pet products.

Before we decide that humans always know better, consider the others. Should a search for “Boston Terrier sweater” return ads for a men’s sweater depicting Boston Terriers or a dog sweater that’s sized to be worn by a Boston Terrier? Now even we naturally intelligent humans can’t definitively match this query to either product.

Knowing that the Boston Terrier is the mascot of Boston University, should a search for “Boston University dog sweater” return ads for a women’s sweater featuring Boston University’s dog mascot? Or for a sweater worn by a dog that features the Boston University logo? Neither the current technology behind Dynamic Search Ads nor we humans can conclusively match it to either product.

When an artificially intelligent AdWords can determine the search intent well enough to create Dynamic Search Ads from any product on any page, with no exceptions, then it has finally arrived. Why are human search engine marketers needed when AdWords can build DSAs for either a men’s sweater depicting Boston Terriers or a dog sweater that’s sized to be worn by a Boston Terrier, based on the query “Boston Terrier sweater?” When a machine can pass this Dynamic Search Ads test, then it will have proven its mettle and the next marketing disruption will be underway.

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

Artificial Intelligence has come to Search engine marketing, Google recently announced something new in the machinery of AdWords:

Smart Bidding is built on Google’s deep experience using machine learning…Smart Bidding can factor in millions of signals to determine the optimal bid, and it continually refines models of your conversion performance at different bid levels…It evaluates patterns in your campaign structure, landing pages, ad text, product information, keyword phrases and many more to identify more relevant similarities across bidding items to effectively borrow learnings between them.

Damn. They elaborate that additional signals like seasonality and price competitiveness are coming soon. So what else is coming? And when will artificial intelligence kill all SEM jobs?

The full transition to a bot-driven world will take decades, so anyone currently working will probably have enough time to retire before being forced to retire. Before that day, there are several checkpoints we can watch artificial intelligence pass on its journey to replace search engine marketers. AdWords will undergo a long transformation and the places where it can benefit from machine learning are the exact places we will see that happen, in ascending order of complexity:

  1. AdWords’ “optimize for clicks” ad rotation setting will begin to work. How many times has ad A produced a 1% CTR and been granted 10,000 impressions while ad B produced a better 3% CTR and only been granted 1,000 impressions? If this feature worked, then that wouldn’t happen. Artificial intelligence will solve this early on.
  2. The keyword diagnosis feature will begin to give consistent information that makes sense and doesn’t contradict itself. Currently, a keyword’s regular status may read “Eligible,” while the diagnosis status reads “Ad Not Showing” and the hover-state speech bubble expansion reads “Ad Showing Right now? Yes.” An end to this will come when AI is effectively put to the task of  diagnosing keywords.
  3. AdWords’ keyword planner will  produce something other than irrelevant oddities. AI gets real when the keyword planner gets good.
  4. The opportunities tab will identify real opportunities and suggest something other than “add keywords” or “raise that budget!” When it can offer insights that human SEMs miss, then AI will become the advisor.
  5. Google Display Network advertisers will see better Dynamic Image Ads. Right now they’re created with the help of the AdWords system’s scan of our landing page which will piece-together a sample ad with images and text found on the page. It often picks images and text that don’t make perfect sense, so some manual re-designating can be necessary. There’s room for improvement, a gap which will be closed by AdWords’ AI before it produces better creatives than we can!

Google has already done such an awesome (the word awesome applied purposefully and literally) job of organizing and making useful the world’s information without artificial intelligence. One more river has just been crossed and Smart Bidding is the very first step taken on the other side. The hills are next!

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

In addition to the recent introduction of specialty certification badges for agencies, the Google Partner Program also announced (in the same post) a new Premier Partners status for agencies. This draws a distinction from regular partner agencies, so what does it take to make the cut as “Premier?”

One qualification is there must be at least two certified individuals in the house. That’s right, two people must pass certification tests, as opposed to just one for non-premier partner agencies.

Google is an innovative group of minds who have a proven record of organizing and making useful so much of the world’s information. Can you guess what the other requirement is for an agency to be anointed “Premier?” You bet it’s their level of spending on AdWords!

While regular partners must spend at least US$10,000 every 90 days, premier partnership is conferred upon agencies “who manage a substantial portfolio of Google advertising campaigns.” There’s no word on how much that is, only that the thresholds “vary by country and market area.

The reason advertisers are interested in agencies that regularly handle similarly-sized accounts is the idea of cultural fit and familiarity with all manner of billing headaches. Enterprise-level advertisers want to deal with enterprise-familiar agency teams who are adept at handling the kinds of problems experienced at their level of spending.

Why not call agencies who spend a lot “Google Enterprise Partners” instead of the disingenuous “Google Premier Partners?”

So what do premier partners get, other than a listing in the Find-a-partner database? There’s off-handed mention that Premier partners “will receive increased support from Google.” But, large spenders already get increased support from Google, whether they concern themselves with partnership certification or not. So what do big-spending premier partners really get that big-spending agencies who don’t participate in the program get?

Permission to paste this logo on their site:

Google Premier Partner Badge

A large agency of thousands will automatically pre-qualify for premier status based on two certified people and significant spend. And how does all of this serve advertisers? They will now have one more piece of disinformation to help them fool themselves into thinking the individual people who service their account who manage their account are somehow “premier.”

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

Google Search Mobile Video Display and Shopping specializations

The Google Partners Program certifies search engine marketing and advertising agencies so that the credential can stand as evidence of SEM capabilities. There has historically been just one certification, but partner agencies can now be designated as having specialized expertise in up to five areas, which are conveniently the same five as individual certification: Search, Video, Display, Mobile, and Shopping campaigns.

If one individual has passed those five tests, then her entire agency would be certified as having expertise in all five areas. Or five people could each take on one exam… or whatever combination can be imagined. The point is that one person’s exam passage qualifies the whole company as a partner agency which an advertiser might assume means there are specialized skills across the entire headcount. It’s always been that way for general certification, the new development is that now an agency can fake five kinds of expertise, instead of just one.

Will this inflame agencies’ chronic capabilities exaggeration?

Advertisers expect agencies to routinely exaggerate SEM competence. and such stretched braggadocio is a big part of the certification program’s necessity to begin with. The biggest threat to the value of the Partner Program the prospect that an advertiser gets burned by the sloppy work of a certified agency. Is the chance of that lowered when an individual account manager is unlikely to be the same person who actually earned the entire company’s specialized accreditation?

There’s Just Gotta be a Better Way

To get and keep partners certification, an agency’s MCC (AdWords Manager Account) is already monitored  to ensure it meets the spending threshold to qualify for the designation and doesn’t raise too many performance-related red flags.

Why couldn’t that same MCC-monitoring bot also confirm that at least 50% of individuals with MCC-level access are certified in an area – in order for the whole agency to be certified in it?

Let’s assume they could. That would help reduce the number who exploit the program to exaggerate SEM competence. There just may be a way to support one of the original goals of the partners program after all.

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

How do ads perform against voice searches versus typed searches?

There’s no way to know. Google says 20% of all searches are voiced, while Bing reports 25%, but neither of their advertising platforms allow segmentation by this variable.

While voice search technologies are mostly supposed to return direct answers instead of search engine results pages with ads, that 20-25% of searches just mentioned aren’t going entirely un-monetized. “OK Google” (a phrase used to initiate a voice search) has begun appearing among search queries that trigger paid ads. There’s also been a steady rise in ad-triggering queries that begin with question-words common in voice searches like “who,” “what,” “when,” “where,” “why,” and “how.”

Ads show for voice searches, but there’s no way to directly measure them. Why not?

Is Google about to spend the next 10 years insisting that advertisers adopt a voice-first strategy while Google’s own platform won’t follow that same mandate?

If it’s treated like mobile, then yes.

The rise of voice search has been compared to mobile, and that may reveal a problem. It seems there’s been an announcement every January for the past 10 years that this was finally “The Year of Mobile.” Google complained annually that their mandate for the world’s publishers to “go mobile” wasn’t being taken seriously. But they were also stuck in the desktop rut.

It was just last month that Google finally allowed bidding on paid search ads to be based on mobile devices, instead of basing all bidding on desktop. So advertisers who echoed Google’s mobile-first evangelism for a decade were prevented by Google from taking action on this belief, until just last month. That doesn’t mean that Google has now completed the transformation. Their money-making flagship is still unusable on a mobile device, violating all the rules Google gives others on how to run a proper website. I know, there’s an app. But, “there’s an app” is not a reason for any publisher to not also “mobilize” their site. After a decade of browbeating others to mobilize their properties, Google still has yet to make AdWords’ site usable through a mobile browser.

Will voice search be like that?

  • When will we be able to segment voice searches to measure and compare them to typed searches?
  • When will there be different text ads served when the query was a voice search (a “voice-preferred” ad similar to a mobile-preferred one)?

Bing or Yahoo have the opportunity to beat Google to the punch by being the first platform to break-out voice searches. The simple ability to segment voice searches shouldn’t be any harder than the existing device-type segment, and can’t be all that technologically difficult to execute. It’s an easy win, all it takes to win the voice search battle is to practice what you preach!

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

An AdWords exam error message isn’t the most encouraging thing to see.

AdWords certification exams are presented as a sequence of individually-served questions. Answering one will load the next one until completion, unless the next one fails to load and is replaced with a server error notice:

AdWords exam server error notice

AdWords exam server error notice

If you see something like this, refresh the page.

I did, and all worked-out and life returned to normal. The subsequent question appeared, no time was lost, and all was right with the world again.

AdWords certification testing is run on Google’s own App Engine which boast of 0% downtime.

That must be rounded-down from a non-zero number.


Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

A Google survey of Partners program participants that closed over the weekend only asked one question, other than the intro qualifying items. The multiple answer choices reveal what kinds of challenges Google thinks agencies face. It was:

What are your agency’s top challenges for the next 12 months? Select 3.

  • Keeping up on reporting for clients
  •  Acquiring new clients
  •  Educating clients about the value of online advertising
  •  Adapting to AdWords product changes
  •  Time management
  •  Hiring talent
  •  Training new & existing Agency team members
  •  Achieving better ROI for your clients’ campaigns
  •  Facing tough competition from other Agencies
  •  Other [Please specify]

Doesn’t “time management” really mean “all of the above?” Some of these are only time consuming but not necessarily challenging to complete:

  • Keeping up on reporting for clients
  • Educating clients about the value of online advertising
  • Adapting to AdWords product changes
  • Training new & existing Agency team members

While others are potentially challenges because they can be hard:

  • Acquiring new clients
  • Hiring talent
  • Achieving better ROI for your clients’ campaigns
  • Facing tough competition from other Agencies

Of those actual challenges, the most interesting ones are “Hiring talent” and “Facing tough competition from other Agencies.” What would Google do about either of those? Wouldn’t responses with those choices just be non-actionable data for the survey makers? If results showed the biggest challenge faced by agencies was hiring talent, then would the Google Partners program have a solution?

It’s unlikely, but let’s hope they do!

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

The world’s Internet users are being trained to take advantage of advertisers who remarket to shopping cart abandoners.

There are online forums that list websites which remarket with discount offers like “Come back and save 15% this time!” Users are advised to visit your site, fill their shopping carts, and then leave without purchasing. They expect to see your remarketing ad offering a discount, which they’ll click-through to buy.

Remarketing with a discount offer isn’t all that creative or savvy. It guarantees some conversions attributable to it are actually forfeited profits to people using this technique. So stop it.

A quick visit to FTD’s site yields these ads immediately afterward…

how to loose money through remarketing

They should really stop conditioning the world’s Internet users to employ this technique!

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

Survey says: Google wants to position the GDN as a programmatic platform.

Recent surveying by Google at least ask if marketers think of the Google Display Network as a programmatic buy, and at the most aim to plant the idea in their minds. Three questions stand out:

Do you consider Google Display Network (GDN) to be a programmatic solution?
• Yes
• No
• Not sure

What best describes why you are not currently considering using Programmatic Solutions?
• No need to integrate third party data
• No need to integrate first party data
• No need for cross engine buying
• No need to push more in upper funnel
• Other [Please specify]
• Don’t know

Over the next 12 months, how likely are you to begin using programmatic solutions?
• Not at all likely
• Not very likely
• Somewhat likely
• Very likely
• Extremely likely

A link to the survey can’t be provided because they are individually issued (and identifiable).

An additional point of interest: A question asking how consultative of a position Google holds in the SEM’s mind compared to others:

Where would you place the below companies on a spectrum between the vendor and strategic partner for your business? Strategic partners are involved in your agency’s priorities to help you identify advertising solutions for your clients. Vendors provide advertising tools and troubleshooting support / expertise.
• Google
• Facebook
• Bing
• Yahoo

In a repeated Faux Pas, these surveys are all personally identifiable and tied to the survey taker’s personally identifiable information, although no notice is given of this. That’s researching bad practice number two after the push-poll technique of using questions to give information instead of get it (like giving us the info that GDN is a great programmatic platform).

Google would certainly insist their clients adhere to higher research and privacy standards than they  practice. That’s probably fine under the theory of Googlian Exceptionalism, which states that it’s OK if Google does it, as long as spammy marketers and national governments don’t.

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

The current bank of AdWords fundamentals exam questions includes a handful that have more than one correct answer. The problem? An uninformed test creator has coded just one as correct which hurts test takers who are left to guess which of the right answers is “the one.”

Let’s see four #AdWordsExamFails:

Your ad can show to a user when your targeted language matches:

  • A user’s operating system language
  • A user’s browser setting
  • The language of websites a user visits often
  • A user’s Google interface language setting

“The language of websites a user visits often” –  Let’s not forget about display. A user’s history of viewing GDN-participating properties published in a given language can mark her for ads targeted to that one even as she views pages written in another.

“A user’s Google interface language setting” – A user can also deliberately set his Google interface language (typically to view Gmail’s inbox in the preferred one), which would prompt AdWords to serve him ads in the targeted language even when he’s searching or reading content in another.

You would advise a client who’s launching a new product line to advertise on the Google Display Network because she can:

  • Reach people who are searching for her products
  • Reach people who are interested in similar products
  • Use text ads that encourage people to call her business
  • Use text ads that encourage people to visit her website

There are two categories of answer here. “Reach people who…” is about targeting and “use text ads that…” is about ad functionality. Our client can “reach people who are interested in similar products” on the GDN (because he was flagged for an interest category) with “text ads that encourage people to visit her website” and “text ads that encourage people to call her business” (call extensions also exist on the GDN). Which of the three correct answers is the right one?

What statistic indicates how often a click has lead to a conversion?

  • Conversion rate
  • Click-through rate
  • Converted clicks
  • Cost-per-conversion

Both “Conversion Rate” and “Converted Clicks” fall under the larger umbrella of “what statistic indicates how often a click has lead to a conversion.” The question isn’t worded specifically enough for only one of those to apply so both are equally correct.

Mimi wants to reach people searching for her baked goods, but only wants her ads to show during the hours she’s open for business. Which campaign type is a good fit?

  • Search Network only – all features
  • Search Network with Display Select – All features
  • Display Network only – remarketing
  • Display Network only – All features

Of the many differences between campaign types, what would make things better or worse for Mimi’s dayparting needs? Isn’t the feature available on all of these?

Are these the result of exam questions being written by someone who has zero experience with AdWords and simply crafts questions from help menu pages? Or are they the result of someone who knows the platform but fails at writing proper exam questions? Maybe under differing circumstances both reasons are correct but we can arbitrarily pick just one reason to designate as “the one.”

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Buffer this pageEmail this to someone

How many liars claim they (as individual people) fluently speak all seven Bing Ads accreditation language choices?

For the Find-a-Pro directory, Bing Ads accredited individuals can self-report fluency in one or more of seven languages: Traditional Chinese, English, French, German, Italian, Brazilian Portuguese, and Spanish. It’s terribly unlikely that any one person speaks those exact seven fluently, yet 94 people report they do! It’s a safe bet that each one of them simply checked all seven boxes when prompted to fill-out their Bing Ads profile.

The Bing Ads Directory provides a place for spammy search hucksters to identify themselves: Just look for those who claim fluency in all seven languages.

The directory lists the names of everyone making such a claim, for anyone on the Internet to see. What are the chances that a marketer who blatantly language-spams the directory is also the exact kind who advertisers are trying to avoid by seeking a Bing Ads pro from the directory in the first place?

Because they also select (just one) country of residence, we can answer the question of which countries are home to the most perjuring Bing Ads pros.

So where do these raconteurs live?



Are these accurate numbers that reflect the successful multilingual education of each individual who claims fluency in these exact languages? No way. Would a Chinese speaker who called each one on the phone be greeted with a pleasant introductory conversation in Chinese with each? Nope.

Simple count by country

(As of this posting)

United Kingdom17
United States13