PPC Without Pity

When creating a new PPC campaign or ad group, there’s an important choice to make with your text ads. Should you set text ads to rotate evenly, or optimize their distribution according to what the search engine thinks is best? Before you choose, you should think about how each distribution is calculated, and how it will affect your data collection for optimization purposes.

First off, you should always be running two to three text ads per ad group. That way, you can test different value propositions in your ads to see which one works best. After a month or two goes by, you can then compare the total conversions, conversion rate, and click-through rates of your ads to find the top performer. If your ad groups tend to get a lot of traffic, run three ads to test multiple factors at once. If your ad groups are low-traffic niche groups, you may need to run only two ads at a time to ensure you collect enough data for a reliable sample.

Choosing rotated or optimized ads will also affect your data collection. Rotating ads is pretty straightforward – each ad gets equal exposure. Optimized ad serving is where it gets a little trickier. By setting your ads to optimized serving, you allow the search engines to choose which ad gets the most exposure, since they will choose the “best” ad after a certain amount of data is gathered, then allow that ad to run the majority of the time.

This may sound good on the surface, but you also need to keep in mind that the search engines will choose the “best” ad based on their own criteria. This probably means that they will choose the ad with the highest click-through rate, since more clicks equals more money for Google, Yahoo, or Microsoft. This may not be in your best interest. The ad with the highest CTR may also have the lowest conversion rate. You could be racking up click costs without seeing conversion results if you trust search engines to “optimize” your ad distribution.

Of course, this is not always the case. Sometimes optimization is on the mark, and you can have a decent performing ad group without the hassle of constantly rotating ads yourself. For advertisers that just want a “set it and forget it” solution, you’ll probably be okay with optimized ad serving. But, if you are a serious PPC marketer, you will need to run evenly-rotated ads for two reasons. First, having even rotation ensures that you will have a valid statistical sample at the end of your text ad experiment, since all ads in the group will have an equal chance to succeed. And second, choosing not to rely on the search engines lets you choose your own success metric. If you want to optimize for conversion rate, you can do that. If you want to optimize for CTR, that’s okay too. You don’t have to rely on a third party to tell you what your “best” ad is, since you will have a much better idea of what is successful for your own business.

Most of the time, I choose to set my new ad groups to rotate evenly. But then again, I’m a huge PPC geek who loves to split-test things. Think carefully about what optimizing or rotating your text ads means for your business before you make a commitment. Don’t blindly rely on a search engine to automatically do your PPC campaign optimization for you.

This just in from the “I thought they had this already” department: Microsoft announces full HTML mobile device targeting for their AdCenter PPC marketing platform. Sure, Microsoft AdCenter has been in the mobile ad game for a while now, but did it really take them this long to work it out so that you can specifically target smartphone devices with full web browsers like the iPhone, Google’s Nexus One, or any number of other Palm and Android phones? Actually, they did have this capability previously, but you had to individually target each type of mobile device. At least now you can target to the general category of mobile devices, which should save you a lot of time.

Regardless of my snarky commentary, this is great news for any advertiser trying to make mobile-targeted PPC campaigns in 2010. Clearly, Google and MSN are betting big on mobile PPC marketing in 2010. Yahoo, not so much. I’m not aware of any ways to target mobile devices in Yahoo (feel free to prove me wrong in the comments!), but we’ve seen a lot of developments lately from Google and Microsoft. I’m still not totally convinced that users are going to be on board with mobile PPC just yet. I don’t think that people are ready to be advertised to on their mobile devices – it seems like a little too intimate of a medium to me. We’ll see if my cynicism is justified by the end of the year.

Either way, now would be a good time to make sure you have mobile-specific landing pages, and optimize a mobile version of your homepage for tiny smartphone screens. Whether you’re ready or not, people are going to start looking for information about your company on their mobile phones. It’s better to have a mobile version of your site ready to go and no one using it than having to scramble to make one once you see your mobile traffic take off in your web analytics program.

Yahoo recently announced some new functionality in their PPC platform. One of the new features was improved ability to import Google AdWords campaigns into Yahoo Search Marketing format. Now, I like the ability to cross-post campaigns as much as the next PPC marketer, but announcements like this just make me feel like Yahoo is constantly admitting defeat in their PPC platform.

I should preface this by saying that you can also import AdWords campaigns into MSN AdCenter, although you do need to take a side trip into Microsoft Excel to get it done. At least you have to use another Microsoft product. Yahoo just seems to be rolling over and admitting that there are much better tools for creating PPC campaigns out there, and you should probably just use those instead of the Yahoo Search Marketing platform.

I assume that most of the people reading this blog have created a Yahoo PPC campaign before. And you would probably agree with me that making changes to a Yahoo account are pretty tedious and boring. Compared to the agility and ease of use of AdWords Editor (or for that matter, the AdCenter Desktop Beta), Yahoo is light years behind the competition in terms of useful tools for large-scale PPC accounts.

So instead of creating a tool of their own, Yahoo decided to take the easy route and just assume that you are using their competitors tools (which you probably are) and outsource their PPC tool creation to the other guys. I admit that Google has some pretty awesome tools for PPC, but I hardly think that’s an excuse for just rolling over and giving up, especially if you’re one of the web’s largest companies. I guess I can’t really blame Yahoo for phoning it in, though, since their PPC platform is probably on the outs, pending the upcoming Yahoo/Bing merger.

If you are still interested in importing AdWords accounts into Yahoo, you can find instructions here. In the future, I’d recommend having a Google AdWords export handy for all of your PPC campaigns, since it seems like it has become the gold standard for pay per click accounts. I doubt that any of the other providers will be taking away their AdWords import abilities any time soon.

If your company is large enough and popular enough, your company name and trademark may suddenly become valuable pay per click search terms not only for yourself, but for your competitors as well. In some cases, your competitors may find that by bidding on keywords related to your company they can reach out to customers who originally intended to purchase from you. Obviously, this is a situation most of us would like to avoid. Unfortunately, most PPC companies do not explicitly ban the use of trademarked terms in PPC keyword lists. And why should they? The more keywords they make available for bidding, the more money they make. If you are having an issue with competitor use of your branded terms, you will need to resolve the issue yourself.

If you see your competitors running pay per click ads with your branded terms, the first step you should take is to contact the advertiser directly. That way, you can ensure that your key terms are removed from all advertising campaigns instead of just one PPC account. Be firm, but polite. There is no need to threaten legal action, but you should be clear that if they do not comply to your request you will register a complaint with the PPC advertising services, and their account will be affected. If your request is acknowledged, you may be able to resolve the issue quickly and easily with no need for intervention by the search engines.

However, if the offending competitor chooses to not take down the ads that are in violation, you may need to lodge a trademark complaint with the search engines. While it is not against the rules to bid on a competitor’s trademarked keywords, there are some restrictions on using trademarked terms in text ads. Be aware of this before you lodge a complaint. You can’t stop a competitor from running ads when users search for your branded or trademarked terms, but you might be able to prevent those competitors from using the terms in their text ads.

If you see some text ads running with your branded terms (and reaching out to the advertiser directly doesn’t help), you may need to issue a trademark complaint. This is about as effective as lodging a complaint with any other major company (that is, it’s kind of a crapshoot), but it may be worth your time if you really think that it is negatively affecting your business.

To lodge a trademark complaint with Google, first read their trademark policy here. If you still think that the advertiser is violating Google’s terms, you can send them a complaint by filling out their Trademark Complaint Form. If your complaint is valid, then your brand terms will be added to Google’s blacklisted terms, and future ads containing the keywords will be rejected through Google’s automated editorial process.

For trademark complaints about Yahoo ads, read their editorial guidelines here. If you would like to report a violation, you can send an email to trademarkconcern-ysm@yahoo-inc.com. Be sure to include the following information in your email:

1. The search term(s) that caused the ad in question to appear.
2. The trademark on which your claim is based.
3. The registration number of the trademark you own (if it is registered at the U.S. Patent and Trademark Office).
4. Evidence of consumer confusion arising from the offending ad.
5. A copy of any communication you have had with the offending advertiser about the matter.

If you have a complaint about an ad on Bing or another Microsoft network property, you can read their guidelines on intellectual property at this link. To lodge a complaint, you can fill out their trademark complaint form.

Keep in mind that all search engines explicitly state in their terms of service that they are not responsible for mediating trademark disputes. However, if you have a compelling case and the offending advertiser is clearly in the wrong, the search engines can be a great help in standing up for your intellectual property rights. Just be civil, communicate with the advertiser directly first, and use complaint forms as a last resort. As long as you remain respectful in your request, you’ll find that protecting your brand in the PPC marketplace is simple and effective.

Now that we’re firmly planted in 2010, I think it’s time for a few predictions for the coming year. And, since this blog is quite obviously concerned with pay per click marketing, I’m going to stick with what I know. Here are what I think the big three developments for PPC will be in 2010:

1. Yahoo and Bing will join forces, but still won’t match Google for PPC marketshare.

In July 2009, Yahoo and Microsoft announced a deal that would eliminate Yahoo search and replace it with Bing. Of course, as in all major business transactions, this deal has taken quite a while to materialize. Yahoo and Microsoft just finalized their deal in December 2009, and are anticipating a rollout of the new functionality in early 2010.

This is obviously big news for the PPC world. With Yahoo Search Marketing leaving the market space, that only leaves two major PPC providers – Google and Microsoft. With billion-dollar budgets at their disposal, this is sure to be a corporate slugfest for the ages.

However, I’m convinced that Google is still going to come out on top on this one. Let’s crunch some numbers. An August 2009 study by Search Engine Watch indicates that Google gets about 65% of total searches, Yahoo gets 15%, and MSN/Bing gets about 10%. Looking exclusively at PPC market share, a Rimm-Kaufmann group blog in March 2009 shows that Google dominates with 80% of PPC market share, while Yahoo gets about 15% and MSN/Bing lags with only about 5%. Even after Yahoo and Bing join forces, they will only get about 25% of web searches and 20% of the PPC money out there. I’m optimistic about Bing, since it has showed some great momentum since it’s launch, but I think we need to be realistic here. Google will dominate PPC spending for quite some time, since their company name is synonymous with web searching in a lot of web users’ minds. Microsoft is going to have to do a lot more than just take over Yahoo’s search market share if they want to put a dent in Google’s profits. This won’t happen any time soon, and certainly not in 2010.

2. Mobile PPC spending and use will continue to increase.

This is a pretty obvious one. More smartphones are being purchased with every passing year, and with that comes an increase in mobile internet use. Google made a big bet on the mobile ad market in late 2009 when they announced the acquisition of mobile ad provider AdMob. Google has also made significant improvements this year to mobile ad tracking in their Google Analytics platform. Yahoo offers a robust mobile advertising platform for display advertising, and Microsoft inked a deal to be the exclusive search and advertising provider for Verizon mobile phones. With the big three making such big pushes into the mobile advertising space, we’re sure to see some interesting developments in the coming year. The only question that remains is how will users of smartphones react to the encroachment of advertising on their mobile experience?

3. Social network PPC advertising rises, then falls flat.

Social networking has definitely been on everyone’s mind in 2009. With the explosive growth of Facebook and Twitter in the last year, advertisers have been wondering how to capitalize on the user bases of these extremely popular sites. To better monetize their products, both Facebook and MySpace have created advertising platforms to businesses who want to reach out to their users.

This is a really tempting proposition. Social networks have access to an unparalleled amount of demographic data that allows for hyper-specific targeting. On the surface, this seems like a great deal, and I think it’s going to attract a lot of advertising dollars in the coming year from businesses who don’t understand social media, and are looking for a shortcut to get presence on these sites. However, from my personal experience (and a few other folks I’ve talked to), social media ads tend to get really terrible results. You can already target MySpace, YouTube, and a few other social sites through Google’s content network, and let me tell you that the results aren’t pretty. You certainly get a lot of impressions due to the massive amounts of users, but you’ll also get a lot of errant clicks. What you probably won’t get, though, are conversions.

People go to social network sites to be social – they’re not there to buy things, fill out lead forms, or learn more about companies. This is the inherent flaw in any social network PPC campaign. Sure, you can pick up some good brand recognition through clever banner ads, but I wouldn’t expect anything good from text ads. The reach is good, the demographic targeting is good, but the user intent just isn’t there.

So here’s what I think will happen in 2010: businesses will get sucked in by the promises of awesome demographic targeting, and consultants telling them they need to get into “the social media thing.” Then, after a few months of mediocre results, they’ll realize that the ROI just isn’t there and they will pull the plug on their social network PPC campaigns. 2010 will be the year of boom and bust in the social media PPC space, unless Facebook and MySpace can figure out how to make ads more engaging to their user base, and more profitable for their advertisers.

There you have it – three PPC predictions for the coming year. We’ll see if I’m right this time next year. Got some predictions of your own? Let’s hear them in the comments.

So you’ve done your keyword research, and you think you have a pretty good, relevant set to run with. But you still have a few choices to make. For starters, what type of keyword match type should you use? Here’s a breakdown of all the PPC keyword match types, and the kinds of situations they are useful for:

• Broad Match: The default match type for the search engines, and for good reason: they generate a lot of clicks, which means a lot of money for the search engines. Having keywords set at broad match means that the search engines will make a rough approximation of what they think is relevant to your chosen keyword. And they will probably make a very liberal interpretation of what is relevant. I’ve pored through search query reports and seen broad match furniture keywords show ads under a search query for housewife porno. Do you really want your ads showing up for everything? Stay away from broad match if you’re on a tight budget and can’t afford any extra wasted clicks.
  
  Broad match does have it’s uses, though. If you’ve got a large enough budget and can afford to waste a little of it in the name of research, broad match can be a fantastic keyword research tool. Since it picks up everything, including those super-obscure long tail keywords that you probably wouldn’t think of targeting, you can run a set of broad match keywords for 30 days or so and then analyze your search query reports to see what you missed in your initial keyword set. Then, you can add those missed keywords in, change everything to a more specific match type, and run a more efficient campaign.




• Phrase Match: Phrase match is a good middle-of-the-road solution to the match type problem. Under phrase match, a user’s search query must match the keywords you specified (including word order), although your ads will also show if they add more keywords before or after the keyword(s) you specified. For example, if you were bidding on “pizza delivery,” your users would see your ads if they typed in “pizza delivery in austin,” but they wouldn’t see your ad if they typed in “delivery pizza” or “pizza fast delivery.”
  
  Phrase match is good because it filters out a lot of that weird, marginally-relevant traffic that broad match brings in, but it still allows you some leeway to overlook some terms that a user could type in. It’s a good idea to start out new campaigns with phrase match, and then switch later to either broad or exact match if you decide that you need more or less traffic to your site.




• Exact Match: If you’re having a lot of trouble with unqualified traffic, exact match is the way to go. Keywords set to exact match will show ads when the user types in a search query that matches your keyword, and nothing more. You’re going to block out a lot of traffic with this setup, but if you see a lot of junk traffic coming through in your search query reports, sometimes you need to take drastic measures to keep your cost down.




• Negative Matches If you see a few troublesome search queries in your reports, but don’t want to take a measure as drastic as setting everything to exact match, then some creative use of negative keywords may be in order. Negative keywords prevent ads from showing whenever a search query contains a keyword listed in your negatives. You can also set negative keyword match types for more precise targeting – negative match types have the same effect as the ones mentioned above.

Yahoo Search Marketing has some special match types, since they don’t use the broad/phrase/exact standard that Google and MSN adhere to. Here are the Yahoo match types:

• Standard: Standard match functions almost like the aforementioned exact match type, but allows for plurals. For example, if you were bidding on the keyword “lamp,” a user that typed in “lamps” would not see your ad under the exact match type, but would see it under the “standard” match type.




• Advanced: Advanced match type works like the broad match type. Your ads will show if a user misspells their search query or appends additional words before, after, or in between the keyword phrase you’re bidding on.

As with any pay per click marketing strategy, it’s important to test out different match types to see which one works best for each campaign. These instructions should give you a good starting point, but don’t forget to analyze those search query reports to make sure your keywords are driving the kind of quality traffic you need.

By now, you’ve probably heard about the Yahoo/Microsoft online advertising deal. Yahoo search is history, replaced by Bing. Yahoo PPC accounts will be merged into MSN AdCenter, except for display and content advertising, which will remain independent to each company.

So what does this mean to us PPC folks? Obviously, the usual online advertising triumvirate of Google, Yahoo, and MSN PPC advertising accounts is gone. Those of us who bothered to deal with MSN AdCenter are rejoicing, since we might actually start to get traffic in our accounts. Also, I get to stop complaining about how much Yahoo search marketing’s user interface sucks.

All kidding aside, this is a huge development. Google is still going to be the king of paid search, but combining Yahoo search traffic with Bing search traffic (which seems to be picking up, thanks to Microsoft’s PR blitz for their new tool) will offer a bigger chunk of the search market than we’ve seen with the respective traffic numbers of Yahoo or Bing alone. This could mean a big influx of advertisers to the MSN AdCenter platform. It also could be the end of good cost-per-conversion in your MSN account, now that there might actually be some serious advertiser competition on there.

This deal also benefits advertisers thanks to MSN’s superior account management tools, like their AdCenter desktop client. I’ve been promised a Yahoo desktop app for at least a year now, but I’ve yet to see it. Obviously, that’s out the window entirely. With AdCenter benefiting from the influx of Yahoo traffic, and Yahoo accounts getting the benefit of decent account management tools, it’s going to be easier to manage larger accounts in the new, combined platform. This means increased competition, but potentially better results from your Yahoo traffic.

What I’m most concerned about is how users will react to this. Are there really any true fans of Yahoo search out there? Will they be disgusted with the merger and make the switch to Google search? Or will there be a smooth migration of Yahoo traffic over to Bing? We have until early 2010 until the deal really goes into effect, so that’s a lot of time for Microsoft to ramp up their Bing PR campaign, and a lot of time for us marketers to mull over how this will affect our accounts.

No doubt some of you have heard of Wolfram Alpha, a new search engine set to launch this month. Word on the internet is that it could be a real Google killer – initial demos show that it offers a much more structured and relevant search on some search results. What’s more, the creator of Wolfram Alpha (Stephen Wolfram) has an impeccable resume – creator of groundbreaking software (Mathematica), Ph.D. in particle physics by age 20, etc. Obviously, Wolfram’s a pretty smart dude, but can he kill a search engine whose very name has become synonymous with search? And more importantly for us, will we need to start learning how to do Wolfram Alpha PPC when he decides to monetize it?

While I think that Wolfram Alpha is a pretty cool tool as advertised, I remain skeptical about its ability to displace Google as King Search. Anyone remember Cuil? The constant, fawning buzz that promised that it would deliver us from Google’s monopolistic grasp? Let me ask you this – when was the last time you did a Cuil search?

Despite Wolfram Alpha’s pedigree, I don’t think it will be taking any ad customers from Google any time soon. First, it seems like Wolfram Alpha is geared more toward the academic set. Not a very juicy demographic for advertisers. Do a couple of searches at Google Scholar and see if any ads pop up. Second, Wolfram Alpha isn’t as automated as Google is – Wolfram himself admits that he’ll need a team of at least a thousand people to curate the information that goes into his search engine. That seems to me like a system ripe for collapse, especially considering all the buzz that’s being built up pre-launch. I’m betting that we see a monumental crash on launch day, and intermittent service until interest cools.

Bottom line, Google is not a perfect system. But it’s the best system we have for online advertising for now. Wolfram is focusing on providing better search results, not serving relevant ads to consumers. I wouldn’t let your AdWords Certification expire anytime soon if I were you.

If you want more info on Wolfram Alpha, I suggest you check out this article here.

Here’s a neat little trick I’ve used to get a slight edge over ad competitors.

If you use granular geotargeting settings in Google or Yahoo (i.e. targeting to the state, metro area, or city level), the search engines append an extra line of ad text to every text ad to indicate the area you’re targeting. Here’s an example:

Note the addition of “Texas” as the last line of ad text. It works in Yahoo, too:

So what does this mean for advertising? Keep in mind that just because you aren’t in a physical location, it doesn’t mean that you can’t use geotargeting to make it appear that you’re a local vendor. People may be more likely to click on your ad if they think you’re a local. Also, any time you can get some additional factor in your ads that distinguishes you from the other results, it will generally get more eyeballs on your ad instead of the others.

A simple solution is to take off the default “United States” setting for your geotargeting, and target all 50 states at the state level. That way, you get nationwide coverage, but you also get that neat little line of geographic ad text on all of your ads.

Last week, I wrote about some changes going on with Yahoo demographic targeting. This week, I wanted to follow up with some more specific reasons why I think demographic targeting for pay-per-click campaigns is less useful than the search engines would have you believe.

There are three main reasons why I think web demographics are unreliable:

1. The inaccuracy of demographic statistics online – a great deal of user demographics that search engines use is pulled from third party data vendors, opt-in data, and algorithms that provide “estimates” of user data (source, source). None of these sources could possibly provide information with 100% accuracy. Third party data providers have an incentive to overestimate the value (and accuracy) of their data, opt-in demographic data is by definition incomplete (since you’re not gathering information on 100% of users), and I’d be surprised if my life could be quantified by a computer program. These are the building blocks that you would base your ad planning assumptions on. If you’re targeting 35-40 year old females, how can you be sure that any of these methods could accurately estimate the actual person who is seeing your ad fits into those categories? That brings me to…
2. One computer DOES NOT equal one person – I’m sure that there are plenty of households out there that share a computer. How many times have you looked up something on someone else’s machine? Despite all the big talk about demographic targeting, there’s really no way to truly target the person. Our ad process ends at the computer screen – you can’t control who is using it. You have no certainty that the user viewing your ad through the lens of the computer monitor is actually the owner of the machine, or the person that all of these fancy demographic programs have data on.
3. Reliance on voluntary data – Of the three sources mentioned in point #1, two of them rely on voluntary data – the third-party data (probably) and the opt-in data (definitely). For starters, you only reach a small percentage of people who will actually take the time out of their day to volunteer this data. I’m willing to bet that a vast majority of people prompted for this information either are too busy to fill it out correctly, or choose to provide inaccurate or invalid data due to privacy concerns. Then, you have to take into account the people who filled out the information, but provide misleading information either through carelessness or willful inaccuracy.

After you take all of this into account, what percentage of your audience do you think has completely accurate demographic information? Ten percent? Five? Maybe less? There’s probably no way to even know. Instead of relying on these vague demographics, do yourself a favor and spend more time analyzing the data that you know is correct – historical keyword reports and analytics tracking. These sources will provide a much greater wealth of insight than the lazy demographics put out by the search engines.