PPC Without Pity

When your PPC account reaches a certain scale, you may be tempted to start using automation software to help manage your account. There are a number of web software services that can help you automate your bids, test out text ads, or optimize your landing pages and display ads. At first glance, this seems like a pretty good deal. You don’t have to spend as much time in your account, and you get to use an automated system to do all the match and difficult thinking for you. Automated PPC management systems can be a boon for the overworked SEM manager, but they can also lead to big trouble if used incorrectly.

First, the pros. Once you start having thousands of keywords, bid management can be a real hassle. Perhaps the best aspect of PPC management software is the bid management functionality. It can be hard (if not impossible) to gauge the effect your bid changes have on your keywords, and to do the math on each individual keyword bid to ensure positive ROI and statistical significance in testing. At this level of scale, it’s best to get a computer to do the dirty work regarding your keyword management.

The same applies for ad testing. Unless you have a real passion for statistical analysis, you probably view ad text testing as a real chore. But, if you get a software package on the case, all you have to do is write the ad text and the computer manages the testing for you, spitting out some handy reports for you to go over.

These all sound like compelling reasons to break out your wallet and buy some software. But, many PPC managers view these software packages as a magic bullet that will solve all of their problems. This is simply not the case. No matter how sophisticated a software program is, all it can really analyze is quantitative data. It can’t research your customers and analyze their motives – you still have to do this. And, since the machine can only look at the numbers, it can sometimes make some grievous mistakes that could cost you revenue.

If left unchecked, automation can run wild. I’ve seen at least one example of an automated system going wild on bid reductions, eventually placing bids so low we stopped showing on the first page, reducing conversions down to zero. An automated system only performs as well as you tell it to. If you put in the wrong formulas, or leave the system unchecked, you could be in for a nasty surprise further along down the road.

Automated systems also lack the ability to make qualitative judgments, making them useless for creating ads or landing pages. Sure, they’re handy when you need to do some hardcore statistical analysis, but they can’t come up with new ideas on their own. You will always need a human being to come up with great ideas.

I’m not totally against PPC automation, but I do remain skeptical. I’ve yet to see a software system that can outperform a skilled human professional. Despite all the numbers and statistics, PPC is marketing, not math. If you are considering purchasing a software package for PPC management, ask yourself this question: will this system perform as well as a comparably-paid PPC manager? If you’re spending two, three, or four thousand dollars on a software system, your account might perform just as well by hiring a competent full-time employee or freelancer to manage your account. They will be able to make the qualitative decisions that a machine simply can’t do.

Long-time readers of this blog have probably figured out by now that I’m a bit of a Google fanboy. Understandable, because I’d probably be out of a job if it wasn’t for AdWords. For the most part, I’ve been pretty impressed by the changes and improvements they’ve made to their AdWords platform over this year. But recently, they’ve changed some things in the reporting features that have me fuming mad.

You might have noticed that you can no longer run account, campaign, or ad group reports in the reporting tab. Instead, Google has decided to move these features to the “Control Panel and Library” section of the campaign and ad group level views. To me, this seems utterly ridiculous. Why would you move a reporting feature from a section of your application titled “Reports”? Or, at least offer users the option to get their reports in both sections of the platform?

What’s more, the part where you actually get your reports is unlabeled and counter-intuitive. When I first found my reports missing, I checked the Google help file. It said to look for the reports in the “Control Panel and Library” section of the campaign tab. After I went there, I spent a good ten minutes looking for the small text link that would take me to the section (it’s on the left side of the page, halfway down, by the way). What really got me frustrated was that after I clicked on the link, there were no reports and no way to generate new ones. After a few days of yelling at my computer and finding no help from the “help” pages, I finally figured out that the solution to generating a new report was in clicking on an unlabeled button with an ambiguous graphic. Of course, even after all of this you can only export your current view in report form. No looking at stats by day or using other segmentation features that Google has rolled out over the last couple of years.

This reporting change screws over both veteran users and customers new to the AdWords platform. Vets lose functionality that they’ve grown accustomed to, and have to change their way of doing things unnecessarily. Newbies are bound to be perplexed by the terrible usability of the new feature and stuck looking frantically for their reports where they should be – in the “Reports” tab. All of this hassle for a feature that wasn’t even broken in the first place.

I really hope that Google wises up to their error and reverts their ill-guided change. You would think with the legendary obsession with user testing that Google supposedly has, they would at least split-test the reporting feature before they rolled it out. Unfortunately, it looks like somebody went with their gut instead of their head on this change, and we’re stuck with this ill-guided “improvement.”

If you’re really into pay per click marketing or web analytics, you have probably heard about the debate around first click vs. last click conversion attribution. So what’s the big deal? Why should it matter which keywords and text ads your conversions get credited to? The answer is more complex than you might think. Let’s break it down.

First, a few definitions for the unfamiliar. When analysts talk about first click and last click attribution, they are referring to which click of a keyword or text ad is credited for a conversion event in a search engine marketing campaign. First click attribution means that the first click in the process should get the credit, even if a user performs additional searches and ad views before they actually convert. Last click attribution means that the last click or ad view should get the credit, no matter how many other searches came before.

The problem is that most PPC metrics use the last-click model due to limitations in tracking. For example, Google’s AdWords tracking cookie expires after about thirty days. That means that if a user saw and clicked on one of your ads three months ago, remembered your business name, and converted under a different search query, then there is no way to measure that first click anyway. The only viable option is to go with last-click attribution.

However, with advances in web analytics software, first click attribution is no longer out of reach. This allows us to ask the question of whether the first click or the last click is more important in a pay per click transaction. When you look at search behavior, a sizable portion of conversions come via a process of multiple searches. People will search for one thing, click on a few ads, refine their search using information they found on the first pass, and continue this process until they find what they are looking for. Users may even click on several of your ads before they come to a decision. Each one of these clicks is important in pushing the user towards the conversion.

There’s no real right answer to this quandary. If you ignore first clicks, you’re ignoring the broader terms that make users aware of your offerings. If you ignore the last clicks, you’re omitting the valuable elements of your campaigns that close the deal. It’s much better to look at this problem holistically. Make sure that you are using keywords that target users in the different steps of the buying process – information gathering, comparison shopping, and highly targeted product terms for people ready to buy. Just remember that both first clicks and last clicks are important. If you leave one or the other out of your campaigns, you could be affecting the entire conversion process.

If you’ve played around with Google’s ad extensions, chances are you’ve come across their sitelink ad extensions. This feature allows you to add several links to pages on your site as part of your AdWords ad. Of course, there is a catch – the sitelinks are only eligible to show on the top-ranked ad for the user’s search query, and even then there’s still a chance it may not show. There are several good reasons to use sitelinks, and one good reason to not use them. This post will cover both sides of the issue.

First off, let’s talk about how to access the sitelinks option. Log in to your AdWords account, and select the campaign you want to add sitelinks to. Click on the “Settings” tab in your campaign menu. Under the section “Ad Extensions,” you should see an option for sitelinks. Click on the “Edit” link next to this option, and you should see a form like this:

Here, you can specify the titles of your sitelinks, and the destination URLs you would like them to go to.

Now that you know how to implement sitelinks, let’s talk about why you would want to use them. First, keep in mind that these sitelinks will only show if your ads are in the top position. If you’re not going for the top position, or can’t afford to make high bids, you might as well not bother with sitelinks.

Sitelinks can be helpful if you have multiple promotions going on. For example, a retailer running an e-commerce campaign in December could add links for Christmas Gifts, Hanukkah Gifts, and Kwanzaa Gifts. This would be especially useful if you’re bidding on some generic search terms, like “gift ideas.” You may not have a good idea what your user is searching for, so adding sitelinks allows them to self-select their route to your site and get a better user experience.

Sitelinks are also handy for branding campaigns. If you’re bidding on your own brand name or company name, you never know exactly what your users have in mind. Sure, you know that they’re looking for you, but what exactly do they need? Adding sitelinks to your ads essentially puts your site’s navigation menu within your ad, saving potential customers a step in their information-seeking process. Plus, since branded terms tend to have a low cost per click and high quality score (assuming that you do own the brand that you’re bidding on) you’re likely to get that top position and actually have your sitelinks show.

There is one very good reason to not use sitelinks, though. If you’re doing structured split-testing of your landing pages (and you should be), then adding extra destination URLs to your ads could really screw with your statistical analysis. If you’re running a split-test in AdWords, Google Website Optimizer, or some other split-testing program, it’s probably best to stay away from these ad extensions. It could really make things unnecessarily complicated for you.

Had any positive or negative experiences with Google Sitelinks Extensions? Let’s hear about it in the comments.

Google recently introduced a new reporting feature in AdWords called Search Funnels. This new reporting allows advertisers to see the sequence of keywords a user used before they completed a conversion. There are a lot more bells and whistles than that for advanced users, but in this post we’re just going to focus on the high-value basics to get you started.

To access this new report, follow these steps:

1. Log in to your AdWords account.
2. Click on the “Reporting” tab
3. Select the option “Conversions”
4. On the left side of the screen, you should see a link titled “Search Funnels.” Click on it.

You should be on a screen that looks vaguely similar to your Google Analytics interface. There are a lot of options here, but I want to look at the two most useful ones: Path Length and Top Keyword Paths.

Path Length

Many users type in multiple search queries before they finally convert. This report will show you just how many of these users visit your site. To view this report, select the “Path Length” link from the left-side navigation menu in the Search Funnels screen:

This graph will give you a breakdown of how many clicks users made on your ads before they converted. Results will vary by advertiser, but chances are that you will see the majority of your customers performed one click, a smaller percentage did two, an even smaller percentage made three, and so on and so forth further down the line. This may give you insight into how to make your AdWords costs more efficient. If you have an abnormally larger percentage of users performing multiple clicks, you may want to find a way to make your ads more targeted (i.e. state clearly your products are for business customers instead of consumers), or more closely align the message on your landing page to the one in your ad. If your customers aren’t finding what they are looking for on their first click, they’re probably going to do a different search and try again. If you meet that more specific need on the next click, you get yourself a customer, if not, you lost your opportunity on the first click. This Path Length report can help you diagnose this issue and take steps to solve problems with your account.

Keyword Paths

Now that you know which percentage of customers are using multiple keywords, let’s take a look at how to discover what those different keywords are. There are a few ways to access this data in the Search Funnels report, but for now let’s focus on top keyword paths. Select the “Top Paths” link from the nav menu:

Next, we need to filter this report down so that we’re looking at keyword paths only. Below the graph, look for some drop-down fields. We’re going to select “2 or more paths” from the “Path Length” drop-down and “Keyword Path (clicks)” from the “Dimensions” drop-down:

Now, you should see how user’s search queries changed (or didn’t change) as they refined their searches before the conversion. If you don’t see a lot of queries, try increasing your date range of the report. Chances are, you’ll see a lot of duplicates as a user searched the first time, mulled their purchase over, then returned with the same search query later. But, you’ll also see a few where a user started with a broad query (i.e. “computer training”) and then refined it after some initial research and consideration (i.e. “A+ certification classes”).

So how can you use this information? This report can give you insight on keywords that you thought weren’t working, but are actually helping other keywords convert. This is known in the industry as an “assist,” and so far, Yahoo has been the only search engine that actively reports it as a metric. These keywords may seem like they’re not working, but they’re actually pretty valuable. While you may not see these keywords getting conversions, they’re actually capturing customers early on in the buying process, introducing them to your company, and priming them to complete a conversion on a later search. Before you start pausing keywords in your account, double check your keyword path reports to make sure they are not assisting other keywords and adding unseen value to your account.

If you’ve ever had the privilege of working with a small business PPC account, you’ve probably had this conversation:

SMALL BIZ OWNER: Our ads aren’t showing up!
YOU: That’s odd…everything seems to be fine in the account.
SMALL BIZ OWNER: I just Googled our name and didn’t see our ad!

Then, you have a long conversation explaining how PPC actually works to your naive client, and hopefully get them to understand that having a PPC account doesn’t guarantee ad coverage on 100% of searches. I’ve actually been on the receiving end of the opposite conversation: when a client congratulates you on getting them a #1 Google result (clearly indicating they don’t know the difference between SEO and PPC). While it’s nice to get noticed for your good work, this may actually be a more harmful scenario for your client.

Here’s what you need to explain to your client (feel free to send them this post as a link). It’s totally impossible for Google to show every advertiser’s ads on the first page 100% of the time for every keyword they are bidding on. It’s a mathematical impossibility. That’s why we have factors like bidding, quality score, and Google’s algorithm to ensure everybody gets a chance to have their ad show up. So, depending on your daily budget, bids, and ad targeting settings, your particular ads get put in a queue to show up, along with the thousands of other AdWords advertisers.

Your ad sits in the queue until a user types in a search query relevant to a keyword you are bidding on. Then, the Google system goes to work, calculating whether or not to show your ad, and where to place it in the results. If the timing is right, your ad will show. Once it has shown up, it gets put back in the queue to wait for the next qualifying query.

Now here’s the kicker. If you just Googled yourself and saw your ad, you just took an ad impression away from a potential customer. This was an impression that might have lead to a click, which in turn might have lead to a conversion. If you sit at your desk all day, Googling yourself to check ad positions, you could be costing yourself money with every Google search.

If you’re worried about whether or not your ads are showing, don’t fret. Google will always let you know if there is an issue with your ads or keywords. If your ads aren’t showing, then Google isn’t making any money from you – they have an incentive to get your ads up and running as soon as possible after any interruption. If you’re worried about which position your ads are showing in, the stats on average position are right there in your account. You’re never going to be guaranteed a top spot every time (and in reality, this isn’t as important as you think), and you can pretty easily adjust what position your ads show in by raising or lowering your keyword bids.

If you really want to see how your ad shows up on a real-life Google search page, there’s even a tool for that. Just go to the Google Ad Preview Tool. Here, you can enter whatever keyword you like and see a sample Google results page. You’ll get results for organic and paid searches calculated by the Google algorithm, but it won’t count as an ad impression in your AdWords account. You can even use the tool to change your domain or geographic settings to see how users in other regions view your ads. This function is pretty handy if you don’t geotarget your own area in your ad coverage, or you have special campaigns for other cities, states, or countries.

Hopefully, this information will cure you of your self-Googling. Sure, we all give in to a vanity search now and again. But if you’re spending money on AdWords it’s best to kick the habit.

Having campaign budgets in your PPC accounts is a good thing. You can keep your spending in check and ensure that your advertising budget doesn’t go totally haywire. You can even target your spending so that certain products or verticals get a certain percentage of your budget, keeping your ad dollars allotted to the avenues that will make you money. But, since every PPC campaign has different content and goals, there’s no set formula for how to make the right budget. This means that sometimes budgeting can go awry.

I had an incident this week that reminded me of this fact. I added a new ad group to an existing campaign. This is usually a pretty innocuous change, but I didn’t count on how popular this ad group would be. I was pretty excited when I saw that campaign’s conversions double. Unfortunately, so did our CPA. In fact, this ad group drove so much traffic that it affected the overall conversion rate for the entire account.

Another consequence was that this new ad group was so popular that it sucked up all the impression share for the campaign that it was in. Since so many people were searching for the keywords in the ad group and clicking on our ads, we blew through the campaign budget pretty quickly. This meant that the new ad group was getting all of the campaign’s budget dollars and conversions (at a pretty expensive CPA, I might ad), and the rest of the ad groups didn’t even get a chance to show up. Ad groups that were converting at a pretty good rate last month had zero conversions month-to-date because they were being crowded out of ad display by the new ad group. This was leaving potential conversions on the table, which is never a good thing in PPC.

In a situation like this, there’s an easy solution: just separate the problem ad group into its own campaign, and give it it’s own budget to spend. There’s nothing wrong with having multiple ad groups in the same campaign, but if an ad group turns out to me extremely popular, it’s best to isolate it so that it doesn’t soak up budget dollars that could be going elsewhere. The toughest part is figuring out that there’s a problem in the first place. You may never know if a certain ad group is getting a disproportionate share of your budget unless you scrutinize each campaign at the ad group level.

So be careful out there. You might be spending too much money in an ad group at this very moment and you wouldn’t even know about it. This is yet another example of how pay per click marketers must be ever vigilant in our accounts. Follow the money, and make sure certain ad groups aren’t getting too much of it.

Well, folks, I’m on vacation this week, so no new post. I can tell you’re all disappointed. In the meantime, you can while away the summer hours by catching up on PPC Without Pity’s greatest hits of 2010 (so far):

1. So You’re Thinking About AdWords Certification…
2. Three PPC Predictions For 2010
3. A Critical Review of Trada PPC Services
4. PPC Text Ads – Rotate Or Optimize?
5. Advertising Pharmacy Products In AdWords

With the Yahoo/Bing integration just around the corner, there’s no better time to brush up on your MSN AdCenter optimization skills. In this next installment of Adjusting Campaign Settings Like A Pro, we’re going to walk through the best practices of MSN AdCenter settings.

To access your campaign settings, follow these steps:

1. Log in to your MSN AdCenter account, and navigate to the “Campaigns” tab on the menu bar:

2. Select the campaign that you would like to edit settings for.
3. Next to the campaign name, you’ll see a link that says “Change Settings.” Click it.

Follow these steps, and you’re in!

MSN doesn’t have many options for campaign settings, but there are still some useful tweaks you can make.

Campaign Settings

Here, you can edit the name of the campaign. Pick something that will help you remember the contents of the campaign, and it’s intended purpose (lead generation, e-commerce, etc.). I also like to tag new campaigns with a date code (i.e. 07-04-10) to let me know when the campaign started.

You can also toggle conversion tracking on and off via checkbox. You should always be using conversion tracking in PPC campaigns, so make sure this box is checked. You still have to install conversion code, though – you won’t get any statistics until that task is completed, even if you do check the box in your settings.

Budget Settings

Here, you can set your campaign budget. MSN offers two options, setting budget by month or by day. Setting it by day is good if you need to test out budget points when you’re not yet sure how much traffic you’re going to get. Once you figure out an average daily spend, you can set your budget to monthly.

There are two very important options here as well. You can choose to divide your budget across the month, or spend it until it’s depleted. If you’re on a limited budget, you may want to divide it across the month. This will ensure that your ads get even coverage throughout the month, even if you have a spike in click traffic one day that could potentially eat up your monthly budget. You might lose short-term clicks, but at least you’ll have ad coverage for the entire month. If you’re not so concerned about how much money you spend, then “spend until depleted” is the way to go. This ensures that you get maximum ad coverage, since MSN won’t be worrying about how to ration out your ad impressions to stay within your budget.

Campaign Targeting

In this section, you can change your targeting options for ad display. You can target by location, day of week, time of day, gender/age, and device.

In MSN, you can set your geotargeting settings to target countries, states, metro areas, or cities. Unfortunately, you can’t do custom targets or radii from a specific location. Hopefully they’ll change this after the search partnership goes through, and they can catch up to what Google has been doing for years now.

Time settings can be useful if you know that your target audience is only active at certain hours or on certain days. If you’re targeting business decision makers, 9-5 on Monday through Friday might be a good choice. Think about when your target audience is likely to be online, and then adjust your settings accordingly.

I’ve already explained why web demographics suck before, so I’ll just say use gender and age targeting at your own risk. Device targeting is a lot more straightforward, though. With this option, you can target your ads to computers, smartphones, or both. This is great if you want to make a mobile-targeted campaign, or if you want to eliminate traffic from busy mobile users who aren’t likely to convert on your multi-step widescreen conversion process.

Campaign Exclusions

This is where you put your negative keywords for the campaign, and your negative sites if you’re running a campaign with content network distribution. Look over your search query reports and placement reports carefully for sources of bad traffic. Once you figure out the search queries and/or placements that are costing you money without generating revenue, put them here so that your ads won’t show up on them ever again.

And there you have it. MSN doesn’t have many campaign setting options now, but perhaps we’ll see some interesting changes after the Microsoft/Yahoo search alliance goes through.

Last week, we continued our series on adjusting PPC campaign settings by looking at how to optimize Yahoo Search Marketing campaign settings. This week picks up where we left off:

Campaign Optimization Guidelines

In this section, you can adjust the settings for Campaign Optimization, if you opted in to this function within the “Campaign Budget and Schedule” section. You can tell Yahoo’s campaign optimization to work toward a certain business objective (clicks, impressions, conversions, or revenue), or try to improve your conversion funnel by assigning relative value to events within your funnel (i.e. a page view, shopping cart addition, or a lead form submission). As I mentioned in Part 1, I’m not very keen on using PPC providers’ account optimization functions since they only have the incentive to get you to spend more money on PPC – they could probably care less about your bottom line. But, if you just don’t have time to manage your PPC account and you’d like to automate the process, this may be a good option for you. But beware – if you set your optimization settings incorrectly, you just might amplify mistakes instead of solve them. Algorithms can run amok if you don’t watch them carefully.

Network Distribution

Here, you’ll be able to change your settings for ad targeting. First, you’ll need to decide whether to adjust your settings at the campaign or ad group level in “Set Level” box. For the purposes of this post, we’ll assume you want to keep the settings on the campaign level.

In the “Network Settings” subsection, you can choose whether you want your search network ads to appear in Yahoo Search (searches on Yahoo.com and Yahoo toolbars), Yahoo Partners (sites that use Yahoo for site search), or both. If you’re just starting out, it’s usually best to opt in to both to judge performance. Once your campaign has run for a little while, you can check back at this setting to see a performance breakdown by network. Then, you can make an informed decision on which network(s) to use.

Next, you’ll want to adjust your targeting. If you just want to reach all of the US (or whatever your home country is) at all times, you can just leave the defaults as-is. But, if you want to get fancy, you can adjust your demographic and geographic targeting, and schedule your ads to show at certain times of day and days of the week.

In the “Demographic Bidding” subsection, you can adjust your keyword bids for specific age and gender groups. For example, if your target audience is women ages 30-45, you can set your bids to be 150% of their normal amount for searches from that particular demographic. You can also block your ads from showing to users age 18 and under on this screen.

Now, keep in mind that web demographics aren’t an exact science. Since you can never really know who is behind the computer doing the searching, accuracy is a huge problem with demographic targeting. Yahoo is just making their best guess at the identity of the user based on if they are signed in with their Yahoo profile, or if they self-reported accurate information. Sometimes you’re relying on third-party data of dubious accuracy. Use demographic targeting if you have a very specific audience in mind, but be aware of its limitations.

Geo-targeting is much more reliable, since search engines can reliably read the IP address of the user making the search. Here, you can set it so that your ads only show in a specific country, state, or metro area. This is really useful if you only offer services at a specific brick-and-mortar location, or if you can only service certain parts of the country. You may also want to experiment with geotargeting if you know your best customers come from a certain part of the country, and you’d rather not waste limited budget dollars getting ad coverage somewhere else. You can even increase your bids in certain geographic areas if you find that certain areas have a more competitive PPC landscape than others.

The last campaign setting to look at is ad scheduling. This is where you can shut off ads during certain days of the week (like over the weekend), or limit ad impressions to certain hours of the day. This is really useful if you need to take orders or leads over the phone, and you only want to show your ads when someone is available to take the call. You can also use this to help stretch your budget, limiting ad coverage to your most profitable hours or days of the week. And, if you want to keep a leg up on the competition, you can even adjust your bids to be higher at certain times to ensure you remain in top positions.

That’s it for Yahoo. Join us next week when we continue our campaign setting optimization series. Next up: MSN AdCenter.