google ads

Google Ads Bidding Options: Overwhelmed With Choices?

Back when I first started learning pay per click (PPC), there was only one bidding option for what was then called Google Adwords, manual cost per click (CPC). That was it. It certainly made the decision on which method to use simple. While the decision was simple, the method wasn’t. Sure you could use a simple formula: CPC * (how many clicks it takes to get a conversion) = your Target CPA. I always tended to bid by feel and still do in manual CPC campaigns too this day. Mostly because I always want to pay the lowest cost per click possible.

Manual CPC

Back then int was much more simple. Bid one penny more than your competitor and show up above them. There were even tools you could get that figured out what your competitors were bidding. You could use an automated bid management tool that would use that data to get the position you wanted. Then came quality score changing the bidding landscape on Google Ads to this day. The new formula became (max CPC * Quality Score) = Ad rank. The ad rank among your competitors is what determines your position on the search results page.

According to Google:

Your max. CPC bid is the most you'll be charged for a click, but you'll often be charged less - sometimes much less. That final amount you're charged for a click is called the actual CPC.

Actual CPC is often less than max. CPC because with the Google Ads auction, the most you'll pay is what's minimally required to hold your ad position and any extensions shown with your ad, such as sitelinks. For example, if your Ad Rank places you immediately above search results (Ex: in the fourth position) with location extensions and sitelinks, you’ll pay the minimum amount necessary to keep that position -- and meet the relevant thresholds tied to that position -- with those extensions.

Very rarely do you actually pay your maximum cost per click, but you’re telling Google that’s the most you’re willing to pay. The difference between your average CPC and maximum CPC is called headroom. Here’s an insiders tip: it’s a good idea to have your headroom be as low as possible. This will typically reduce your average cost per click from what we’ve seen over the years.

Manual CPC bidding is a bit of a lost art in today’s PPC landscape. Personally, it’s been by far the hardest thing to train anyone in PPC on doing. Most PPC managers that excel at it have been doing PPC management before there were any automated solutions to bidding.

When Should You Use Manual CPC bidding?

Manual CPC bidding is available for search, shopping & display campaigns. If you ask Google, they’ll likely tell you NEVER. In fact, when you change your bidding settings in the interface it actually says this:

google ads manual CPC.png

Now don’t get me wrong. This is a 100% accurate statement, but so is Setting bids manually (when done properly) may result in higher performance. Trust me when I say, don’t believe everything Google has to say. They have a way of almost forcing you to do what they believe is the right way. Since their entire business model is based on making money, that alone should at least make you question the recommendation.

The truth is there’s no exact science to this. For any given advertiser, campaign, ad group & keyword the results may differ. Personally, I start nearly every new campaign on manual bidding to create a baseline. When I have that baseline I start testing other bidding options. There is most likely many that feel the complete opposite and there’s nothing wrong with that.

Enhanced CPC

Enhanced CPC is a box you can check along with manual CPC bidding. It’s not applicable to any other bidding type. For Google, this was the beginning of any automated bidding. Here’s what it looks like:

google ads enhanced cpc.png

Enhanced CPC uses manual bidding with a smart bidding layer on top of it. According to Google:

This strategy raises your manual bids in situations that seem more likely to lead to a sale or other conversion on your website, and lowers your bid for situations that seem less likely to lead to a conversion.

The major difference to keep in mind when using enhanced CPC is that Google can raise your bid above your maximum cost per click if they believe it to be more likely to convert. Again, you’ll see various answers on the results from this. From what I’ve seen most PPC managers have had good results with this and tend to use it. I know many that feel the opposite though. What’s the answer? Test.

Smart Bidding

There are a variety of bidding options across different campaign types that use smart bidding. While Google recommends Smart Bidding in virtually any case it’s not always the ideal solution. Even when you decide to implement or test smart bidding, there are a variety of choices and depending on the advertiser one might work better than another.

There are many reasons why Smart Bidding has an advantage over manual bidding, but that doesn’t always mean it performs better. If you’re using manual CPC bidding you have the ability to adjust bids based on the following: time of day, day of week, device, audience, location & demographic. In the Google Ads interface we have access to see the performance on each of those data points and choose to bid up, down or exclude.

However, Smart Bidding has those plus additional factors: location intent, ad characteristics, interface language, browser, operating system, actual search query, and depending on the type of campaign you’re running a few others. Even if we had access to every one of those options with manual bidding, it can get far too complex to manage. This is where artificial intelligence comes in. When a search is completed on Google, artificial intelligence will look at all of those factors and decide what your optimal bid should be. This is done every single time a search is made.

While that all seems amazing, one of the challenges artificial intelligence is up against is it relies on the data coming into the algorithm. A good rule of thumb is the more data you’re feeding it, the better it can determine your bid. Now the opposite can also be true. The less data your feeding it, the harder time it has determining your optimal bid. Even Google recommends at least 30 conversions over a month or longer to let the algorithm learn. What happens if you don’t do that much volume in a month? The algorithm still tries to determine your bid, but the likelihood of accuracy can go down.

Does this mean that if you don’t get 30 conversions within a month you should always use manual bidding? Although a good rule of thumb, not exactly. In fact we have two cases just within the last 30 days where we tested Smart Bidding on low volume campaigns and in both cases the results beat our manual bidding. In both cases we went out on a limb with this test.

Here’s one example:

smart bidding test.png

This one has an almost identical cost per conversion and we’ll continue to run this experiment until we have statistically significant results. What’s most interesting here is when you look at the cost per click and conversion rate. Both are nearly double. When they algorithm used all the factors is had for certain keyword searches, it determined a bid significantly higher than our bid would have a high conversion rate. They were right.

Here’s another:

I have plenty of examples of the opposite on either low or high volume campaigns, but what impresses me the most on these is how good the smart bidding is doing on such low volume. Does this mean we switched all our low volume campaigns to smart bidding? Nope. However, we are now going to run more similar tests for lower budget clients.

Now we’ll get into the different types of smart bidding.

Target CPA

This is likely your most common. With Target CPA you (the advertiser) determines the most you would like to pay for a conversion and the algorithm will try to match or beat your target CPA. Keep in mind this does not mean it will. It only means it will try.

Target CPA is a good starting point for many advertisers to test assuming your tracking conversions. It can work well for both lead generation and eCommerce. The best way to start testing this is a campaign that has historical performance data that will give you a baseline target CPA to start with. Typically you’ll want to start with the recommended Target CPA and let the algorithm learn. From there you can start to bring it down. However, Google does claim you can start a campaign from scratch with this bidding type and it will learn on the go.

Maximize Conversions

Maximize conversions uses your budget to try and get you the most conversions while spending your full daily budget. This can be great for campaigns that are either hitting their daily budget or campaigns that aren’t hitting their daily budget, but you would like them to even if it means a higher CPA.

While target CPA focuses mostly on hitting your target CPA, maximize conversions focuses more on hitting your daily budget while getting you the most conversions.

We don’t use this one as much, but we do use it for most clients if the campaign consistently hits our daily budgets.

Target ROAS (Return on ad spend)

Target ROAS can be a great option for eCommerce clients. Often times with eCommerce, ROAS is far more important the your cost per purchase. This smart bidding type works virtually the same as Target CPA, but instead of focusing on your cost per conversion column, it focuses on your conversion value divided by cost (ROAS).

Maximize Clicks

Maximize clicks just like it says, focuses on clicks. Mostly intended for branding advertisers who are looking for click volume ahead of any other metric, there are other use cases as well. Like maximize conversions, the algorithm will focus on spending your budget in full while getting you the most clicks.

A great case for using this bid type is if you aren’t tracking any conversions. Another way that we actually use quite often is when starting out a new campaign. Sometimes we’ll start out with maximize clicks because it may help speed up the process of getting clicks and spending our budget quickly. When we start out with manual CPC we are typically starting as low as possible and increasing bids until we get enough volume. Maximize clicks can help do that for us.

We highly recommend setting a manual bid cap when using this because it can increase your cost per click more than you’d like. While you should always keep a close eye when testing a new bid strategy, keep an extra close eye when testing maximize clicks.

Target Impression Share

Last, but not least target impression share. This is replacing target search page location as Google Ads is getting rid of average position anyway. The algorithm will set your bids with the goal of showing your ad on the absolute top of the page, on the top of the page, or anywhere on the page depending which option you choose.

Here’s what it looks like:

target impression share.png

First you would choose where on the page you want to show and then you choose the percent impression share you would like in that location. You also can set a CPC bid cap.

This can be a great bidding type for any client focused on branding and also can be good to test for brand campaigns with sales or lead goals that want to own the search results for their brand.

Conclusion

Overwhelming isn’t it? That’s what we’re here for at RelayPM. We highly recommend using an agency to manage your Google Ads account for this very reason. We’ve run hundreds of campaigns using every single bidding option available. This allows us to know which is best for each client to start with. We always test other bidding types though, whether it’s manual CPC or any type of smart bidding.

So which bidding type is right for you? The answer is all of them or just one of them. The only way to figure that out is to test and then keep testing. We love that Smart Bidding allows us to take one thing off of our plate so we can focus on another piece of the pay per click puzzle, but it’s not always the most efficient type.

Google Ads Auto-Tagging: What Is It & Who Should Use It?

Google Ads Auto-tagging has been around for as long as I can remember (and I’ve been doing this since 2006). While it’s not new, many companies using Google Ads and even some PPC managers either have never heard of it or aren’t quite sure exactly what it is.

Google Ads Account Settings

The default setting in Google Ads is auto-tagging turned on. The way to find this has changed a few times. Currently, On the left side you’ll see a list that looks like this image below. It doesn’t matter where you are in the interface you can click there. When you get here you’ll want to make sure and click Account Settings on the top.

google ads account settings.png

Once you are here you’ll see a full list of account level settings. One of them you’ll see is Auto-tagging and if you click in you’ll likely see the box checked off.

google ads account settings list.png

What Is Auto-Tagging?

Auto-tagging is designed to be used with Google Analytics. What it does is automatically append special tagging to all of your ads so that they data can be parsed out inside of Google Analytics. If you’re familiar at all with Google Analytics tagging you might be thinking about UTM tags. If you didn’t know any better you might think that if you use auto-tagging, all Google does is automatically create UTM tags for your ads. That would make sense wouldn’t it?

Look what happens when I click an ad on Google for an account that has auto-tagging enabled (we know this because it’s one of our clients)

https://www.troyleedesigns.com/?gclid=CjwKCAjwr8zoBRA0EiwANmvpYNCXRntcV592cPlZjm0ytjZfA_IW-MLfSFDkkpae7-Okj-v087UwBRoCJ6IQAvD_BwE

No UTM tags are to be found in this URL. What you see instead is a gclid tag followed by a bunch of letters, numbers and symbols. Why do they use a gclid and not UTM tags? After all, UTM tags were created for Google Analytics originally.

Why Does Google Use The GCLID?

A gclid is simply a Google click ID. There are a couple reasons Google uses a special click ID (which only Google Analytics can read) and not UTM tags. For starters, they want you to use Google Analytics for all your website tracking needs. Another reason is that UTM tagging only has 5 parameters that it can pass: source, medium, campaign name, campaign term, content. Google passes dozens of other data points from their click id into Google analytics that can only be passed this very way.

If you login to your Google Analytics account, under Acquisition you’ll see an entire Google Ads section:

google analytics google ads.png

You’ll see shopping data, hour of day, ad group and more. None of that would be possible without the gclid. If you’re Google analytics and Google Ads accounts are linked together you’ll also see cost data. If you have auto-tagging turned off and are manually UTM tagging you’ll still see data in here, but only for what you can pass which is limited to source, medium, campaign name, keyword and content.

Using auto-tagging and linking up your Google Ads and Google analytics accounts will provide you with all sorts of new data points that you can slice and dice up to get a deeper view of your advertising and how it integrates with your other website traffic.

Why Not Auto-tag?

The limitation of auto-tagging is when you are passing data into another place outside of Google analytics. This could be Omniture, Salesforce or any other type of CRM. Only Google analytics can parse the data in a gclid so you won’t see any of this passed to other tracking platforms.

If you are doing this already, make sure and double check your data. We’ve seen all sorts of funky things happen over the years when doing this.

Can I auto-tag and manually append UTM tags?

Technically it is possible, but I would advise against this. Even if you match up all the naming perfectly, we still will typically see data issues in Google analytics. To be safe and keep your data clean we would advise against this option. The better solution would be to create something outside of UTM tags that Google analytics wouldn’t read for your other tracking platform. In some cases this is farily simple, like with Salesforce and Omniture, but other cases it isn’t like if you’re using Unbounce and want the lead data passed.

If you must use UTM tags for your other tracking sources, you are better off missing out on the extra data passed from Google Ads to Google analytics, turn off auto-tagging and use only manually created UTM tagging.

If you have any other further questions on tracking your Google ads, please message, text or call and we’ll help you out.

Google Ads Smart Shopping: How Smart Are They?

What are smart shopping campaigns?

how smart is smart shopping.png

On May 3, 2018 Google Ads announced a new type of shopping campaign, called Smart Shopping. Continuing on their path of releasing new products driven by machine learning, Smart Shopping campaigns allow you to hand over your shopping ad management to the machines. At the root of the product, what separates them the most from traditional shopping campaigns is that the ads don’t just run on Google search, they also run on Youtube, Google display network & gmail. As we continually see in eCommerce product touch points are increasing and this product takes advantage of that.

The second feature that separates Smart Shopping from traditional shopping campaigns is that they are fully automated. While there could have been an option for Google to expand shopping campaigns to enable the advertiser to choose to expand the networks the ads run on, they chose not to. We would love to see this feature added in the future.

The third feature is that because they also run on the display network, Google chose to include remarketing (with no opt out). The intention is that you no longer need to run Shopping and Remarketing campaigns, only Smart Shopping.

That all sounds pretty good doesn’t it?

What are the downsides to smart shopping campaigns?

Smart shopping campaigns are a black box. The advertiser gets no insight into whether the ads ran on search, display network or gmail. All you see is "cross-network” for placement data. There is also no audience data shared. How much of your budget is being spent on remarketing? No idea and no way to find out.

You cannot do any of the following with Smart Shopping campaigns:

  1. Add negative keywords

  2. Change location targeting (feed runs in country targeted in the merchant center)

  3. Adjust bids by device

  4. Change ad scheduling

  5. Do anything with audiences

What can the advertiser control?

The advertiser can still see product level performance data which is great, but there’s nothing you can do with the data. When you setup the campaign you get to choose your shopping feed, a budget and ad text. The default bidding type is maximize conversion value, but you can choose a target ROAS if you please.

Here is what the ad setup looks like:

diamond supply smart shopping.png

Can any advertiser run smart shopping campaigns?

The requirements to run Smart Shopping campaigns are:

  1. You must have conversion tracking installed and passing revenue data

  2. You must have the remarketing global site tag installed

  3. You must have at least 20 conversions over the last 45 days across existing shopping campaigns

  4. You need an audience lists of more than 100 users

How do smart shopping campaigns perform?

This is what we’ve all been waiting for isn’t it?

There are certain clients we have tested this on in the past 6 months. You have to be careful because smart shopping campaigns are shown before traditional shopping campaigns. Even though you technically can have both on at the same time, your traditional shopping campaigns will receive almost no impressions.

What we have seen is that half of the clients we’ve tested smart shopping with, we have continued to run long term. Here is one example:

smart shopping results.png

The results above are over a 30 day period. Traditional shopping campaigns ran for the first 15 and smart shopping for the ladder 15 days. As you can see the cost per conversion dropped from $47.60 to $19.06. What’s interesting is you see similar conversion rates, but the cost per click is dramatically lower on smart shopping. That’s because the Google display network is less expensive to run ads on.

We also have cases where the cost per conversion is almost the same, we have cases where the cost per conversion dropped the same amount, but we lost too much volume and we have cases where the cost per conversion increased. Just like any other advertising products in Google ads: test, test, test.

What else should you know?

The most difficult piece in a real controlled test between traditional shopping and smart shopping is the retargeting portion. At Relay PM, we all but phased out running display retargeting on Google for clients because we’ve seen much better performance with Facbeook ads. It’s difficult to factor that into the campaign performance. The analysis can get complex.

Conclusion

For many advertisers we highly recommend testing this feature. Be aware of all the pros and cons. Be aware it includes retargeting. Understand in most cases you’ll never be able to have a real controlled test. There are too many pieces of data Google doesn’t share.

As we’re able to bring in more data we’ll continue to update this article. Please feel free to message us with any questions.

If you’re looking to test shopping campaigns, let us know.