Tracking conversions is an essential part of PPC management so that you make data-driven decisions about your account. The deeper you dig into the conversion tracking setup, the more common it is to see mistakes. Many companies regularly audit their search terms, keywords, and ad tests, however, they don’t always audit their conversion settings.
When your conversion tracking settings have issues, or your conversion action sets aren’t consistent across campaigns, then your conversion data is incorrect. As automated bidding, and many of your optimization efforts stem from accurate conversion data, having incorrect conversion data creates problems. There are companies who are managing their accounts ‘correctly’, however, their decisions aren’t correct due to their conversion data being inaccurate.
The worst mistake is around the consistency of decisions. One type of conversion is using one attribution model ad a different conversion is using a different model. These types of consistency issues can have dramatic effects on your account’s data and efficiency.
To make it easy to see the top mistakes and know exactly what to look for in your account, we’ve put together a video that will walk you through conversion settings and the top mistakes you should look for in your Google Ads conversion tracking settings.
Here is the full transcript of the video:
0:00:00.2: Hello, this is Brad with Adalysis. In this video, we’re going to look at the top conversion tracking mistakes and how to fix them. When thinking of conversions, this data, it’s easy to set up, and it informs Google ads how to do bidding. It tells you what search terms will be negative keywords. It informs your ad testing data and so forth. A couple of mistakes in conversion tracking can have massive implications for your bids being wrong, to other poor decisions being made. When we go to the conversion action section of our account, we can see all the conversions, our source category, attribution, etcetera, it’s a nice overview. When we click into an individual action, we have different options to look at. Our number one mistake is not using values for conversions. If you’re e-commerce, then of course you’d want to use the actual checkout value. But if you’re doing lead gen and you have some conversions that have low value and others with high value, you can just use a point value system.
0:01:04.4: Our next biggest mistake is the count. If you’re a lead generation and you’re tracking form fills, phone calls, whitepaper downloads, and you have every turned on, that means if one person fills out your form 10 times, the count is 10 conversions. If you only want uniques, your count should be one. If you’re e-commerce then you may use every because of course, every time someone checks out you make more money. Next, we have included it in the conversion column. This is a very important one. If you include the conversion in the conversions column, it will be used for automated bidding. For most third-party systems such as Adalysis, we only pull in what’s pulled in your conversion column. You may have other things being tracked such as page views, time on site, etcetera, and those actions will show up in the All conversions. The conversion column is the most important one because this is where all the big decisions, Ad testing decisions etcetera, come from.
0:02:09.4: Next, we have not chosen an attribution model or being inconsistent in attribution models. By default, we have last click. If you have a lot of conversions, you could use data-driven or you can use time decay, position-base, etcetera. Now, where this really comes into play, thinking about your attribution models, is when we may go back to all of our conversions, and we say, “Who’s being included in conversions?” And we have one, that’s time decay. Another last click, another position-based. This is going to cause a problem because if you really want everything time decay, but there’s a last click in here, or you want everything position-based but there’s a time decay in here, the data gets really messy, and bad bids are made for your account. Your conversion data is not put correctly across all your data. So often when we look at our overview, we’re saying, “Are we consistent on our attribution? Are we consistent on our counts?” Now, you may have some that are different. You might have, this one is a count of every because we’re e-commerce, but we want to count catalog leads, so therefore we’re going to count those as one but e-commerce as every. There are exceptions. And then, do we have some that had different conversion values? It’s important to be consistent.
0:03:29.9: Our next biggest thing to look at with mistakes in conversion tracking is how you choose to handle local actions. Is a website visit a conversion that you want Google to optimize for? What about driving directions? Store visits are often one that is included. When there are other conversions that are being added to your account, have you thought about how you want to have them included? Many people don’t want driving directions included they’d rather just use store visits. Our click-through conversion window is how long a user has to convert after clicking an Ad. If these are inconsistent that can also lead to inconsistent decisions by a bid system Ad testing, etcetera. Consistency. If you’re using the call extension, those are also automatically included with the count of every which is often not what you want. Now, in your settings, you can choose the call length for what you want to be counted as a conversion. So, you could say, “Well, only if a call is at least 30 seconds or 60 seconds, do I want to count it as a conversion,” so leaving it as any call is usually a mistake. And thinking about when someone calls you, what’s a successful call look like and how long does that take? That’s an important one and that can be your seconds.
0:04:52.5: Another mistake is tracking conversions on the website that never make it to Google Ads. The most common issues with this are phone call tracking, where someone tracks phone calls and they put it into Google Analytics, or they put it into Adobe. And in those cases, if they’re not being put into Google Ads, you can’t make decisions on them. You can either do an import from Google which is easy to do right within the system or for third-party systems you can do an offline import. These are the most common problems with the conversion actions themselves. We have a bonus one: Conversion action sets. Conversion action sets let you bundle conversion types together. We can go in, and we can say, “Count these different conversions as a set.” Now, what’s nice about this is when we can go into a campaign, and then we can say, “Let’s use these sets, these track conversions for this campaign. And let’s use a different set for a different campaign.” This can be useful when you’ve got some sets that are awareness, top of funnel, maybe do some discovery Ads, and you’re really focused on time-aside metrics.
0:06:04.9: But then you get your search ads and your remarketing ads, and you expect them to convert. In those cases, you want to focus purely on who converted the lead information and not just a good visit. So, when we look at the conversion action sets, we are often looking to see how many campaigns are they used in? And do they have different conversions being used in different campaigns? Because if you have four or five conversion action sets, and one is being used in one set of search campaigns and a different one is being used in a different set of search campaigns, that may lead to an inconsistency issue and you need to examine that very, very closely. What are conversion actions? Our number one thing we’re looking at is consistency. Right? Consistency in the attribution. Consistency in the counts. There could be some exceptions here. Consistency in the click through window. And then, ensuring only conversions that should be used for bidding, Ad testing, search query data, etcetera, are in the conversion column, because it’s important. In systems like Adalysis, we use counted conversions for automated Ad testing, automated analysis, etcetera. We’re going to look at what’s in the yes column, how that data has been divided up, and then give automated recommendations based upon that information.
0:07:28.5: So, the takeaway from this video is, number one, make sure you understand what these conversions mean. Number two, be consistent. Number three, double check any conversion action sets, and other values or conversions not being imported. And if you can get through those three things, then you should have great conversion data, making all your optimization efforts much more efficient.