Dimensional Reporting in Google AdWords
Exploring Dimensional Reporting
Hello fellow advertisers! Welcome to another episode of my video blog series. Today I will be reviewing some different types of dimensional reporting you can review within Google AdWords and/or Analytics. I’ll discuss how/where to find these reports, and also what kinds of things you should be looking for when analyzing them. You’ll be able to implement the findings from these reports to help you reallocate spend from your underperforming campaigns so that you can optimize for a better ROI within your accounts. Enjoy!
Hello everybody, and welcome to another episode of my video blog series. I am your host, Nick Rennard, and today we will be talking about dimensional reporting within Google AdWords. Let’s go ahead and get started.
All right. There are 4 reports I’m going to be going over with you today. The first one is called the day of the week report, which is going to analyze the performance of your campaign based on the day of the week. The second one is going to be called the hour of day report, which is similar in that it just analyzes the performance of your campaigns based on the hour of the day. The third one is called a mobile report, which is going to compare the desktop traffic that you receive compared to things like mobile or tablet or anything else, so it compares it based on the device that people are coming from. The last one is a geographic report, which is going to look at the performance based on the, most of the clients that we have are within the United States, so it’s going to look at it based on the country or state or city that you’re advertising in, and yeah.
There’s a final note at the bottom here. All these types of reports are viewable in both AdWords and Analytics, so if you prefer one platform or the other, that’s totally fine. I usually prefer to use Analytics, just because the data is more accurate within Analytics, and it’s more customizable. If you feel more comfortable with AdWords, then that’s fine, too.
All right, we’ll start with the hour of the day and the day of the week reports. I’m going to combine these, because you can actually find these within the same tab within AdWords. If you start off here, I have all the steps that you need to do laid out here. Number 1, always check your date range. Make sure you’re looking at the appropriate portion of data. With all of these reports, you generally don’t want to review these reports over a short time frame. By short, I mean even 30 days is going to be relatively short. It’s going to be dependent on how much you spend, but I generally look at these reports every 2 months for my clients.
If you’re running a campaign or you’re running campaigns for other people, 2 months is generally a good rough estimate. However, if you spend a lot more, you could review them monthly, or if you’re not spending that much at all, maybe you’re only spending 1,000 a month or something like that, you may want to review this every 4 months or so. The reason we do that is because the data isn’t going to be statistically significant if we aren’t looking at a large enough quantity of data. It takes a while to get this kind of data. If you’re looking at a 7 day period, then the patterns that you are seeing may not translate to the future months. Yeah, you get what I’m saying.
All right, second step here. At the top here, you have your set of tabs, starting with Campaigns. If you go all the way over to the Dimensions tab and click that, there is a drop down box that says View. You’re going to want to click that and go down to Time, and the 2 reports are going to be under this Time drop down box. You can see Day of the Week, Hour of Day.
Click either one of those, and we’re going to go ahead and start with the day of the week report, which looks exactly like this. You can see here, it says, “Make sure you have statistically significant data.” I’m pointing to the number of clicks here, because that’s going to tell you whether or not, it’s going to give you a good idea whether or not you have statistically significant data. What you’re looking at here, seeing this many thousands of clicks, this would definitely be statistically significant. If you’re wondering where that kind of break point is, I would want to see at least 100 clicks or so on each one of these in order to be able to take these reports seriously.
If you look at this report over the last 2 months and you only have maybe 70 clicks, or maybe even some of the days have 100 clicks and others have less than 100, it’s one of those things where more is better. If you have 200 clicks, that’s going to be more statistically significant than 100 clicks, and so on and so forth. This report is definitely statistically significant, and what we’re looking for within this report is the conversion data. We’re going to be very conversion focused, because that is what determines the success of your campaign. That’s what we’re trying to do, is generate leads from these campaigns, for whatever company is running the AdWords campaign.
We look at the cost for conversion, and we also look at conversion rates. Also, a side note here, if you’ve added conversion values to your campaigns, so let’s say you have multiple conversion types, like you have phone calls and form submissions and whatever else, you can assign values to those leads. If you do have conversion values set up, then I highly recommend you also add the columns that revolve around conversion values, conversion value divided by cost, your total conversion value, you can look at those as well, but cost for conversion and conversion rate are going to be pretty good.
You can see, in this example here, that weekdays actually perform a lot better than the weekends. You can see that they have a lower cost for conversion, lower is better, and the conversion rate is also almost double that of the weekend, sometimes more than double. In this situation, if we were to look at the past 3 months’ worth of data and see a pattern like this, this would tell us that we want to be decreasing bids on the weekends and increasing bids on the weekdays. That way, we can reallocate some of those funds from the weekends where we’re not getting as many conversions to the days that are performing better.
All right. Next report here, the hour of the day report. This one’s going to be much bigger, and since this data is going to be spread over 24 hours throughout the day, you can see here hours 1 through 24, and with a day of the week report, you have it spread across 7 days, so it’s going to take more data for you to have a statistically significant hour of the day report, since it’s more spread out. Just keep that in mind. Again, make sure you have statistically significant data by looking at your number of clicks. Make sure you have 1 or 200 in there on each one of these. Then again, look for patterns in performance.
One thing to keep in mind is when you look up these reports, you want to be careful not to over analyze them too much. In order for me to actually make changes in an account, I want there to be a pretty glaring, obvious pattern throughout the day. In the last one, in the one before this, I’ll go back to it, we see that the weekends are just obviously worse than the weekdays. I would definitely make a change based off of this.
You will see reports where you’ll see one hour has a great conversion rate, maybe 4 a.m. is great, and then 5 a.m. is bad, and then 6 a.m. is kind of medium, and then 7 a.m. is great, and it just flip flops over and over. If you’re not seeing any consistent pattern, like a chunk of time where it’s like, “Okay, 11 a.m. to 3 p.m. is just fantastic.” If you see a block like that, then go ahead and add bid modifiers during those hours in your ad scheduling, but if it’s too spread out and too sporadic, then there’s a couple things; 1, maybe there isn’t a pattern, you don’t need to make a change; or 2, maybe you need to wait and look at a longer time period, because if those clicks that you’re looking at are too spread out, then again, it goes back to that whole statistical significance thing. It’s going to be the buzz word of the day for this video here, is statistical significance.
You can see in this one here that as it gets later in the night, so more around 12 a.m. or 1 a.m., or even late like 11 p.m., the conversion rates are much lower. As we get to the core of the day, between probably around 9 to 5 or so, you can see the conversion rates shoot up pretty drastically, so this is a situation where I would definitely create some ad scheduling to emphasize the bids during these peak hours and try to de-emphasize some of the spend that we have during the off hours, because you can see here that really early in the morning, they’re spending thousands of dollars for conversion rates that aren’t as good as the, during the day, sorry. I’m fumbling on my words there. Yeah, you get that pattern, we’ll go ahead and move on to the next report here.
All right, next here is going to be the geographic report. Now we’re going to move over from AdWords to Analytics. In Analytics, it’s going to be almost the exact same process, it’s just the buttons are going to be a little bit different. In Analytics, again, start by setting your appropriate date range at the top. You can see here, looking at 3 months’ worth of data, which is a good chunk of time to look at. One thing you’ll want to make sure in Analytics is, Analytics tracks the traffic from all sources, so that’s organic, that’s direct, that’s paid, that’s referral, that’s everything. If we’re only interested in the data from paid traffic, then make sure you set up a segment to only be looking at the paid traffic. You can see we have a segment here for AdWords paid traffic, and then the last step here is going to be click on the Audience section over here on the left, and then click Geo, and then click Location. It’s going to bring you to this next screen here, once you click those 3.
It’ll actually start by, this is actually 3 different screen shots merged into 1. This top part here that says Country up in the top left is what it will start with. It’ll start at the most zoomed out version, so it’ll start, in this case, we’re looking at by country. This client only advertises in the United States and Canada, so that’s only where they’re going to see traffic. The cool thing about Analytics is we can actually drill down on those countries and look at the particular states, and then we can drill down on those states and look at cities, so you may be thinking, or maybe you were thinking, that, let’s say you only advertise in southern California or something like that. This report is still relevant to you because you can drill down to see if Los Angeles or San Francisco or San Diego, you can see how each one of those cities is performing and adjust bids based off of that data.
Yeah. Another note here, kind of like setting your date range, there’s going to be a little Conversions tab right here. Make sure that you’re looking at the right conversion. A lot of times people will set up goals within Analytics for things like remarketing lists and whatnot, so make sure you’re looking at the conversions that are actually relevant to you and so that way you’re not reviewing the wrong data. Again, it’s going to be the same process. We’re going to be looking for patterns in performance, so if we drill down on the United States here and we see that the conversion rate in California is drastically better or worse than Georgia or whatever …
Actually what I usually do is if we’re going to organize it by state, I’ll download this report into an Excel file. Then what I’ll do is I’ll organize the data based on highest to lowest conversion rates. I’ll take the, maybe the top 20 or 30% conversion rates, the best conversion rates and highest conversion rates, and I’ll add a positive bid modifier to those. Then I’ll take the bottom 20 to 30% and I’ll add a negative modifier for those, so that way you’re reallocating the spend from the under performing states to these states that are performing better for you. That’s kind of a way to do it.
You can also organize, when you download these reports into Excel, I also like it because you can organize it by these other columns here. You can, user acquisition, user engagement, bounce rates, pages per session, stuff like that, so you can really look down at a very, very granular level as to how interested these users are. If we see something like, there’s a handful, let’s say we organize it by bounce rate, and we see that there’s a set of states where the bounce rate is really, really high, 90% or something like that, regardless of the conversion data we can actually add bid modifiers to the account simply because 90% of the traffic is bouncing. That’s not so great. Yeah, there’s more that we can look at than just conversion data in these reports, but again, conversion data is going to be the most important one, because that’s the one that determines performance for us.
All right, and then the last one that we have here is the mobile report. Again, same process. Set your date range in the top right. Make sure you’re looking at the proper segment. You’ll want to click Audience, just like you did with the geographic report, but instead of going to Geo, you’re going to go to Mobile, and then go to Overview. This report is actually a lot simpler, because it just breaks it down into 3 sections for you: desktop, mobile, and tablet. You actually can’t, well, I don’t know if you can yet. Last time I checked, you weren’t able to adjust the bids for tablets. The only thing that you can actually adjust bids for within AdWords is either mobile or desktop, so those are the 2 we’re going to be interested in.
You can see here, again, set your conversion, and you can see here that the conversion rate on desktop within this particular campaign is 2.75% where the conversion rate on mobile is 3.36%. Again, if you see a pretty big gap like that, between conversion rates, and that’s actually pretty big. Going from 2.75 to 3.36 is pretty big. It’s less than 1%, so looking at it that way sounds little, but if you think about it, how much per percentage increase that is, yeah, I mean that can make a pretty big difference. If I were to see data like this, I would definitely increase the bids on my mobile and decrease the bids on my desktop. Again, reallocating spend from the under performers to the campaigns that are performing well. Well, in this case, it would be based on the device, but yeah. You get the idea.
All right, so for all of these reports, the process is going to be pretty much the exact same. You just need to know where to click within these platforms, whether it’s AdWords or Analytics or if you’re using, I’m sure they have them on other platforms like Amazon as well. Again, make sure you’re looking at the proper date range. Make sure you’re looking at a relatively large date range, 2 months or more is the rule of thumb. Check to make sure that you’re looking at the proper goals, and conversions and goal completions. You want to make sure that if you set up a goal for people who didn’t bounce off the website, that you’re not merging that data with your phone calls, because it’s going to make it look way inflated or deflated.
If you’re using Analytics, make sure you’re looking at the correct segment. A lot of people, this is a rookie mistake, and I mean, honestly, I even make this mistake from time to time where I start doing a report and then realize that I’m not using the right segment and have to go back and redo everything. Make sure you just do that in the beginning, so that you’re filtering the traffic that’s relevant to what you want to look at, probably paid. Check the statistical significance. You want to be looking at clicks, or if you’re in Analytics, sessions, to make sure that you have enough data. That way you’re not looking at a bunch of percentages and it’s only based off of 14 people who visited the site, because that’s not going to be enough people to determine whether or not that pattern is going to continue over a long period of time.
Lastly, look for patterns in performance and adjust with bid modifiers based on that performance. If there’s good performance, you want to increase your bids; and if there’s bad performance, you want do decrease your bids. It seems obvious, but a lot of people would just do, I could see a lot of people doing just one or the other. You want to make sure you do both, because if you only increase your bids, it’s going to jack up your impression share and your cost per click, so you want to make sure you do a little bit of both so that way the spend balances from the under performers to the good performers.
Anyways, that’s all I have for you today. Thanks for watching, and happy advertising. I will see you guys in my next blog.