Reporting shouldn't start and end at an automated email, sent once a month. You need to build better reports, ones that illustrate how your PPC campaigns are performing but also where they're lagging, and what you can do to improve them. In PPC, every penny matters, so why not maximize that spend with reports that matter?
Presented at CTA Conference, June 27 2017.
Accompanying document can be found at kickpoint.ca/ctaconf2017
First off, grab this URL. I’m going to be covering a lot of tools and some custom reports. There is a document here with more details on everything I’m covering and links to more information.
Good reports keep the money coming in for your campaign. You never want to say “I think”, instead it’s “I know”.
Big hotel chain who wasn’t seeing results from their PPC campaigns and so shut it all off, and then saw a 20% decrease in bookings. That’s why you need better reports. That 20% was in there, but it was hidden.
The first report you can build is a chart that shows conversions by channel. Super easy!
This should be the very first thing you do because it’ll show you if paid is awesome or lagging behind. This is made in Google Data Studio, I’ll come back to that later. Also will come back to CQI!
Consider personas when you build these charts. Sometimes there are visitors to your site that are just going to happen but they’re not necessarily the ones you want to convert.
Here is the same data set from above but now only showing sessions from one of their key segments. See how organic goes from 0.94% to 5.47%, but paid actually stays pretty similar? That’s how you know you’re getting the right segment in via paid.
But this is a pretty basic report and this AdWords is doing well. Exporting a PDF of a Google Analytics report is the first way you can get your PPC budget cut.
I have an example of what happens if you’re not recording the right goals.
Check out fullstory.com. HotJar is similar. Everyone should watch these. Some companies watch them as an entire team – I love that idea!
This is only 10 seconds long, but I’ll tell you the rest of the story. Alexa here ended up filling out every single field in the form, and then just as her mouse was hovering over the submit button…
… she scrolled back up and downloaded the guide instead.
The great thing about FullStory is we can see how many form abandons we had with a gclid, or really, any source or UTM parameter – then you can silo out a specific source. For this client, we only have 13 sessions. Totally doable to watch these.
We got lucky with only 13 form abandons, but this is just one client. You can’t watch it all.
What else is possible?
I’m going to assume that most people here have a basic understanding of Google Tag Manager and what you can do with it. If you don’t, come find me later and I’ll happily talk your ear off about it. GTM is great.
Even if you’re not sure exactly what you’ll do with it, but you think it might be interesting to know.
Based on that FullStory session we watched, probably!
For this same client, we know this.
But now let’s segment that for just paid users.
A pretty good signal that the guides are interesting to this audience. Knowing this, you can then consider building retargeting audiences, or mentioning these guides in your ads to improve click through rates.
Dwell time itself isn’t enough to determine a goal. It has to be associated with an action.
It’s also a smart place to start when you’re considering if your landing pages are working. Use the gtm.timer built in. (If you’re doing well on time talk about making it smart.)
So if you’re using one of those snazzy single page landing pages with ajax form submission, you probably have no sense of how long people are actually on the page.
At this point, we often get asked why we just don’t use the time on site that is built into Google Analytics. Well, here’s the problem and it’s very specific to a paid audience. Generally, paid audiences visit less pages than other audiences. Particularly if you’re driving people to a specific landing page — that might be a single page, and even if they fill out the form, the visitor might not be directed anywhere else. And that means that the time on site won’t record their visit. Time on site is only recorded if someone visits another page. So if your pages per session are very low, it’s entirely likely that your time on site, or average session duration as they’re calling it now, is totally wrong.
Maybe you don’t have the budget for something like FullStory – you can get some useful data from GTM.
Get this. It’s great. Then you can see how far people are engaging with your landing page. There is a sample of how you can use this for reporting in the link I provided.
How can you see what leads are coming in – not just that a lead happened, but which one?
This feels like sorcery but it isn’t.
This is an important question you need to ask yourself when you set up your GTM.
Here’s a good rule of thumb — if the event you’re tracking is a thing that can be directly related to making your company money, it’s a goal. For example, a phone call, a form fill. If it isn’t, it’s an event. Time spent on site? NOT A GOAL. Make it an event. It really, really isn’t a goal.
Now that you have a sense of the kind of data you could be grabbing, now let’s get back to reporting. The standard reports in Google Analytics are kind of meh.
Keep an open mind about Data Studio. Now you can use a lot of different things as a data source, including Google Sheets and databases. Basically, anything that has an API or export function can go into a Data Studio report. And even if you’re not comfortable with using APIs, tools like Zapier can automate the process for you. For example, FullStory has an API and you could put that into a Google Sheet.
You can do some crazy math in GDS.
Analytics with math added on is enormously exciting! Yes, Google Analytics has calculated metrics but you can do some real next level stuff in GDS.
Let’s cover some fun ideas for metrics.
This assumes that you have one profit margin for everything. You probably don’t, and in that case you can use Google Sheets to make the math more precise.
Useful if you have a ton of repeat visitors in a short time window before the conversion.
Who didn’t bounce. Then you can combine this with other metrics to get data on say, events for unbounced vs bounced and see if there is an event tipping point or a particular event that means that the visitor is more likely to convert.
Thinking back to the earlier example – how much more is a form fill worth compared to a PDF download? Change the conversion rate to better reflect value.
Beyond the calculated ones that I’ve already gone through, here are some more ideas.
This is a pretty basic report to pull but how often do you look at top of funnel and bottom of funnel keywords? And consider the CQI, profit margins, intent, etc.
Do people download PDFs more on a specific day or time? Or book appointments? For example, bankruptcy lawyers get more calls on Mondays. Home builders have more leads towards the weekend. Do you need to use AdWords scripts to swap out landing pages based on day or time? Both?
As you improve page speed, check out how conversion rate changes. I guarantee it’ll go up. If it doesn’t, well, you may have other issues.
Don’t limit yourself to the standard percentages in the GTM code – add your own! Fire an event when a new section of your landing page is visited, for example.
You can use the regex functions in GDS to do this. You can also group keywords or pages by persona.
We haven’t actually done this but we think it’s a fun idea and we will test it out. Let me know if you run with this!
There’s a bunch of ways to do this – most social tools can give you a list or there are other tools.
The Links Shared API will get the links that they recently shared.
You can use the addPlacementToAdGroup function to add in all the placements you pull from step 2. Now you’re advertising on everything that people who are engaging with your competition are paying attention to.
Did it work?
PPC Reports That Matter #CTAConf 2017
#ctaconf @danaditomaso#ctaconf @danaditomaso
PPC REPORTS THAT MATTER