Behind the Scenes: How this team Growth Hacked Facebook Ads to the Perfect Audiences

Want to achieve high conversion Retargeting Audiences and scalable Lookalike Audiences for growth? Read how this team unlocked their perfect audience targeting for growth...

Today we get behind the scenes to chat with Cynthia Pucheanu (Demand Generation Specialist) and Stefan van Eerde (Demand Generation Manager) at Trimble.com.

Cynthia and Stefan’s growth marketing team at Trimble have been using Popcorn Metrics to track user events into Facebook Ads (via FB Pixel) and Google Ads (via Google Analytics).

By a series of Growth Hacking experiments they unlocked their perfect audiences for growth.

First, by tracking behaviours like scrolling and visit seconds on each page.

Then, experimenting to discover the perfect formula for “behaviour based” audience targeting.

They were then able to scale up using “behaviour based” Lookalike Audiences, and ultimately to Conversion Based Lookalike Audiences using visitor “in page” click behaviours.

Let's join the chat
with Paul, Cynthia and Stefan...

Interview with Cynthia Pucheanu (Demand Generation Specialist) and Stefan van Eerde (Demand Generation Manager) at Trimble.com. Cynthia and Stefan’s marketing team at Trimble have been using Popcorn Metrics to track user events into Facebook Pixel and Google Ads (via Analytics).

PAUL: Hi Cynthia, just to kick off, can you tell us a bit about you and what you’re doing in Trimble and Stabiplan?

CYNTHIA: I’m part of the marketing team. We are in charge of generating leads and following them up further through larger campaigns. We are involved in all marketing processes throughout the organisation.

CYNTHIA: I’ve been with Trimble for about a year now, since Trimble acquired Stabiplan (where I was working previously). MEPcontent is a brand of Stabiplan.

PAUL: What do you do specifically within the team? What’s your role?

CYNTHIA: I’m mostly in charge of the paid campaigns on different platforms. Also, coming up with different experiments and growth-hacking the ideas and techniques to generate more leads, to increase our visibility and to achieve our goals.

PAUL: How much of your work would you say is just marketing/advertising versus growth-hacking?

CYNTHIA: I don’t see them as different, I see them as intertwined. Marketing includes the traditional branding of a product or of a company, whilst growth-hacking is just a fast way to get more growth. I can’t really say that 20% of my time is marketing, because marketing for me is growth-hacking. They’re not interchangeable. I like to have a growth-hacking mindset to whatever marketing activities I’m doing.

PAUL: You’ve been using Popcorn Metrics for a year, maybe more. Why did you look for a solution like Popcorn Metrics?

CYNTHIA: It was actually a colleague in our team. He did a Growth Hacking course in Amsterdam – Growth Tribe. He said “guys, I found this really cool tool and I think it will save us a lot of time” and we said “okay, let’s try it”.

He said “guys, I found this really cool tool and I think it will save us a lot of time” and we said “okay, let’s try it”.

It worked well and it did the job that we were looking for so we went with it.

PAUL: How were you solving your problems before you used Popcorn Metrics?

CYNTHIA: We were tracking conversions based on data from different platforms, and we weren't sure of the accuracy of that and how that translated to real actions performed on our website by our users. Also, we wanted to get really narrow, retargeting the audiences.

"we wanted to get really narrow, retargeting the audiences"

We started with people who scrolled 50% of the page and spent 50 seconds. But we discovered that it was too narrow. So we decreased it to 20 seconds, but then the scroll length was still too high. Then we ended up with 10 seconds. It was good because we were able to exclude the bounces.
PAUL: What difference did you see that making in terms of your retargeting campaigns?

CYNTHIA: We wanted really to experiment and to combine the audiences, so we created lookalikes from the audiences and then later we created lookalikes from the converters.

Recently we found out that the retargeting audiences of our product that we were using to retarget the same products to those people were being saturated. So those people already knew about us, and maybe they were just not interested anymore.

So what we did is we created ads for different products and advertised those to people who visit other products on our site. We did a sort of cross-selling technique, and that’s worked really well and it generated a lot of conversions for us.

PAUL: So you went down to the 10 seconds to narrow it down to remove the bounces, which I guess then left you with the audience that were more interested because they’d stayed on your page for that amount of time. That gave you a smaller audience. So were you saturating because you were using the same budget on a smaller audience or were you able to reduce your budget?

CYNTHIA: In the beginning it worked well and then we didn’t make any changes to the budget. But over time – the campaign ran for a couple of months – gradually we saw a decrease in conversions. So we just concluded that those people were just not interested, on that specific product. So that’s when we started cross-selling and its worked really well.

PAUL: You mentioned lookalikes. Were you building lookalikes from the 10 second retargeted audiences?

CYNTHIA: Yes. That, and per country. So, 1% lookalikes on Facebook per country, so then that worked really well. We made lookalike audiences from our 10 second audiences.

"We made lookalike audiences from our 10 second audiences."

And also, after we built up enough conversions on our lead generation ads on Facebook, we took that audience of converters and we created lookalikes in the same way as before from those.

"we took that audience of converters and we created lookalikes"

PAUL: Okay. So you started off with lookalikes of people who were spending time or scrolling to get a more targeted audience, and then once you were building up more conversions you were then able to switch across to then making lookalikes from the people who were actually converting as opposed to people who were showing that they were interested?

CYNTHIA: Yeah, and so on.

PAUL: That makes perfect sense. Did you try out lookalikes built from different groups (30 seconds, scrolled 50%, 20 seconds, 10 seconds etc) to work out which level worked best for lookalikes? Or did you start with a smaller audience but build a more accurate lookalike off that?

CYNTHIA: We actually first used the retargeting audiences themselves and then once we found the correct formula for that audience, then we went on to create lookalikes from that. Whenever we figured out that 10 seconds worked well for the retargeting, we then built the lookalike off the 10 seconds audience.

PAUL: I get it. Okay so, you experimented to find that 10 sec level. Then you built lookalikes off that 10 sec audience to reaching into a similar market until you’re getting enough conversions. Then you changed again by creating lookalikes based upon the conversion audience rather than the 10 seconds “I’m interested” audience.

CYNTHIA: Yeah.

PAUL: That’s a great strategy. I like what you were doing there.

STEFAN: For us it is important that we have clear attribution in place. That’s why Cynthia is looking after Popcorn Metrics (and the other tools we have running) to make sure we use all the tools in an optimal way.

PAUL: It’s always a balance using enough tools to make your life easier, but not so many tools that you’re standing on your own toes and tripping yourself up.

STEFAN: Yeah, absolutely.

PAUL: Stefan, what do you do in the team? You’re Cynthia’s boss right?

STEFAN: I guess so, yes. We have a team spread around the globe doing cool stuff together.

PAUL: Cool. So how did Cynthia convince you that you should use Popcorn Metrics?

STEFAN: Really by showing what it’s capable of. It was a good addition to the tools that we already have. And she’s just also really good at coming up with the numbers to see if it’s worth the effort. So that convinced us.

PAUL: Did it justify itself in that you were making better returns on your campaigns or did it justify it because you were moving faster as a team or because you didn’t need to hire a developer?

CYNTHIA: Yeah, all three I would say.

STEFAN: Yeah.

PAUL: Was any one more important than the other?

CYNTHIA: Yeah, I guess the fact that we didn’t need any developers, because it takes a long time until something gets implemented. We have an amazing development team, but they are so busy. So for the sake of speed and just trying it out, we wanted to do it this way and then we were free to do it ourselves.

PAUL: That makes perfect sense. So can you give me a quick overview or example of where you use Popcorn Metrics in your Facebook ads? Please tell us, what was the challenge, how did you go about it, and what were the results?

CYNTHIA: As I was saying earlier. We really wanted to get the correct audiences for our ads. We wanted to optimise our budget, so as not to just throw around money. We wanted to really be on point with our message to the people who had already shown interest. And also, we didn’t want to lose those people that showed that particular kind of interest by trying to bring them back. I think that was the challenge.

PAUL: And what would you say were the differences that made in your campaigns?

CYNTHIA: It allowed us to be faster in our experimenting, and so from those experiments we were able to find the correct formula to increase finally our conversion rate.

PAUL: And did your conversion rate change significantly once you’d done that work? Or did your revenues change significantly?

CYNTHIA: Well, we didn’t really monitor revenue. I don’t remember exactly what the conversion rate was before Popcorn Metrics.

STEFAN: It was hard to attribute it, so we had a lot of technicalities as to what source led to what kind of conversions because we had internal registration systems which was also received as being a referral channel.

PAUL: Of course, yeah.

STEFAN: That was always messing up our stats before we had Popcorn Metrics in place. Popcorn Metrics helped us to clarify and get more of an understanding than the data that we gathered before.

PAUL: Attribution can always be a challenge, especially with people coming in from internal references and from other parts within your ecosystem. How were you able to separate that out using Popcorn?

STEFAN: I think it was a dual approach. Popcorn Metrics was set up by our colleague, with help from Cynthia. At the same time, we convinced the team to make adjustments on the registration system. So it was sort of mutual cooperation. It was a joint effort that actually made sure that we had a clear understanding of where our traffic came from and how it all came together when it comes to conversions as well.

PAUL: And that’s hugely important, because you need to know what works in order to know where to reinvest and put your efforts in.

STEFAN: Yeah, absolutely.

PAUL: And with Google Ads…were you doing similar things or was it different? Could you maybe give just a quick example of Google Ads? Because I know that the infrastructure is a little bit different with them.

CYNTHIA: I would say that the challenge was the same.

What was really important for me was to see exactly what type of conversion we were getting, because we created separate events for the separate apps that we have. And then we were able to see maybe one app, one product, in one campaign.

We were expecting all of the conversions to be tied to that, but we were surprised to discover people were downloading other apps, because they’d just visited our page and were scrolling. So it was interesting to see what relations there were between the apps. That was also one of the ideas that brought us through to using a more general audience to cross-sell. We saw that people were interested in other things as well.

PAUL: I see. So basically you were running various different Google Ads, and expecting particular ads to be generating particular conversions, but you were found that people were buying other apps – i.e. different products that you were already offering – that you weren’t expecting from those particular campaigns. That gave you the idea to do these broader campaigns because it’s a lot of cross-sell. And you were able to see that because it was easier to set up conversions or faster to set up conversions? Were you putting conversions on a successive page or on a download button or?

CYNTHIA: On a download button, yes.

PAUL: Okay, so it’s in page on that button click.

CYNTHIA: All of the products are from different pages, but you have the homepage where all of them are listed. The navigation is pretty easy, so although they always arrive on the landing page of one product, then they can easily go to the homepage because you have the navigation by the top. It just says – ‘All Apps’ – you can go there and you can see all of them.
So it wasn’t like we were advertising for one product and then we were getting the conversions for a totally different kind. The biggest chunk was what we were trying to get, but what was interesting was we were also seeing these other conversions. That data gave us the idea for the cross selling. It’s been great.

PAUL: That makes really good sense.

PAUL: I’m curious… with AdWords, people tend to track conversions either: by sending an event into Google Analytics and then turning that into conversion goals and importing that into Google Ads. Or just creating a conversion straight inside AdWords on a particular page. Why did you choose Google Analytics with your Google Ads? What advantages did that give you?

CYNTHIA: We worked with goals set up from events sent by Popcorn Metrics.

PAUL: So, you had events coming off Google Analytics and you would then turn them into a goal for a conversion inside AdWords. Did you also use those events to create retargeting audiences like you were doing in Facebook?

CYNTHIA: Ah, yeah yeah yeah. Sure. The same logic was behind it, only that we didn’t use lookalikes in Google.

PAUL: Okay, but you could create retargeting audiences of people who were either showing interest?

CYNTHIA: Yes!

PAUL: Did you do it off the same 10 second audiences - did it work the same? Or was it a different number?

CYNTHIA: 10 seconds, yes, the same.

PAUL: And did you also do it off the conversion events as well?

CYNTHIA: Yeah, we also set up conversions.

PAUL: It’s so interesting talking to you Cynthia, the way that you’re approaching is really interesting and I really love the way you’ve used your strategy to narrow that down faster and move faster in your work. Narrow that down to find what was the right kind of interest level – that 10 seconds retargeting audience – and then be able to use that to, on Facebook, create your lookalikes and then once you’d gone further create your conversion based on those lookalikes. And a similar kind of thing in Google except for the the lookalike audiences – but you can retarget.

PAUL: Does that mean that you’ve created campaigns where you would bid higher for audiences that had already arrived and had shown interest?

CYNTHIA: That would’ve been an idea, to increase the bidding but we didn’t actually do that.

PAUL: It’s a nice strategy because, if you’ve got somebody who’s already visited you and you know that they’re going to be interested, then if they put in a similar search those can be audiences where it’s probably worth overbidding. Those are bids that you probably do want to win because they are still looking. Anyway, you can experiment with that in the future.

PAUL: Were you sending the same events then into Google and Facebook?

CYNTHIA: Yes.

PAUL: You mentioned before the link between Google Analytics data and Facebook data.

CYNTHIA: To be honest, for me Facebook Conversions were a bit confusing. I didn’t really trust it. I didn’t really know if it was accurate. So that’s why I really wanted to rely on analytics data.

PAUL: I understand. You wanted to see the same data in Facebook as you have in Google because you weren’t trusting Facebook data, or you weren’t trusting Google…?

CYNTHIA:We wanted to use the same source of data for everything, as much as possible, just to be sure that we had the same benchmark. For example, if we looked in Facebook the conversions they are skyrocketing, but then we looked into analytics and we saw that it was not the case. So that’s why we wanted to use one source of data.

PAUL: And what did you decide to be your “true” source of data?

CYNTHIA: Google Analytics.

PAUL: I think it’s a fair call. From a Facebook conversions perspective, you’re seeing which ads are converting – you see which ads work, which ads don’t work.

Even if the numbers aren’t exact between Google and Facebook – sometimes that doesn’t matter, because you can go “okay, let’s put more on these ads because they’re getting the best results”.

But, yeah, if you’ve got different sets of data like Facebook and Google, you just have to decide “who are we going to call the truth?”

PAUL: Plus, by sending the same events into Google Analytics, you can look at your Channels Report and you go “all these are coming from these channels” and you can get that cleanly.

CYNTHIA: Yeah.

PAUL: Do you also use UTM characters on your parameters on your Facebook ads?

CYNTHIA: Yeah.

PAUL: Cynthia, we’re about to wrap up. Is there anything you’d like to add today?

CYNTHIA: Yeah, you said before that you were surprised by how we approach things but how do you see usually marketing teams approaching things?

PAUL: Well, I think it’s the fact that you took a full growth-hack strategy. Usually when I speak to people it tends to be more “we’re working on this little particular conversion that we’re tracking” or “we’re going to create this lookalike audience”.

But what I found interesting is that you’ve taken a whole strategy approach that used Popcorn Metrics to get to that page level of interest based upon scrolling and time, and you’d done that until you’d figured out what was the right kind of level, which was the 10 seconds mark – and then turning that 10 seconds retargeting audience into a lookalike to get you the best lookalike you can get.

Which is giving you a clean lookalike because you’re stripping out all of the bounce people from it, effectively is what you’re doing. That part particularly I thought was very interesting because that was very strategic in what you’re doing – to go from the lookalike until you’ve got enough conversions, and then switching over to the conversion lookalike.

That was really interesting, I’m impressed, good job!

CYNTHIA: Thank you.

Read on to learn more about Cynthia's team at Trimble, the challenges hey were facing, how they evaluated Popcorn Metrics,how they use the service, what they like best and what they would like to improve.

QUESTION: What challenges were you trying to solve? Be as specific as you can.

We were trying to gain valuable insight into key interactions people performed on our website. For instance, we were able to create retargeting audiences with the native tools from Facebook or Google, but what we couldn’t do is create audiences composed only out of those people that clicked a particular button, or viewed a certain section of our website. We wanted to get more granular, and more to the target.

QUESTION: How were you solving the problem before you used Popcorn Metrics?

By using goals and custom audiences in Google Analytics and Facebook’s Pixel for retargeting audiences.

QUESTION: What was the previous solution costing your business?

The previous solution didn’t imply any additional costs, but the value of Popcornmetrics comes from the fact that we were better able to understand the performance of our ads through custom conversions, hence we could implement the necessary changes to boost performance. Additionally, by targeting the very hot prospects and excluding bounces or visitors of unrelated pages, we were able to drive better results from our advertising campaigns.

QUESTION: How did you hear about Popcorn Metrics?

Growth Tribe Amsterdam

QUESTION: What made Popcorn Metrics stand out as a fit for your needs?

Simple solution that didn’t require coding.

QUESTION: What was most important to you when evaluating your options?

Simplicity of use, data accuracy and integrations with other platforms.

QUESTION: What was your process for choosing Popcorn? Who was involved?

After attending the Growth Tribe training in Amsterdam, one of the colleagues in our growth team presented the idea to us and together we made the decision to go for a trial. After we were able to explain and prove the value of the tool to our Marketing Director, we got approval to go for the full version.

QUESTION: What were your initial expectations for Popcorn Metrics?

Simplicity in use, accurate data sent to Analytics and Facebook.

QUESTION: How long did it take to get up and running with Popcorn? Was it easy? Who was involved?

Setting everything up, from Popcorn events to Analytics goals and retargeting audiences in Facebook was quite easy. I was in charge of getting started with it and throughout the process got feedback from the team. It didn’t take more than one week.

QUESTION: How is Popcorn Metrics helping you save time, increase productivity or gain competitive advantage?

Popcornmetrics made it very simple for us to track key user interactions on our website. Just right click on the button or HTML element you want to track -> add event, and that's it, your events will start popping up in the Popcornmetrics interface and from there be sent to several other useful assets. For instance, we mostly used our events to create custom conversions and retargeting audiences for Facebook Ads and Google Ads. It was a relief to see that Facebook and Google recognized the events instantly and we could just start working with them, without further coding or tweaking. I definitely recommend Popcornmetrics if you want to get the job done fast, without any headaches.

QUESTION: How long did it take for you to get results from Popcorn?

The results show at least one month or a few weeks after setting the events, because of the time frame required to observe and optimize advertising campaigns based on performance.

QUESTION: Who on your team is using the service?

The whole team is involved in analysing data, but one person is in charge of all operations associated with Popcorn Metrics.

QUESTION: Is there a particular aspect of the product that you rely on most?

Retargeting audiences, custom conversions.

QUESTION: Can you give a specific example of using Popcorn Metrics with Facebook Ads?

• Campaign objective: desktop app conversions
• Custom conversions from Facebook based on website button click data.
• Custom audiences and lookalikes based on visitors and converters.

QUESTION: Can you give a specific example of using Popcorn Metrics with Google Ads

• Campaign objective: desktop app conversions
• Analytics goals for custom conversions
• Retargeting audiences based on goals (visitors that have spent more than 20 sec on website)

QUESTION: How has Popcorn performed compared to your expectations?

Accordingly. I would say the most important benefit that comes to mind when thinking about Popcorn Metrics is that it allowed us to be able to rely on a single set of data that was closely tied to actions that users were performing on the website.

QUESTION: Why would you recommend Popcorn Metrics?

Because of its ease of use.

QUESTION: Anything else you would like to add?

Great customer service! The popup that requires me to install an add-in every time I go into edit mode, is quite annoying because I already have to add-in.

Your Name & Position:

Cynthia Pucheanu (Demand Generation Specialist)
Stefan van Eerde (Demand Generation Manager)

Company:

Trimble MEP

LinkedIn Profile:

https://www.linkedin.com/showcase/trimble-mep/

Twitter:

https://twitter.com/TrimbleMEP

Facebook:

https://www.facebook.com/TrimbleMEP/

Website:

https://mep.trimble.com

How many company employees?

5,001-10,000

How big is your marketing/growth team?

30

How long have you been in business?

35 years

---