Patrick Gilbert was invited to speak at the Friends of Search conference, a two-day event held in Amsterdam and Brussels. Friends of Search is the largest search conference in Europe, bringing together nearly a thousand PPC advertisers from across the globe to learn from Google, Microsoft, and other industry experts. Patrick's presentation, titled "Be More Like SpaceX: The Elon Musk Approach to PPC" presents a new budget framework that PPC advertisers can use to profitably scale their ad channels and achieve their long term business goals.
The Falcon Budget Framework is an approach that AdVenture Media has been piloting for more than a year. It has been instrumental in helping improve communication, expectations, and goals between the agency and our clients.
Host: Would you give a very warm round of applause for Patrick Gilbert from AdVenture PPC, all the way from New York City, ladies and gentlemen!
Patrick Gilbert: Thank you. Hi everyone, it's great to be back. My name is Patrick Gilbert. I'm the COO of AdVenture Media Group. I'm based in New York, but our team has more than 50 people worldwide. We're a digital advertising agency that helps clients like Sports Illustrated, Sennheiser, Forbes Magazine, and others. I'm also the author of Join or Die: Digital Advertising in the Age of Automation. It's a playbook for how to profitably leverage automation in both Google, Facebook, and other emerging ad platforms.
It feels so good to be here for so many different reasons. It's great to be back in a room filled with energetic, exciting marketers. I have really missed this. But another reason I'm really excited is because today is actually the one-year anniversary of the publication date of Join or Die. After all this time, and all this quarantine, writing and talking about this stuff in private, to be able to be here talking to all with you today really makes it feel full circle and fantastic. So I appreciate the opportunity.
As Reuben alluded, what does SpaceX and that crazy Elon Musk guy have anything to do with PPC? To get us started, I do have a question. I have a copy of Join or Die here to give out to someone that's brave enough to answer the question. But does anyone know what the point of SpaceX is? Does anyone know why they exist? What is their ultimate goal? Hand at the back?
To colonize Mars...that's correct. The ultimate goal of SpaceX is to put one million people on Mars, which is quite the vision. But we don't currently have a million people on Mars, and SpaceX is still generating money somehow. So, does anyone know, what's their current business model? The right answer is that they have a space delivery business. They're basically Uber for space.
Private companies and government agencies will pay SpaceX to deliver cargo, mostly satellites, up into orbit, which sounds great. But is Elon Musk settling for less? Did he set out to colonize Mars, realized that that is a really hard thing to try and solve, and then just say, "You know what, let's just make a space delivery business."? No, that's not quite. That was actually part of the plan all along.
Because if you really want to do something as complex as colonize Mars, you need a lot of research and development. You need to get really damn good at launching rockets into space. You need to revolutionize the cost of space travel. And you also need to, in the medium term, generate a ton of money.
So the space delivery business kind of is a two-birds-with-one-stone accomplishment. It allows them to practice and get this research and development and test new technologies. But it also generates capital, in the short term, and that's fantastic. This whole idea here, colonizing Mars, is one of the most complex problems that anyone in human history has set out to solve. But it becomes a lot easier and more simple when you break it down into smaller parts.
This is a simplified version of SpaceX's business plan:
Phase one identifies the first 15 or so years of their existence. They basically had to figure out how to get stuff up into orbit. Phase two is where we currently are. This is the ongoing space delivery business. Now, this generates a profit in the short term, which is great but there's a few components here. They need to drastically reduce the cost of space travel. You'll see here that the goal is to bring down the cost per ticket to $500,000.
Now, that was the price that Elon Musk figured out was the breaking point to be able to get the first group of settlers up into Mars to let this whole process take place. But in the early 2000s, it was estimated that a ticket to Mars could cost $10 billion or more per seat. So when they set out to do this, they figured out, okay, well, we need to reduce the cost of that ticket by 20,000 times. And that's insane. How could you possibly reduce the cost of something by such a large factor?
And this is something that PPC advertisers ask themselves all the time when they see very high CTA numbers that don't really match up with their business model. So they didn't just give up. They didn't say, "Okay, well, that's not going to happen. We're going to throw up our hands," or PPC advertisers often do, we might blame Google for the high costs, or we'll say that PPC doesn't work for us.
No, they stayed the course. They realized simple optimizations will make this economical. "We're going to have to invest in technology." They were willing to make the investments. And that's really what I'm here to talk about is the way to frame your testing and your investments in R&D to be able to ultimately achieve your grand vision.
The best example of this is their attempt to reuse materials. For 50 or more years, every single rocket, regardless of who set it off, put into orbit, most of the materials were one-time use, which is pretty expensive. So the rocket in this photo is called the Falcon 9:
Does anyone happen to know what makes the Falcon 9 so unique? This freaking rocket lands straight up on its own after being in outer space! This is incredible! This is a miracle of modern science. But it's also really good for the bank account.
This wasn't just something that could be solved overnight. It certainly took a great deal of investment and testing. And the key here, though, is that the real focus on it, it wasn't just about, okay, how do we get technology to get us to Mars? It was about how do we reduce our metaphorical CTA in the short term?
So I thought this was really impressive:
I really liked this idea. I was really blown away when I first discovered it. And this approach can be summarized by the following: you first break down your most complex problems into simpler parts. Let's call them phases. Then you have to identify the most pressing challenges and goals of each of those phases. And then, and most importantly, you have to manage expectations throughout the process.
So, does anyone here happen to have an ultimate goal that is more complex than that of SpaceX? Does anyone here attempting to colonize Jupiter perhaps?
Probably not. I don't think so. So why don't any of our PPC strategies take on a similar approach?
So what I want to present is the Falcon Budget Framework, named, of course, after Falcon 9 rocket (which, to be fair, Musk stole the Falcon stuff from Star Wars anyway). The Falcon budget framework allows you to do five things: add clarity to your long-term vision, identify your starting position, break down your most complex problems into smaller parts, identify the primary challenges and goals of each of those phases, and, most importantly, manage expectations.
Regarding long term vision, I won't spend a lot of time on this because every single business is unique. But it's really important to be specific about that and be very clear with all stakeholders, whether it's the CEO or talking to an agency or just talking to the rest of the team. It should be specific. Oftentimes, this goal will be a specific revenue target with an EBIDTA framework.
For the sake of this exercise, we're going to assume that your ultimate goal is to colonize your market:
But the question is, where do we start? Because not all of us are starting at point A. Most of us are already doing something and advertising. We're off and running in some capacity. So this is what I refer to as our new client matrix:
Where you are in the matrix, currently, depends on the quality of data that you have access to in your accounts, as well as how profitable you are.
On the lower left-hand side, we have dark horses. They're new to the game. They have little to no data and they're not yet profitable. Above them are sprinters. They are also pretty new, but they've seen some early success. But they probably don't have the right attribution tools in place. They're going to have a hard time scaling because they probably don't have process and organization and the quality data that's going to allow them to scale.
On the lower right, are opportunists. These are ones that have been around for a little while. They might have seen some success. They have some data, but performance is either starting to decline, or leadership just feels that they should be doing better than they are.
And then, finally, you have incumbents. They're in a good place but their primary challenges are usually related to scale. We often hear incumbency things like, "I can't spend more on similar CTA." Or, "Every time I try to increase my budget, my return on ad spend goes down."
I think incumbents are probably the most challenging clients to work with from an agency perspective. Because I think really anyone in PPC can somehow provide value to any of the other three categories here. There's kind of a process that we all know to be able to do that. But to really prove your worth to an incumbent, you need to be able to solve complex problems. And that's where I think the Falcon frame work is particularly useful.
We're talking about quality of data here:
I don't care if you've been running Google ads for 10 years. The things I care about are what's the first-party data, CRM data, or whatever you're doing, post-click that you're importing back into Google? And how can we leverage that to scale algorithms? That's the stuff that really matters. And that's really what's going to make the difference here.
Before the automation era, PPC advertisers were able to kind of use their intuition. You didn't need to import your CRM data back into Google because you kind of knew what worked and what didn't work. But success in the modern era requires a seamless feedback loop because your data plus confidence plus profitability, equals progress.
So here, we have a dark horse:
They're starting out. They have an idea: We should do PPC!
I think that's a great idea! So where do they start? What is the test budget they start with? What is the first campaign? I think a lot of this, this probably isn't the answer that you guys want to hear, but it's really gut instinct. This comes with experience and saying, "Hey, listen, I think we should take a certain budget and we should put it in this direction."
For a lot of advertisers, that's going to be bottom-funnel search, or maybe even branded search. For an e-commerce store, it might be Smart Shopping, or maybe a look alike audience on Facebook. You just need to get going with something and have reasonable expectations to try to get that going. The real goal here is to acquire a single customer. You're really looking for a proof of concept. You're trying to prove, can I be successful? Can I get customers on this channel? Can I get customers with this campaign type?
Regarding expectations, this is generally where people start to mess things up. We might be launching a bottom panel search for a brand new client and a dark horse. And they might say something like, "Listen, I know that there's a learning curve here. But how long is it going to take until we see some success?" And I would answer that question by saying that that is not how you should be thinking about this test. It's not about when will this be successful? It's about when will I know if this could ever be successful? Those are the tools that you're looking to solve.
You're looking for when will we know if this can be successful or not when will this be successful. And this is a really important point because we're going to have a test budget at every stage of our framework. And this expectation about what happens within the test budget needs to be consistent throughout every step of the process.
So this is our initial test budget:
The goal is to acquire a customer; a proof of concept. Well, then what happens? Well, eventually, we find something that works, something that isa little bit repeatable, something that we can rely on versus some consistent customers. That is what I refer to as High Confidence Response Campaigns.
Examples of high confidence response or HCR campaigns: most search campaigns, shopping campaigns, some Facebook campaigns. Basically, your bottom-funnel; we all know what this means. Basically saying it's a pretty linear process. You can get a reliable source of customers from this on a consistent basis. Your CTA and return on ad spend targets are going to be somewhere within your profit expectations but volume is finite. There are only so many customers that can be bought in this category.
Now, when we say confidence, we're really thinking about how do we measure success? What does that really mean? How easy is it to measure success? With high confidence response campaigns, it's a pretty simple process. We can usually just rely on Google to tell us whether or not it's profitable. For search campaigns, this is pretty much what we've always relied on. And for the most part, it works. It's not perfect, but it works. So if you're more confident, the attribution process is simple. It'll probably come with a High Confidence Response.
The yellow is your HCR campaigns. They generate an ROI in short term. This is your Uber for space, your space delivery business. This alone isn't going to help you achieve your ultimate goal but this is a necessary step. And what's important here is you also have the test budget. This is fueling the research and development that's ultimately going to be used over the long term to help you scale. And a lot of times, these two things work hand in hand. That's the beauty of where we're at with the current technology.
Because Google, in particular, will aggregate your signal data at the account level, meaning every click and conversion that you get from any of your campaigns, branded, whatever, all that information is going back to feed confidence into the algorithms. So the larger your data gets, the larger your account gets, the more data you get, the more confidence you can have as you scale your test budget over time.
The next set of campaigns: modeled confidence response. It's going to be a little bit more complicated. The examples here, YouTube, display, discovery campaigns, most Facebook campaigns. We're all familiar with these. We've all decided they don't work. The real value here is harder to ascertain. It's foolish for us to think that we can really judge the value of a YouTube campaign by looking at the conversions column. The real value really has to be modeled out.
This is something that you might need to be doing a regression analysis with somebody on your team that can do that. Maybe not me. That can actually look a control for certain variables and be able to say, "Okay, we're measuring this budget against a KPI like total site visitors or total site traffic, or total site revenue," or different KPIs that you can also measure.
And you're probably thinking, "Okay, he's just describing branding advertising and trying to describe it as something else now. And that's not what I'm doing. This is not brand advertising. And that's why it has a different name. And this is why the Falcon framework is, I think, so interesting and so unique. Because, for about100 years, brand advertising and direct response advertising were mutually exclusive. Any campaign you ran or budget you spent would fit very nicely into one of those two categories. But in the modern digital era, that's just not the case anymore.
You can take action, you can have a response anywhere at any time in any setting, whether you're sitting on your couch at home, or if you're commuting to work, or if you're standing inline in the grocery store. So really, any ad message now is not going to clearly fit into one of those two categories. They're more intertwined than ever before. So we need to get rid of our previous connotations about how this stuff work to look into a modern lens.
So the real value of the MCR campaigns is realized in three different ways. The first is actually direct response. Becky gave a presentation earlier about displays, shows. Listen, sometimes these campaigns actually can drive action, can drive short-term conversions. We're measuring it in the Google Ads conversions column. And that's great. These things actually happen.
But just that alone won't give you the full picture about the value that it's having. Sometimes, a lot of times rather, display ads and YouTube or really any sort of advertising is having an additional value, and it's swaying the decision-making process and you're not measuring. It's not coming back through Google attribution the way that you'd love it to. And this is referred to as an "halo" effect.
I actually want to give a shout-out to James who spoke this morning about share of voice. He helped me articulate this a little bit. But the halo effect is the idea that if you show somebody a YouTube ad, and it's a good YouTube ad, you get your message across, and it's received well by the audience, if that person then goes to Google and performs a search, and one of your ads shows up, the expected conversion rate on that ad is now going to be greater as a result of that YouTube ad. The two are working hand in hand. They're influencing one another.
And in the modern age, expected conversion rate is our currency. If we can influence a higher expected conversion rate in any of our ads, then we can bid more aggressively, which will allow us to enter more options that were previously unavailable to us. This is how we scale. This is how you incrementally scale your shopping campaigns, your search campaigns. This is how you answer the questions of the incumbent when they say, "You know, every time my CTA just goes up."
It's because they're probably not investing enough or realizing the value in these other sorts of ad mediums that are actually having an impact. This is how the whole ecosystem should work.
The real point here is that MCR campaigns and HCR campaigns are apples and oranges. it is foolish to try and think of these two things in the same boat.
MCR campaigns, we introduced them, the main goal here is that we should also be looking for opportunities to reduce our HCR CTA. This is your version of figuring out how to land a rocket straight up. Opportunists generally fit in this phase, but they will cut off their MCR campaigns because they're not measuring it correctly.
And finally, we have our incumbents. They also will panic every time they see a CTA increase. But with the right framework, you can allow them to be able to see this through. So this is the Falcon framework in all of its beauty. This is how we look at how to solve complex problems and scale our client accounts.
We're going to go through a couple of examples here. I think I'm already running short on time. But first, let's review.
We have the new client matrix, which shows us that where you're at currently depends on your current level of profitability, and how much quality data you have in an account. And then we have the Falcon budget framework, which shows us that R&D, which is really data, plus confidence, plus profitability equals progress.
So we start off, we have our test budget, we're dipping our toe in the sand. We're not expecting a very high return on ad spend.
As we go into phase two, we're introducing our high confidence response campaigns. But we also don't give away our testing. We don't give that up. We know that that's going to be a part of the process, and we continue to invest there.
And then eventually, we start to increase our MCR budgets. We're getting more data, we're doing more testing.
But what does this mean practically speaking? So let's give an actual example here, not just something that looks good on a model. Let's say we're an e-commerce store; we're starting now. We're going to start with this adorable little smart shopping campaign.
The goal here is to acquire customers—proof of concept. After a while, we figure out how much we could spend in smart shopping, we get somewhere in the profit matrix, and we start getting customers on a reliable basis.
And then we start another test:
In this case, maybe it's a remarketing campaign. And we have reasonable expectations about the outcomes that we're going to realize from that campaign. Eventually, that remarketing campaign gets optimized, gets tweaked, maybe reduce the budget a little bit, and now it's moved to the MCR stack.
We introduce another test. In this case, a non-branded search:
Eventually, we turnaround and our campaigns will look something like this:
This is what our account might look like. We have a couple of different search campaigns, a couple of MCR campaigns, and another test because testing never stops.
Now, let's say this e-commerce store has a profit margin of 40%, which means that their breakeven return on ad spend is going to be 250%. But it's reasonable to expect the profit. So they ideally want to see a return on ad spend of 300%, which is totally fair. The HCR campaigns are averaging a 300% return But these other campaigns are dragging down the overall account below that.
So most people, before they've heard of the Falcon framework, they'll come to us and their account will actually look something like this:
This is how most people see their accounts. This is how they're understanding. They're focused on the return on ad spend in the short term, and they're not investing in the future growth of their campaigns. Campaigns are pass/fail. They either work with our short-term understanding of return on ad spend, or they don't work for us.
And that's not how like to look at things because if we really look at this, let's say we put out these new test campaigns. We have the MCR budgets doing what we believe them to be doing, in a positive way.
But this is what the actual return on ad spend looks like, 230%.
That's not just below their goal of 300% but this is actually not profitable. This is losing their money.
So let's say we've done a full round of optimizations, whatever our fixed period of time is. Let's say it's a month or a quarter, however you guys like to measure these things. And they're like, "What are you doing? We're losing money on this? I'm not into this." And you say, "No, listen, we have to stay the course."
And here's how we're going to break this down. We're going to look at all these different things individually. Our HCR campaigns at 300%, MCR at 200%, and our test campaign 100%. Okay, let's look at this if we just exclude the test budget altogether and then group MCR and HCR together. In this case, we're not in our ideal 300%, but at least we're profitable. So, okay, we'll see how this goes. We'll continue to invest in this whole budget framework thing.
And you might be looking at this and say, "You can't just exclude things to make your numbers look better. That's not fair. No client is going to be okay with that."
I disagree. I challenge them on it and say:
When you were here...
...you were here as that aspiring bushy-tailed, bright-eyed Dark Horse, you were willing to invest in PPC. You leveled your expectations because you knew that this was very important to be able to achieve your long-term vision. But along the way, you changed your expectations.
You thought you'd be able to get to Mars without investing in new technology, without revolutionizing the cost of space travel. And that's just not how it works. And that is why it's so important, at the beginning of this whole process, step one was identify your long-term vision and be very specific and clear about that. Because if consistent results or a space delivery business was your ultimate goal, then that's fine, and you should be very happy with that.
...but if you really want to stick to that ultimate goal, it does require continued investment.
So this is what our campaigns look like after the second round. We now have top-funnel searches graduated into the HCR campaigns. We're testing the YouTube prospecting campaign. Let's look at the results.
The overall return on ad spend has stayed about the same across the board. However, we've just spent more. So we're generating more revenue, but we're still operating at a loss. And our clients are now going to start to say, "You guys don't know what you're doing. We're losing faith. We got to start pulling the plug. We're shopping around for a different agency." And that is where the expectations come into play. That is why is it so important to be able to understand the way that these different segments and types of campaigns benefit and work intertwined over time, and how it's important to be able to scale.
Eventually, we get to phase four. We now have more MCR campaigns, we're spending more. But the key difference here is that our overall return on ad spend for our HCR campaigns has risen from 300% to 350%. Let's look at the results.
We're now at a 270% return on ad spend across the whole board. There's the segments and results. And then if we just exclude the test budget and look at everything, we're at 300% return on ad spend. I know this isn't perfect but this is what it takes to scale. Because if we look at these three rounds and chart this out and actually measure profits, given the 40% profit margin and everything else we've done over the last couple of slides, this is actually what the growth trajectory would look like here.
And these two dips, they're scary, they're terrifying. There's a lot of uncomfortable conversations that you have to have with your clients during those periods of time. But if you really are invested in this, and if you're willing to figure out how to get that long-term vision, then it's absolutely necessary to be able to achieve that.
So my last question; how would Elon Musk manage client expectations? It doesn't seem fair that this guy gets away with this 30-year vision. He has no idea how much things would cost or how long it's going to take. And I think it's really interesting because he said he'll never take the company public.
Now, a public offering would allow them to generate a ton of cash that could fuel R&D. But he's been vocal about this and said that the problem is that the markets operate with a short-term time horizon. Everyone says they care about the long-term vision, saving humanity, until it's time to report on quarterly earnings. And the reason that they'll never go public is because he wants to make sure that none of the stakeholders or none of the employees ever take their eyes off the grand vision.
And I really admire that. Honestly, I envy it. And I think that this is the most important takeaway of the entire SpaceX model. Because if you really do want to achieve that long-term vision, it does take an investment. It does take those periods where it's not as profitable. And all of us here know that but we still put up with it.
We know that it's not possible to operate with short-term expectations and these crazy time horizons. Forget quarterly results, we're reporting weekly, daily. How many times have you received a call from a client midday saying, “We had a slow morning, we need to change things immediately.” That doesn't work.
So it's time for all of us to really buy into these concepts; to draw a line in the sand and say: these are the ways we're going to set expectations moving forward. These are the ways we're going to break down campaigns in a different way. We're not going to think about them in terms of brand campaigns or direct response campaigns. The modern era is going to require more of us.
Host: Thanks very much, Patrick. We have a couple of questions. Stay on just for a second before we go to our penultimate speaker today. And actually, the main question was, from an AdVenture point of view, to your clients have to buy into this in order to work with you? Are you as strong with your clients as you are on stage here, I think, is the point?
Patrick Gilbert: I'd like to think yes. I think that we're having more tough conversations with clients than ever before. The issue comes, I think, our points of contact that are hiring us. We talk about this during sales pitches. So the person that we're working with will likely buy into it, the problem will come if their boss won't understand.
So that's something that I think is a real concern. We have to continue to simplify this messaging and be able to figure out how to get this up the chain. And that's where I think the rest of you come in. But yeah, I think that we're pretty hard on this.
Host: And one final question before we have a little leg stretch before our penultimate speaker, how would you measure more upper funnel campaigns? On total revenue or goals?
Patrick Gilbert: It really depends on what you're doing. You have to figure out...if you're advertising content and if you have good content on your site, then you should be looking at things like repeat visits and time on site. If you're measuring something else, if you're just an e-commerce store trying to get traffic, I think I would look at it a different way by saying, how can we breakdown affinity categories to understand this is our target market? How can we understand we're getting the right people to our website?
So there's all these other metrics in these accounts that oftentimes we ignore because they're not necessarily a part of the KPI that really show profit. But that's when you answer the stories. It's all about the narrative saying, "We know that this top funnel campaign is good kind of because we get a sense that the people that are coming through that campaign type are the kinds of people we want to be talking to." So it's a lot of proxy metrics. You have to really figure it out and use your best judgment.
Host: Thank you very much, Patrick Gilbert. He's around for the rest of the time together so grab him some more questions.
We'll get back to you within a day to schedule a quick strategy call. We can also communicate over email if that's easier for you.