We're going to be discussing the Evolution of the Digital Advertising Agency as it relates to the shift within the martech landscape and how it impacts staffing models.
Stay tuned to learn how they built a technology stack that allowed their agency to pull 450k reports in Q1 while reducing the time spent creating reports and allowing their account manager to focus strategy.
- Learn how agencies are leaning on their Martech stack as a competitive advantage
- How solving for data pipeline and deeper funnel metrics leads to better optimizations
- Plus, how they're automating as much as possible so account managers can focus on strategy
Listen to the Episode
Innovative companies work with us to help them gain an unfair advantage in the highly competitive digital advertising world. We deliver superior results by combining senior practitioners, deep ad industry expertise and our own ad technology in a way that enables our clients to grow faster while hitting - and often exceeding - their ROI goals.
Show Notes and Transcript
JD Prater: Lance and Amanda welcome to the PPC show.
Lance Loveday: Thanks for having us JD.
JD Prater: Yeah, so I got to meet up with Lance and Amanda at Hero Conf just a couple of weeks ago, back in Austin, and I got to learn a little bit more about their agency, Closed Loop. Instead of me giving introduction Lance and Amanda quickly introduce you guys yourselves, and tell us a little bit about Closed Loop.
Tell us a little bit about Closed Loop
Lance Loveday: Sure. This is Lance, I'll go first. I've been in the business for about 20 years. I've been in advertising, and digital advertising for that long. Was one of the first advertisers on Goto.com, which became Overture, which became Yahoo, which is now dead, but also I was one of the first advertisers on Google before they had AdWords. They didn't know how to value their inventory. We buy these impressions, blocks by the quarter and we buy all the overage it was really cheap and found easy to make money at that point.
And then had stayed Paid Search and expanded into Paid Social and display overtime and grown the agency organically and now we manage about 50 million in ads and annually and got clients all over the world. And a great team and we're having a lot of fun in a quiet corner here in Sacramento but doing what I consider to be world class work so in my mind, it's the best of all worlds.
And I'm fortunate to work with the very talented Amanda Evans. She's our chief advertising officer and I'll let her introduce herself.
Amanda Evans: Sure. Thanks. So yeah I'm Amanda Evans. I'm the Chief Advertising Officer, I run the client services team over here at Closed Loop. I am almost at 20 years being in the industry. I think I'm at 18. And yeah, like Lance started out very early on with Goto and Overture and Yahoo which is still around. And our Yahoo reps would be upset if they heard you say they were dead.
Lance Loveday: Yahoo search, all right.
Amanda Evans: Okay, yeah, and like Lance there, happy to be here. Thanks for having me.
JD Prater: Yeah and I got an interesting fun fact. I remember seeing Lance presented here at Hero Conf Portland so if anyone was there at Hero Conf Portland, that was my first run in with Lance. And then surprisingly getting to run into him again in a very different world and a very different space and it's like, Ah, wow, that was like three years ago. Like that was so long and then Amanda spoke last year at Hero Conf on reporting and visualization so.
They're very well known within the industry and so we're actually going to be talking about this evolution of digital advertising agencies, which I think is absolutely fascinating conversation especially in talking with them at Hero Conf because of what they're doing, how their solving it and then what they're using to solve these problems.
I'm going to turn it over to you guys as kind of a first question, and that's really around like what you guys are trying to solve as an agency and then what are you guys using to solve it, because that has drastically changed in the last five, ten, 18 to 20 years that you guys were talking about.
What you guys are trying to solve as an agency and what are you using to solve it?
Lance Loveday: Yeah I mean, it has evolved quickly. I mean, what we're trying to solve for at high levels if we're battling this perception that I think some people have that advertising management is a commodity and that kind of is like a spike in my heart on the one hand, but you know we see the evidence all over the place.
In some advertisers do really well, some advertisers do very poorly and some, majority are somewhere in the middle, and most are blind to the opportunity the money that they are leaving on the table. And so we fight this constant battle of people having this check box mentality when it comes to who's managing their campaigns and how effectively they're spending their ad dollars. So that's a constant battle for us. Just have to fight that perception of commoditization, and we're fortunate. We find clients who kind of self qualify, people who get it and are willing to make sure that they got right person in the seat managing their campaigns.
You know and as for how we do it, you know we've had to evolve as you mentioned partially on the technology side over time, and we built our own technology tool. After a while we got frustrated that nobody has done it and there weren't any good third party tools.
So we built our own data pipeline tool that pulls data from all the front end ad platforms and very importantly pulls data from our clients back end systems so that we can blend it together and do a full funnel analysis in our own environment and get down at a really granular level to do analysis and understand where we could squeeze performance out of these campaigns and then we combine that with a really strong data visualization tools that allows us to tell the story more visually as opposed to having to get you to long tables and so on and tease it out.
And we did this kind of on a speculative basis to begin with but when we started using it, it unlocked all kinds of additional value that we hadn't really anticipated. It made our people way more efficient and started powering all of our report and automating all of that which our clients loved, and our people of course loved they're not having to manually sort data from different places and marry it all together in Excel.
And most importantly, it contributed to massive performance gains. Because partially yeah we had more time to do more strategic things and optimize, but it also enabled deeper forms of analysis and intelligence that we couldn't get any other way. So it's really been a game changer for us developing our own technology, and I know a lot of agencies struggle with how to approach that. Do you build or buy and how do you go about it? In our case, building our own has been the right answer. But it's something I think every agency has got to solve for.
JD Prater: Yeah, I mean even adding to that one, build, buy or staff. I mean I've seen agencies just throw people at it. You know you can get a lot of people coming in, maybe, right out of college and teach them Excel, right, and you can pay them cheaper wages, right, but I think at the same times, I kind of want to hit on a couple of things that you talked about is, solving for the data pipeline, automating as much as possible.
Amanda, I'll kick this one over to you. So when you guys are thinking through on the services side of this reporting in this visualization like talk to me how you guys are pulling this all in, trying to automate it and why that's so important?
How are you automating reporting and visualizations at scale?
Amanda Evans: Yeah, it's really critical you know when we started looking at the way that our account managers were reporting out to clients was maybe five years ago now, we realized they were spending roughly 50% of their time putting things into spreadsheets, to turn it around to give to people. That's just not an efficient use of time when we've got really smart people that don't need to be copying and pasting data into a spreadsheet. And then running a pivot table and then formatting and all that.
We realized they were spending roughly 50% of their time putting things into spreadsheets, to turn it around to give to people. That's just not an efficient use of time when we've got really smart people that don't need to be copying and pasting data into a spreadsheet. And then running a pivot table and then formatting and all that.
So we started looking for different ways to get our data into something that would work. So with the proliferation of data, we started seeing that our clients were asking for different levels of key word report or geography or search query reporting and it was just getting to be bigger and bigger and bigger amounts of data. So now what we've got is, we've got a data warehouse that is pulling directly in from the platforms API's so we're connected to AdWords and Bing and Twitter and LinkedIn-
Lance Loveday: And Yahoo.
JD Prater: That's right. Yeah. Gemini and I don't leave them out.
Amanda Evans: -And all of that data gets pulled in and what I mean by all of that data, I mean ALL of that data. So we're pulling geography down to the city level for every day for the past few years. We're pulling key word level data, search query data, placement data, you know different cuts and slices and dices of social level data. All of that gets pulled in to, we use Tableau but you know different agencies use different type of BI Tools but the BI Tools are critical for enabling this kind of on-the-fly analysis.
Our philosophy is that reporting needs to be able to answer people's questions on the fly as well as provide analysis as well as tell the story. To be able to find the tool to do all that is really difficult. So that's what we've come up with.
Our philosophy is that reporting needs to be able to answer people's questions on the fly as well as provide analysis as well as tell the story.
JD Prater: Nice, yeah that's something you know when you think through like reporting, I've been trying to think through it differently as well. So like for me, this is my internal definition is like report is like the last thing you do, right?
And really what you should be doing is measuring and analyzing and spending your time doing that. And if you're spending all of your time like you were saying, like 50% of your time just getting the data in, just so you can build a report, right, your kind of missing on this measuring aspect and this analyzing aspect, which is honestly, that's way more fun. I would much rather spend brain power on that, so yeah.
I don't get the people that just like are so ingrained in Excel and they don't want to leave it and I get it. It's powerful. I get it. It's great but at the end of the day, let's automate that. Let's use our brains power to solve greater problems like data pipeline and understanding deeper funnel metrics and how do we use that to optimize, right?
Understanding deeper funnel metrics and how do we use that data to optimize
Amanda Evans: Absolutely.
Lance Loveday: Yeah. Well you gotta think about just what the natural limit is on how much ... how many reports you can pull ... how much data you could pull in and what you can ... what one brain could possibly internalize and just to give you a basis for comparison, our CTO published some numbers from the use of our tool. And in key one alone, we ran 450 ... we pulled 450,000 reports. And we pulled in 256 gigs of uncompressed data.
JD Prater: That's a lot.
Lance Loveday: You know no reasonable size collection of brains is gonna be able to process that kind of data. You just can't do it.
JD Prater: Definitely not quickly. Yeah. No, I think that's like ... that's a really good point. You're talking about the amount. You're talking about the volume of the data and so when you guys are thinking through using this data, so like let's say, you guys have got this automated, you're pulling it in like what's that next step now when you guys are thinking about optimizing and blending this data from the front end kind of, in the interface type of metrics, from your AdWords, from that backend data that's living maybe in a CRM or maybe it's living somewhere else maybe for eCommerce in their own data warehouse. How are you guys thinking through the blending and really using that information to drive performance?
How are you guys thinking through data blending and really using that information to drive campaign performance?
Amanda Evans: So we will typically tag all of our ads and ... with a parameter that get passed into a client's CRM system. They will then, either send us some sort of a download of that CRM data, whether it be in the CSV format or directly ... we can connect directly to Salesforce as well. We, then, blend that data with the front end data, so we get a true understanding of a user that came through on a given key word, did they eventually convert to an opportunity and a closed sale. As you can imagine for a B2B business, that's extremely important. We're not just driving leads at that point, we're driving revenue. We're driving qualified leads for the Salesforce there.
So that all happens on automated basis. We set it up once, as the client sends us their reports each week, that gets blended in automatically and pushed up to a web based reporting tool.
JD Prater: Yeah. I like to kind of diving into that. I'm sure that our agency folks over here, probably have two questions. One, like Oh man do I really have to? Is gonna be one because it's so easy to optimize for the lead, right, and not go further because we can track that lead, probably digitally, right, like we can use Google Analytics for example, to track that. But when you're talking about the CRM data, like marketers are probably, like, "Ah that's for sales, right?"
So how do you guys kind of work with clients like to actually get that access, or to get that information? Lance, you had mentioned in a lot of clients that are opting into your guy s's philosophy to begin with, do you think that's really kind of where it starts in the beginning?
Getting client buy-in and access to CRM data
Lance Loveday: It's part of it and you know but sometimes there is an education process. And you know and sometimes its just a matter of kind of showing the kind of waterfall analysis of like, all right, well, here's what we think is happening on a cost per lead basis and it looks like these leads are cheap. But here's bases on our past experience what reality has been, so we'll show what kind of a case that's ... and look here's how the economic look on the front end. And we were buying leads from this source all day long because they were cheap.
But when we ran it down and finally got to, what's the cost per MQL? What's the cost per SQL? What's the cost per close and then what ultimately over time, what's the true ROI? It turned out that as cheap leads, were cheap for a reason. You know they were worthless. And we were over paying for them. Whereas some of those expensive leads, we thought we couldn't afford to buy, were worth it when we should have been paying 500% more for it.
You know and so we show based on real world performance, how our spending mix changed over time based on incorporating that back end performance into our biding and spending decisions. And it completely changed how we thought about the value different sources of traffic.
You know and so we show based on real world performance, how our spending mix changed over time based on incorporating that back end performance into our biding and spending decisions. And it completely changed how we thought about the value different sources of traffic.
It also had the beneficial effect of changing the Salesforces' perception of the quality of those leads so they ended up working the leads what much harder. And we got an additional multiplier effect on it that way. And so you combine all the those things together and you get into a really virtuous cycle and that's the Holy Grail of all this force so. Once we can tell that story based on actual performance data, and we tell people that, "Hey, the action required to actually instrument things and enable this, it isn't that hard. It's doable." People generally start nodding along going, "Yeah, okay we should totally do that. That makes sense." And we have a solution. It's pretty turnkey.
So it doesn't create a lot of work on the client's side so we are able to not just preach at them and tell them this is how it should be done, we're actually trying to do that heavy lifting for them. And show it for them. And then over time, clients kind of build their own internal systems around that as well.
JD Prater: Yeah I think, you know, where we're using Zoom right now to record this and so right behind Lance is a map of the Martech landscape put out by Scott Brinker and you should see this explosion, right. If you haven't seen it, I think this slide just came out last couple of weeks. I think it's close to 7,000 now. When we think about marketing, when we think about where it's going, there so much data and I would say it's getting easier to track. It's getting easier to automate. It's getting easier to measure and I think the advertisers like yourself that are building out these types of solutions are the ones that gonna succeed because I think you're already ahead of the curve.
I was talking with Andy Taylor whose now at Merkel, and he was at RKG before but they started out with the warehouse. Like that was what Merkel bought them for.
Lance Loveday: Yeah.
JD Prater: They're tagging down to the key word and they've been doing this and they've been able to run these type of reports. I mean, that's what the Merkel reports are, so valuable is because they're able to use a data warehouse to be able to use a visualization tool to run these types of analysis. And I think that's something like we could all understand, we can all grasp, but you guys are actually using it to drive performance.
Before this, we were talking about some quality score stuff. Tell me a little bit more about that story 'cause I thought that was fascinating because I'm sure everyone's rolling their eyes once I said quality score and that it actually matters.
Gleaning insights from Quality Score
Lance Loveday: Yeah, you know its funny. I mean this is something that has kind of reared its head again for us more recently. Partially because we've had kind of deeper data access and analysis and we've been able to correlate quality score to performance a little more closely. And it's just driven home for us via, a lot of the audits that we've done for people over time are just ... how low a starting point so many can be in from starting from on a quality score basis and I think it's a really under appreciated drag on a lot of campaigns performance and people are kind of blind to it. And so yeah, I kind of roll my eyes a little bit about quality score because it has been discussed and almost beat to death over time.
But having said that, we've seen repeated now via so many of these audits that the average weighted quality score for, even larger advertisers sometimes, is really low and is absolutely a drag on the economic of their campaigns and again via case studies we're able to demonstrate, say, "Look, this is what has happened in terms of the campaign economics." When we've been able to improve average quality score by even a point or two over time, it's just had a massive impact. So I think for as much ink has been spilled on quality score for as much as people might be tired of hearing about it, it isn't enough to just nod your head and say, "Yeah, yeah I quality scored your report." You got to take action. If it turns out that you're suffering as a result of those things and when you do, it must creates so much upside potential. You can't ignore it.
JD Prater: Yeah. When you guys are kind of thinking through these deeper funnel metrics, you guys are automating, yet you have a great visualization tool within the Tableau and you're freeing up your account managers' times, so how do you guys think through staffing? What are they doing now with that 50% of time? Talk to me a little bit about that. I mean, Amanda, you ... probably under your purview here. How do you guys think through that?
Amanda Evans: Yeah you know it's interesting. We were originally thinking the 50% of their time would be able to be spent on, maybe, an additional client. That hasn't happened. And the reason that hasn't happened is that 50% of their time is now shifted over to being more strategic. They're able to analyze deeper, much, much deeper levels than every before, which I think is tracking along perfectly with the industry. The industry is giving us more and more targeting options. And so it's been a good change for us because that 50% of the time, now they can actually identify, which of those targeting options need to be used? And start to use those levers strategically.
It's hard to find a lot of that information within the interfaces or as we said before, downloading it into Excel. But by having the reporting automated, by having the ability to drill down and filter and segment in real time, that's made our account managers much more strategic, I think. So we didn't get to use the 50% of the time on other clients but that's okay.
JD Prater: Oh I'm sure the clients appreciate that.
Amanda Evans: Yeah.
JD Prater: But kind of like the next phase of that, right, so now they're being more strategic. Talk to me about how even changed for like a staffing model within an agency, within the last five years of thinking through who you guys are hiring now and the types of people you guys are hiring now.
Amanda Evans: Yeah, we're hiring mostly for a mindset rather than skill based. So think back five years we were looking for somebody who knew AdWords inside and out. Now, we're looking for somebody that understands how numbers fit together, how metrics are working towards improving a campaign. So I think the skills and how the mechanics of how you build on ad group, that can all be taught. We're looking for somebody who is driven to identify, kind of mirrored anomalies in the data, think through questions that a client is gonna ask, and really look at this from a qualitative perspective rather than just pure mechanics. So I'd like to think we're looking for somebody that can think a little bit bigger rather than just somebody whose in there doing the daily bits.
Lance Loveday: You know think of as the systems thinkers, right, people who appreciate how changing what input is gonna change other downstream metrics and impact the ultimate Northstar metric that we're optimizing to as well but who appreciates the quantitative end and the quantitative components that come into play. And I talk about how over time people develop the art of advertising management and I really believe that there is an art and creative components to the work that we do and it's not fully quantifiable, it's not something that is easy to build an algorithm for. And for that reason, I actually feel pretty good about our ability to continue to differentiate and to not to fall into that commodity trap. Because we're already using people's minds for what people are good for and that's the strategy teasing out meaning, understanding the interrelationship between things. And then we're using the technology obviously but technology's good for.
JD Prater: Yeah I mean it seems like you guys are having a really good frame of reference for what can be automated but also like what can we focus on as humans, right? There are things that if you're spending five, six hours a day pulling reports in Excel like, you're probably going to be automated in a couple of years but the things that won't be automated are that strategical thinking. I still maintain that and I think the people that can learn to use the technology, and use the automation tools, those are the ones that are gonna succeed in my opinion kind of moving forward.
So I think you guys are doing a really great job. As far as kind of like wrapping up, what would you guys like wanna leave everyone with? You know like what would be your main take aways to agencies or to in-house teams as they're, maybe, some are on this journey with you guys and trying to build out their own text techs. I know many are not. So how do you think through that?
Lance Loveday: You know I would encourage people to just start experimenting. For us, if you'd have asked us five years ago if we were gonna have our own technology tool that might rival what's some of these other agencies have built over time, I would have said, "No way. We're too small. We can't afford it. It's too hard." But the reality is we got frustrated enough that we decided to go ahead and take a shot at it and you know God Damn It, if we didn't pull it off, you know, over time.
But it had to start somewhere and in retrospect, God, yeah, I wish I started a lot sooner. And so I would encourage people just to take the first step. You know start download a trial of Tableau. Get to familiar with it. Start to think about ways you might be able to use that. Hook it up to a tool like AdStage. I mean, we're using AdStage to power part of our funnel, right? You know you can absolutely kind of do your own home brew data pipeline and visualization platform and I would encourage everyone just to take that first step. That would be my take away.
JD Prater: Yeah. Amanda, do you have anything you want to add to that?
Amanda Evans: Well I mean, I would agree. I think taking the first step with the client in mind. I think you know advertising agencies are much more than advertising agencies at this point. We've transitioned into really consultants on a business front and I think using a lot of the data automation tools and data pipelines that are out there help us to continue in that capacity. That's all I'm gonna say.
JD Prater: Well fantastic. I mean great advice. Thanks for taking us along that journey of this evolution as we kind of think through. I think we really are in this kind of this turning point I would say in 2018, specifically where I think we say a lot trends bubble up into 2017 around CMOs wanting to track revenue being held responsible for revenue. And I think we're starting to see people, agencies, in-house teams really starting to think through especially on the marketing side of how we can track further down the funnel.
You guys have been doing it for five years which is absolutely incredible to think through. But where can people find you guys? Where can they reach out if they had some questions? I'm sure there are people that are gonna wanna ask questions about your stack and what you guys are using, what tools. I know that was always a big question. Where can people find you?
JD Prater: And you guys were talking about who you guys are hiring for? Are you guys currently hiring if anyone is listening?
Lance Loveday: We're always hiring. Phone yes. And in fact if you could advertise for that on all future podcasts, I'd appreciate it.
JD Prater: You got it. You got it. All right. So thank you guys again, for coming on, talking us and then for those listening that's closedloop.com. You guys can find them on Twitter as well @closedloop, and they are hiring. So reach out to Lance and Amanda if you guys want to learn more about what they're working on or hey, if you want to become a potential client of theirs. They are very smart people, and I think they're gonna do a really job for you guys.
So that's it for today's show. We will see you guys next week.