This is the fourth installment in "PII," a series featuring the talent that makes the wheels turn in our data-driven advertising world. Read previous interviews with Nazanin Jazayeri, Andrew Klein and Ella Chinitz.
When Freddie Liversidge joined Essence in 2013, the agency’s programmatic practice he was tasked with building out had just three people.
Now it has 25.
Liversidge helped evolve Essence’s programmatic philosophy, which began as a performance marketing channel and now focuses on branding.
“It’s really exciting to get marketer's brand dollars into the programmatic space,” he said.
Programmatic has always excited Liversidge, who entered media as an analyst but wanted to assess the media he bought.
Today, he runs the Hewlett-Packard (HP) account. His team chooses platforms agnostically, but leans heavily on Google’s DoubleClick for Bid Managers (DBM), given Essence’s relationship with Google.
Liversidge spoke with AdExchanger.
AdExchanger: What about you personally makes you thrive in this role?
FREDDIE LIVERSIDGE: I came into the role from a data point of view. I loved the fact that I got real-time information and could make real-time changes. But I think you can come into programmatic from multiple different ways now. People are coming from traditional buying because they love working with publishers and building strategy. They’re pivoting themselves on a contextual basis. You don’t have to be a data nut who wants to look at numbers all day.
How do you use data in your daily life?
I do a lot of racing and Ironman triathlons, so I analyze my workout data to understand my performance. I’m hooked up all the time when I’m running, cycling and swimming. I always joke that I spend my whole day analyzing media data and then I go home and analyze workout data.
Is data-driven advertising creepy to you?
If the advertising is better and more interesting for the person, then I’m not bothered by it.
I used to be a huge mountain biker. Red Bull advertises to me constantly about their mountain biking events. But because it's so relevant to me, I love consuming that information and I don't find it creepy. The creepiest kind of advertising is retargeting that just bombards me with something that I have no interest in because I clicked on a website by accident.
I’m super excited about geolocation data. That's right on the limit of where it starts to get creepy: understanding where someone is at a given time and then what their daily habits are to power our advertising. But, again, if that makes advertising more relevant, it’s interesting.
What does your team look like?
We’ve broken the team into three different pods. Planning and strategy building comes up with the media strategy and how we leverage data for each campaign. Media managers are focused on optimizing campaigns and making sure things go smoothly. Last, but not least, we have an operations person. Programmatic is this beautiful thing, but it breaks constantly, so we have a person that’s ready to fix it.
How does programmatic break?
There are a lot of different parties involved in programmatic. In a PMP deal, we talk to the publisher directly, who works with a supply partner. We have to make sure those plugs connect. Tiny little things will knock that off, such as creative approval or manufacturing errors, like the bid ID is different, or the deal ID is chopped off coming to the demand-side platform (DSP). Little things add up and we spend quite a lot of time fixing them to make them go smoothly.
How often are you doing PMP deals, and what’s your relationship like with publishers?
For HP, about 60% of our media goes through the DSP, and 95% goes through “programmatic” as a title.
This idea of technology removing the need for sales people, I don’t think it's true. Our relationships with publishers haven’t changed that much. I go for less lunches than traditional media people, but we still talk to publishers. We prefer to work directly with them. The conversation has shifted from, “What inventory do you have?” to “What data do you have?” I want to understand what insights I can use to better target my audience.
What’s the reality of programmatic creative?
To be brutally honest, it's a dream. It's 100% doable, but the clients haven’t shifted yet. We're still talking to marketers who live in the world of this lovely storytelling, TV advertising. To say, “Let’s change our creative for all of the different audiences,” isn’t really happening. We’ve had really good traction with Google, but with our other clients, it’s been a battle.
How do you educate your clients on these shifts in marketing?
People are starting to understand they spend quite a bit of money with us to dissect audiences to serve them ads at the right time and the right place, and they all have the same creative. But they’re so set on the way that they build creative that they can't change it quickly.
Quite often we'll be briefed on the campaign the day we receive the creative. You can't then build a programmatic strategy off of that. Getting the media guys involved with the creative team [before] all of that kicks off is key. That's not the way the industry has done it. That’s the change we have to make.
How did you train those more traditional media buyers on programmatic?
Focusing on audience is key. If you’re not buying based on audiences in programmatic, it's a waste of time. I make sure they understand the way data management platforms are structured, the way you can leverage data, what is good data and what is bad data, because there’s tons of bad data out there and what we can do with a DSP to target.
How well do you need to understand technology to work at an agency today?
I don’t think you need to know it coming in the door, but you do need to have a love or at least an interest in it. You’re going to spend a lot of time analyzing good technology and bad technology. You’ve got to have some kind of excitement at looking into that.
Do you feel overwhelmed by the number of vendors in the space?
It’s an arduous task. We talk to a lot of people. We do a semi-annual audit to skim out vendors, or at least bucket into different areas. We’ve got our name DSP [HP uses Turn]. Then we have special DSPs, like Amazon or AOL, that have unique data.