Q&A with the Director of Automation Solutions at Dentsu Aegis Network on the organization's digital transformation journey
Major advertising agency Dentsu Aegis Network began its automation journey in January of 2018 with the creation of its Automation Center of Excellence (COE), with Brian Klochkoff as one of two founding members, and Catalytic as one of its first pilot automation solution providers. In its first year, Dentsu automated about 60,000 hours of work. Currently, the team continues to scale those solutions.
Below, we talk to Klochkoff about how Dentsu took the easiest path to scaling automation for success.
What has the importance of Catalytic been in your COE’s success?
Our COE wouldn't have matured as aggressively as it has without the great people at Catalytic. They've been the beacon of light guiding us where we should have been going from the beginning—the kind of use cases we should be having, to the point of just navigating discussions, holding workshops. The partnership is at the core of our team.
How did you go about choosing which projects to start with?
When I started the group with Max [Cheprasov, Chief Automation Officer, Dentsu], we wanted to incubate a group, because we have a lot of belief and passion in the future of this industry, in advertising and media, but also how automation can kind of act as the intersection of people, process and technology.
To begin finding these use cases, we would discuss with CFOs at our agencies: “What are the pain points? Do you have something for us to go try and solve for?”
Everything started pointing toward back-office functions, such as client accounting, accounts payable, accounts receivable.
So Max, the Catalytic team, and I started doing some typical process discovery work in February 2018, because we didn't want to slap technology on a problem just to solve for that problem.
You wanted to treat the cause not the symptom?
Yes, we wanted to trace the problem to a root cause, and make a surgical incision where that root was, and watch manual work fall off so that a.) we didn't disrupt the change management at scale, and b.) so we could use our resources effectively.
We discovered the primary issue was that invoices were coming in and no one knew what to do with them, because they weren't syncing up with data in the accounting platforms. We kept hearing about a specific process where the accounts payable folks said, "If the mid-office media buyers were performing this the way that they needed to or we need them to, it would really save us a lot of time. We wouldn't be working weekends, and we wouldn't have to stay late, and we'd have work-life balance."
So we connected on a human level with guidance from Catalytic to understand the holistic process, and that humans were on either end of the conversation. People in the mid office realized, "Oh, there's a small task that I should be doing on a monthly basis. And if I don't do it, people way downstream are severely impacted."
After we introduced that human component, we started to get a lot of advocates in the mid office to give us information on what it would take to introduce automation to that process. And we also received a lot of advocacy from the back-office functions as well.
So our first POC was to introduce automation to that process in the mid office. We knew that this was a process that was done across many different accounts, in many different agencies, but it was taking about 66 hours per month, or three hours per brand for this process to occur. And most of the accounts have around 20 brands.
What happened when you automated it?
When we introduced automation, we had a huge win for one account, and then other accounts said, "We heard that you introduced automation to this process. Can you do it for us as well?"
Now we're looking at scaling it out to about 200 accounts, and that is mostly due to Catalytic's flexibility and quick adoption for introducing new requirements and rapidly developing them on top of existing products.
How many hours of work does that save your employees?
Thanks to the technology Catalytic offered us and the guidance they gave us, we were able to target almost 200,000 hours' worth of mid-office work to be repurposed almost immediately. And that doesn't take into consideration the back-office work that fell off down the line.
How has this inspired ideas for more processes to automate?
Almost by word of mouth, Max and I started receiving meeting invites from senior people at our different agencies that we own saying, "Can you do this for us? We have an automation idea."
So going into 2019, we decided to formalize the structure around how we’re going from a horizontal use-by-use case process, and focusing on verticals across our different agencies.
How did you prioritize these ideas?
We held workshops with people from the c-suite down to assistant levels to show what automation did in 2018 at Dentsu through the use cases and case studies. We explained what we're capable of, what we've done, what we want to do, and then opened it up to those attendees in the latter half of the workshop to say, "Let's go into ideation sessions and extract ideas."
Catalytic actually co-hosted, which resulted in eight successful workshops gathering about 123 use cases, which we categorized across all of the media agencies, accounting for a couple hundred thousand hours' worth of work.
How did your team begin to implement them?
We gave a pitch to our board about why automating these processes would be important, then told them that with our current team, we can do one of these projects at scale across all of these agencies. We asked, “Which ones do you want us to do?”
The chairman said, "You delivered such great results in our POC with Catalytic in 2018, what do you need to do all those categories of work?"
They approved the funding for all of them on the spot. Now we have the resources and we're quadrupling our team to handle those use cases at scale, across all agencies.
What do you see as the biggest value that these time savings give your organization?
I think that if you were to speak to that chairman of the board that we got the investment from, he would point to generating more value out of entry-level and mid-level employees.
Within this industry, we see a lot of turnover and a lot of boomeranging, where people come in for a year and a half. Their expectations are shattered when they're not doing new business pitches, like they do in "Mad Men." Instead, they're entering information from a spreadsheet into a system, into another system, and then watching Excel crash because there's too much data, and then getting frustrated. They leave and they elevate their role at a competitor, find the same thing there, maybe move on to another competitor, or eventually come back to us.
What happens on our end is we lose that knowledge, and we lose that training, we lose that creativity from those people. In an effort to reduce churn, automation is hopefully going to help retain people by eliminating the high-volume, tedious, repetitive work, which is causing people to want to leave, so they can focus on the more human things.
And if we were using people right now to move data across different platforms, the manual risk there is tremendous, which means we're at risk of telling the wrong story.
How do see Catalytic as different than the RPA solutions you have?
Most of the media buying platforms that we have, have either come from acquisition or have just existed for a very long time. We only have a small set of folks who are using on-prem platforms to buy media. We've used RPA for those types of things and it's been successful but it's not scalable.
We saw Catalytic as offering a solution for democratizing automation across end-users, process owners and subject matter experts. I have a full stack developer background, so do some other people on my team, and with our RPA providers, we're able to go into the studio and do things based on requirements that solve for problems without going through process transformation workshops or training people to code to crowdsource our development.
The ultimate goal, is for people to interact with Catalytic in a way that allows them to focus on the smaller day-to-day problems that they experience and build automations to solve them. Where we see Catalytic taking us is doing to Excel what Excel did to physical spreadsheets 30 years ago, by putting tools into a simple-to-learn cloud-based place to solve their problems independently.
What would be your top pieces of advice to other businesses considering automation?
Small, incremental progress over perfection is definitely number one. You're not going to solve for a problem 100% and if you do, it's probably not the right problem to solve. Involve people and workflow with automation, because humans complement machines. The way we see automation is kind of creating a digital exoskeleton for our employees so that they can lift more data, handle things better, and make decisions based on what that digital skeleton is doing for them in a smarter way.
My other advice is to keep stakeholders engaged actively throughout the process to make it less scary. Develop in an agile way to ensure that stakeholders are seeing that small incremental progress and understand that it's a journey. It introduces a human element, and it makes sure that at the end of the day, everyone's here to work and make a more effective and collaborative environment. Doing that just makes the whole process much more enjoyable.