Three years ago, Mediabrands CEO Mark Coad warned of burnout, overload and soaring stress amid a crunched talent market.

Taking the heat out of the kitchen, he said, was not a sustainable fix, because more heat kept being pumped back in. A “complete redesign” was required – with automation at the heart to free up staff to do the smart stuff they thought they were signing up for.

“It's pretty hard yards [starting out in media agencies] but we're trying to find ways to remove some of the menial and less glamorous tasks, not just for our own financials but for the benefit of staff retention and recruitment,” said Coad at the time.

“It’s about looking towards smarter, efficient, and more productive ways of luring the most intelligent, creative souls out of universities and not burying them under media confirmations for a couple of years only to see them change their mind on the industry,”

So Mediabrands, alongside UI Path and Cognizant, set out to automate what could be automated.

Robotic process automation

Robotic process automation (RPA) makes it easy to build, deploy, and manage software robots that mimic how humans interact with digital systems and software. The robots can ‘understand’ what’s on a screen, complete the right keystrokes, navigate systems, identify, and extract data, and perform defined actions faster and more consistently than people – and they don’t need a break, or wish they were doing something else.

Mediabrands set out to automate over 75 per cent of the overall media agency investment workflow and 90 per cent of the media agency TV campaign tracking, monitoring and post evaluation process along with other media investment task. It worked with the tech partners to build three types of bots:

• Trigger bots: Bots triggered by the user via an email template instructing actions.
• Automatic bots: Following a user triggering a bot, these automatically carry out actions once the trigger bot has completed its task.
• Scheduled bots: These bots are scheduled to carry out tasks based on agreed scheduling at a certain time. They require no individual interaction.

The bot combination allowed the automation of the following manual media agency tasks:

• TV campaign set up: Sets up TV buy tracking sheets populating all the relevant data directly from the media plan.
• TV campaign tracking: Every week the bot runs campaign tracking, populating TV buyers’ tracking sheets with updated bought TARPs (GRP’s) and reach.
• TV campaign post: Every week during a campaign the bot matches TV post analysis at spot level populating the investment trader’s TV tracking sheet with actual campaign data.
• TV makegoods: Evaluates performance at a spot level, creating the investment trader’s TV makegood tracking sheet so people can spend more time negotiating with the networks, not running reports.
• Press & OOH bookings: At the investment traders request, the bot loads all the OOH and press bookings into Mediaocean SpectraMD (Australian industry media book, buy and pay system) directly from the media plan and makes the booking financial.
• TV holding exceptions: The bot runs the investment traders weekly and monthly reports in holdings, so they always have a record of activity.
• Budget reconciliation: After invoicing, the bot keeps the investment trader’s budget reconciliation sheet up to date automatically.
• Proforma business validation: This bot automatically checks the investment trader’s proformas and sends the exceptions directly to the relevant person five days out from invoicing. This process is completed each day for five consecutive days until invoicing.

65k hours saved

Since launching in 2023, the program has reduced employee manual burn by 65,329 hours with the bots completing 158,701 tasks previously completed by agency staff.

Time saved since launch represents circa 36 people and it has allowed Mediabrands to re-allocate resource budget into upper funnel craft specialisation roles.

The upshot is higher quality of work, client service standards and improvements in organic revenue while maximising client returns.

Qualifying results

When completing initial project discovery work, a lot of time was spent standardising end-to-end workflow, duplicating roles and responsibilities and carefully measuring what the human benchmark hours were from that improved process.

Mediabrands divided staff into three group quadrants: highly experienced people at that process; people still learning; and people in the middle. That helped create human effort process benchmarks – and means the results and ROI calculations are solid.

With the automation in place, agencies are able to calculate effort improvement based on how long a human would have taken to complete that processes versus how long the bots take.

With tracking stored in the cloud, solution data is extracted and put into a dashboard. The dashboard measures agency efficiency and client group efficiency down to the individual.

Data indicates which bots are used the most and individual processing times. That means which client teams are getting the most value out of the program is a known quantity.

Within that, individuals who are using the system to its maximum potential have been made ambassadors to promote the benefits internally in a bid to ensure all client teams extract maximum value – which is a key part of onboarding and will continue.

Automation = human growth

As an industry, we go to great lengths to employ highly skilled people only to stick them in front of a computer and get them to do rudimentary manual jobs that are burdened through complicated, inefficient, fragmented processes. It's not a smart way to run a business, nor does it return value back to clients.

For every minute someone is completing a manual process, it is a minute they're not delivering value to clients by thinking about better investment decisions or strategic solutions. Nor are they putting in time improving their own craft skill and IP value.

This is only the beginning of positive change through technology.