Why and When to Automate Media Buying

Imagine you are the marketing lead at a startup. Your team comes up with an idea, and you test it, pivot once or twice, and boom! You have success!

LTV/CAC = 4 or bigger

You think of scalable channels like Meta Ads, Google Ads Campaigns, you name it. You hire a creative producer and a media buyer… two media buyers, three? You are scaling the channels successfully. Looks good, right? You are on fire, but you’ll like it.

Digital marketing has a dramatically dynamic nature. Algorithms are constantly changing and updating, new rules are continually emerging, creatives are fading out, and you must develop new approaches, etc., not to mention new channels.

You feel that your team is drifting to constant overhead, and occasionally, you may see yourself drowning in the routine, and your team may start to lose motivation while failing to see ways to grow professionally. It may become more challenging to have a helicopter view and develop new ideas for tests and so forth. Sounds familiar?

Hiring more people may look like a good idea; however, justifying such a headcount to the board and finance advisers is not an easy journey under any circumstances.

As such, we see that most media buying operations are routine processes. These processes may be standardized; the more standard they are, the quicker your team performs these tasks and the fewer mistakes they make.

Thus, you’re faced with two options:

• Try to justify the headcount, hire one or several media buyers, and struggle with helping them with motivation and career growth;
• Automate your routines with no-code or code automatization.

Well, automation is another type of task for you and your team. When will it become an excellent option to invest time and resources here? Usually, when you have:

• A product-market solid fit for your main product (use-case);
• One or several scalable digital marketing channels with sufficient LTV/CAC ratio;
• Routing tasks to manage each channel take at least 2 hours per day.

Let’s consider the possible benefits of this approach.

• Optimize FTE (Full Time Equivalent) per channel. You’ll decrease or maintain the exact headcount by performing more tasks, which makes the board of directors happy.
• Spare time for analysis, ideation process, and other things that help you grow the product via marketing and help your team grow professionally.
• Decrease response time to typical situations (high CAC, creative burnout, relaunch of ads, budget maintenance, etc.)
• Decrease the number of inevitable mistakes during manual operations due to human nature.

Absolutely, along with the pros, there will be cons, such as:

• What to begin with? You may struggle with questions like “How to start?”, “Should I hire a developer or a project basis?”  “How do I describe what we need?” or “How do I support it through time?” etc.
• What should the prototype perform? How do we build the roadmap?
• Communication with the team: who should be responsible for which part?
• Auditing system: what should stay manual while other things are automated? How do you catch issues before it’s too late?

At first glance, these issues may look scary and outweigh the benefits. But let’s take a closer look at those benefits. For example, the FTE optimization may lead to not only maintenance of headcount but an opportunity to decrease headcount; thus, you’ll have the opportunity to hire much more qualified professionals who not only manage the particular channels better but also make a significant contribution to the marketing strategy or the channel growth strategy. Thus, the share of “creative” tasks will increase, which usually motivates the team.

And decreased response time will help you cut the losers and shift the budget to the winners’ hours quicker because even if the person performs the daily auditing, they work only 8 hours a day and another 16 hours each day the things as they are, not meaning the weekends and holidays.

Let me share a particular case from Scentbird regarding these two points.

Case Study

Scentbird is a subscription-based fragrance service that allows customers in the United States to discover and try new perfumes and colognes monthly. Scentbird utilizes the subscription model and business growth directly connected to active user base growth. It was critical for business margins to optimize customer acquisition costs.

For any startup, it is critical to demonstrate steady growth with all the KPIs met, primarily including ROI. At Scentbird, I started my career as a media buyer and was responsible for user acquisition on Facebook. I managed 100% of performance traffic with up to a $2M monthly ad budget.

Facebook was the primary user acquisition channel for Scentbird, which brought the most valuable users from the ROI perspective.

My tasks were to:

• Optimize customer acquisition costs;
• Keep the pace of at least 3-4 A/B tests on each inventory;
• Scale the user acquisition meeting the KPI on ROI.

I determined and automated the routine processes to fulfill these tasks and maintain the growing channel.

The first areas of improvement were:

• Turn off ad sets that are not performing “good enough” right away after I knew it;
• Scale the “best” performing ad sets after the action above to maintain the daily budget.

To accomplish these tasks, I had to determine the following:

• Formulate the exact rules when the ad set is not performing “good enough” to turn it off;
• Find the earliest time when we can say so;
• Calculate how much daily budget was allocated to those ad sets (it was the case for the ad sets with budget allocation on the ad set, not the CBO);
• Scale the “best” performing ad sets after the action to maintain the daily budget.

The first version of automation I’ve completed using no-code tools. It was a simple set of if-else rules.

I formulated the following hypothesis:

• Automated hourly auditing will decrease FTE by at least 20% since we can only review new turned-off ad sets for the first short period after the automation launch and then shift from daily auditing to twice a week or even weekly auditing;
• Automatization would decrease customer acquisition costs by at least 15% due to timeless budget reallocation between ad sets.

As a result, after the implementation of this first automation, I had:

• Decreased FTE for Facebook routine tasks by 30%;
• Reduced customer acquisition cost by 16% (based on statistically significant A/B test results).

The measures mentioned above contributed to the growth of spending on Facebook while still meeting the KPI on ROI.

And this was only the first step. After a while, we implemented the automation of budget management, ad set relaunch, bid management, creative rotation, and other operations. As a result, we optimized the media buying team from 4 people managing a $500k monthly ad budget to 1 person managing up to a $2kk monthly ad budget.

This approach is still in place in Scentbird, and the same ideas are used in Playrix, one of the world's top mobile gaming companies by revenue and where user acquisition is performed on a considerable scale.