Karolina BendrykThe adtech market is no longer debating whether automation in ad operations is necessary — it is actively adopting it. According to research by Theorem in partnership with Hypothesis Group published by Digiday:
How ad ops teams estimate the value of automation
This raises a new question: if market consensus is so strong, why does operational drag persist? Manual work continues to take a significant share of time in many teams, forcing them to wrestle with spreadsheets, chase approvals, and handle reconciliation.
The reason extends beyond the tech stack or the skill of the teams. It comes down to a misplaced focus: companies automate isolated tasks while leaving their broader operational architecture entirely manual.
Inside 79%: What “tools could be better” really means
Manual work is the direct consequence of AdOps process complexity. Most of these processes lack standardization, which further complicates the workflow.
Partner onboarding is one such area. AdMonsters, quoting Addy Atienza, Managing Director, Global Programmatic and Ad Operations at Time Out Group, describes it plainly: “There’s no template. Everything’s manual.”
Without standardization, AdOps professionals have to navigate every onboarding case independently — an approach Addy calls "just figure it out" energy. As a result, specialists spend the bulk of their time on manual setup drag — communicating with colleagues, gathering technical requirements, resolving mismatches — rather than testing creatives, fine-tuning floor price rules, and validating ad latency.
According to AdMonsters, SiriusXM faced the same bottleneck before shifting to structural workflow automation:
Before automation: The AdOps team was buried in spreadsheets. Most of their time went toward mechanically collecting, exporting, and manually verifying data across disparate systems. Strategic analysis was deferred, and experienced specialists were bogged down with low-leverage tasks beneath their skill level.
After automation: Automating routine workflows freed up bandwidth to focus on partner relationship management, generate revenue insights, analyze campaign performance, and identify new growth opportunities. The AdOps function shifted from mechanical data processing to directly driving business decisions, enabling one specialist to transition into a fully strategic role.
Reconciliation is not just a reporting problem.
Data reconciliation is one of the most complex tasks in ad ops manual processes. Yet it is often automated most simply: by adding another analytics dashboard.
The issue is that dashboards provide visibility, but do not necessarily support the decision and follow-up workflow. They can highlight a discrepancy, but cannot process the underlying operational context:
They don't know the root cause of the discrepancy or whether it falls within an acceptable tolerance threshold;
They ignore data source conflicts across platforms that calculate metrics using different logic;
They cannot evaluate the financial risk to the final invoice;
They cannot trigger escalation or approval workflows.
AdOps teams still have to connect these dots manually. Consequently, dashboards don't eliminate ad ops manual reporting — they merely serve as an intermediary step before it.
AdExchanger highlighted the complexity of reconciliation back in 2016. The fact that it remains a pain point a decade later underscores the persistent, structural nature of the problem.
The work everyone automates is not the work taking the most time
Businesses often focus AdOps automation on campaign optimization. However, adjusting bids or targeting is only a fraction of the daily AdOps workload.
According to data from The Drum that was published by Influencers Time, AdOps teams in agency and in-house media spend their working hours as follows:
Approvals & Reallocation: ~30%
Reporting & Reconciliation: ~25%
Creative Versioning & Trafficking: ~22%
Targeting & Optimization: under 15%
This means investing exclusively in optimization tools shouldn't be the top priority for automation. Instead, priorities should be driven by the business model, overall strategy, and specific workflows. Moreover, even the most powerful optimization engine remains a point solution. It doesn't relieve teams from the regular necessity to:
secure budget sign-off from the client;
validate whether creative assets meet technical specs;
align changes and hand off context to cross-functional teams;
reconcile reports and pinpoint the root cause of numerical discrepancies;
confirm that the final execution was carried out.
The solution lies in rewiring how systems, data, and teams interact. Effective workflow automation must tie every operational step into a single pipeline that connects:
Signal: A data event or state change that indicates a need for action.
Business rule: The logic that defines what specific action should be executed for a given signal.
Owner: The designated person or system responsible for execution.
Approval: Securing the required sign-off (when the action exceeds auto-approval thresholds).
Action: The execution of the operational task (launching, adjusting, or trafficking).
Follow-Up: Verifying the outcome and closing the operational loop.
It is a comprehensive process that requires structural change at the workflow level, rather than just within isolated tasks.
Adding ad ops automation tools does not always solve the fragmented workflow
Point solutions rarely fix root-cause issues; they push the ad ops operational bottlenecks further down the chain.
The reconciliation example above illustrates this clearly. :
Without a dashboard: AdOps toggles between five platforms, manually pulling CSVs into Excel to spot mismatches and calculate the reason behind them.
With a dashboard: The system automatically aggregates data and flags a 15% anomaly. Yet, AdOps must still export data and investigate to solve the issue.
The company automated only a single step: signal detection. The surrounding execution chain remains fragmented, complex, and manual.
Effective operational automation bridges context, routing, and execution. It doesn't always mean replacing manual work — and that is not the point. The goal is to structure how work moves forward by encoding:
How the business defines and categorizes the issue;
Which error margins trigger critical alerts;
Who owns the remediation step;
What exceptions apply;
What ad ops reporting automation is needed and which automated action triggers once a signal is detected.
The ability to orchestrate the complete lifecycle of an issue — rather than just generating a report, alert, or recommendation — is what separates an automation layer from limited enhancements.
Effective automation starts with the workflow
When specialists say their tools "could be better," this does not always mean they need shinier software. But it is definitely a sign businesses should evaluate their existing workflow to uncover where context is lost, ownership is blurred, or next actions stall.
Operational drag doesn't always happen during technical execution. More often, it occurs at the friction points between steps where context drops, systems fail to communicate, and actions depend on manual handoffs with no system-level control. This is where workflow automation actually adds value: not by stacking more point solutions, but by connecting systems, rules, roles, and actions directly at the workflow level.
