Don Draper would be overwhelmed by the sophistication of digital advertising. But he’d laugh at how unsophisticated financial reconciliation remains.

We can buy impressions in milliseconds and continuously optimize, yet, on the back end, we’re still exporting spreadsheets and manually reconciling invoices against contracts and delivery. The more sophisticated the front end gets, the less sustainable the reconciliation becomes.

This creates pain for accounting teams and economic friction for agencies. Manual work, meanwhile, erodes margins, and delayed billing means cash-flow pressure across the business (not to mention that it drives CFOs crazy).

The industry’s first wave of AI focused on optimization. Google optimized Google. Meta optimized Meta. DSPs optimized DSPs. The result was tremendous efficiency gains inside of individual systems. But AI did little to solve how those systems work together. In many cases, it accelerated fragmentation by making disconnected platforms operate faster.

Amid the growing media complexity, agencies ultimately need to know: What was contracted? Planned? Bought? Delivered? What did each platform report? What did the publisher invoice? And what should the client be billed?

Answers to those questions live in different systems, and financial reconciliation is where all those versions of reality finally must become one. When that data doesn’t move automatically between systems, people become the connective tissue by exporting files, compiling spreadsheets and manually investigating the almost-inevitable discrepancies.

Say an impression is contracted at $10 but comes through reporting at $11. Now the agency has to determine which number is correct, where the discrepancy originated and what the client and vendor should actually be charged.

1. Expensive human labor eats into profitability.

Manual data aggregation is costly on its own, but the more expensive part is the manual investigation to determine why two (or more) systems disagreed in the first place. Discrepancies are constant. When agencies need more employee hours to administer the same media spend, their service cost goes up and margins go down.

2. Slow reconciliation creates cash-flow problems.

Multifaceted campaigns can take weeks or months to reconcile, delaying billing and collection. If an agency owes a publisher before receiving payment from the advertiser, it effectively becomes a bank for its client, sometimes needing to draw on lines of credit and paying interest while waiting for cash to come in. This is the stuff of CFO nightmares.

3. Payment delays ripple through the media supply chain.

These payment delays propagate downstream to publishers and other media partners. In some large agency relationships, payment terms stretch as long as 120 or 150 days, creating the same wave of cash-flow pressures within the publisher world.

Large vendors have a traditional answer for fragmentation: Consolidate everything into one platform (aka ours). But this doesn’t fit agency reality today.

With technology advancing so quickly and completely new capabilities appearing at a rapid rate, agencies see best-of-breed strategy as the only way to remain competitive.

The answer, then, can’t be forcing agencies to choose between having the best tech for the job and having manageable back-end operations. The future won’t be won by the platform that replaces everything. Rather, it will be won by the platform that connects everything.

Leading agencies are building a new kind of “AI orchestration layer” to resolve fragmentation across their core operations, connecting data and workflows across planning, activation, measurement and actualization. This leaves agencies free to fully embrace a best-of-breed strategy.

Historically, those functions operated as discrete steps, with reconciliation sitting at the end. Orchestration creates a more continuous information flow. Instead of manually reconstructing a campaign at month-end, an orchestration layer can pull together the relevant data continuously, identify discrepancies earlier and coordinate the information needed to resolve them.

That makes financial reconciliation less of a standalone process and more the byproduct of a well-run system.

Tech vendors are supporting this agentic enterprise model through new partnerships. For example, the integration between Basis and Mediaocean Prisma brings together planning, activation and actualized omnichannel campaign data, then passes that clean data set into Prisma for financial reconciliation and invoicing. Changes in Prisma can flow back into Basis as well, helping both systems operate from a consistent source of truth.

Removing repetitive manual work reduces the labor required to administer the same media spend.

But an even bigger economic boon could come from better visibility. A continuously orchestrated approach gives agencies a much clearer view of accounts payable and receivable – what clients owe, what publishers are owed and where discrepancies remain – so they can understand how cash moves through the business.

There may even be a strategic upside. Agencies with clean accounts receivable and accounts payable and a track record of paying partners quickly can have more leverage to negotiate favorable payment and credit terms with vendors.

This story is part of a broader emerging trend. Some of the most valuable applications of AI in advertising may be around the administrative work that keeps campaigns moving by coordinating repetitive workflows, connecting fragmented data and, now, orchestrating financial reconciliation.

This isn’t flashy, but it is consequential. It’s about how healthier cash flow supports a healthier marketplace.

The first generation of AI helped marketers make better decisions. The next generation will help agencies run better businesses.

For more articles featuring Michael Olson, click here.