WHY DIRECT SOLD CAMPAIGNS BREAK BEFORE THEY LAUNCH
Trafficking sits between a signed IO and a live campaign, absorbing every unstructured agency instruction and manual re-entry along the way.
The risks are real — delayed launches, delivery gaps, makegoods — and most revenue leaders never see them coming.
Here's where they start.
Direct Sold Operations
The gap between signed and live
Direct sold campaigns don't break at delivery. They break in the hours and days between a countersigned IO and a campaign that actually goes live. This is where it happens, and why it keeps happening.
Campaign lifecycle — annotated
Phase 01
The IO is signed. Now you wait.
Somewhere on the agency side, someone is preparing trafficking instructions: a document, a spreadsheet, an email chain, or some combination of all three. There is no standard format, no SLA, and no visibility into when they will arrive. They land when they land.
Instructions arrive late, incomplete, or both
Your trafficker has no control over when agency instructions land. A campaign scheduled to go live Monday might receive its trafficking specs on Friday afternoon, or Monday morning. The publisher absorbs this uncertainty entirely, and the clock keeps running regardless.
No standard format means every IO requires interpretation
Agency A sends a PDF. Agency B sends a spreadsheet. Agency C sends an email thread. Each requires someone to read, interpret, and manually translate those instructions into your systems. Every translation is another opportunity for error to enter the campaign.
Phase 02
The clarification loop opens.
Instructions that are not human-readable, complete, or accurate get sent back to the agency for clarification. The agency responds when they respond. Your trafficker waits, follows up, and documents the thread. The go-live date does not move.
Clarification cycles eat directly into the launch window
Every email thread asking to confirm targeting parameters is a day closer to the campaign start date. Late launches compress delivery, creating pacing pressure and increasing the probability of underdelivery that will require a makegood.
Phase 03
The campaign gets built across every system.
For a campaign to be traceable across its lifecycle, it has to exist in every system that touches it. That means the same information gets entered by hand, repeatedly, in platforms that do not share a source of truth. Every pass is another opportunity for a number, a name, or a targeting parameter to drift from what was originally specified.
Manual re-entry across systems creates data inconsistency
A line item budget entered correctly in the OMS but transposed in the ad server will run wrong until someone catches it. The campaign optimizes to the wrong number, and by the time it surfaces in reporting, the damage is done.
Phase 04
The campaign goes live. Mostly correctly.
Pre-flight reviews catch what they can see. When the inputs feeding into them, including specs, targeting parameters, and creative assignments, were entered manually across disconnected systems, there is a limit to what any review process can surface. Errors that slipped through Phases 01 to 03 go live with the campaign.
Mis-trafficked campaigns do not announce themselves
A campaign running against incorrect targeting parameters may still be delivering. It can take days or weeks for the discrepancy to surface in reporting. The window between go-live and discovery is the window where the problem compounds.
Phase 05
Ad ops catches what trafficking missed.
An in-flight discrepancy surfaces. It might be a pacing issue, a creative serving on the wrong placement, or a targeting parameter that was transposed at setup. Ad ops investigates, finds the root cause, and fixes it. Now they own a delivery gap that should never have reached them.
In-flight fixes do not recover lost impressions
Catching an error in week two of a four-week campaign means the first two weeks ran wrong. The setup can be corrected. The impressions cannot be recovered. What comes next is a conversation about a makegood.
Phase 06
The debrief nobody wanted.
The client is owed impressions. Your team calculates what it will take to make them whole. The delivery happens, but the revenue was already recognized. You are delivering against margin you have already spent.
Makegoods are a trafficking problem on a delayed fuse
The makegood looks like a client relationship issue. It originates in a setup error that happened before the campaign ever went live. That is the connection that rarely gets made, and rarely gets fixed as a result.
80–90%
time savings on trafficking versus manual setup, across Swivel customers
Most of that time was waiting for instructions, correcting entries across systems, and chasing down discrepancies that should have been caught before go-live. Swivel automates the setup work that precedes every direct sold campaign, giving operations teams the speed and accuracy to execute at the pace the business actually demands.
What Swivel is building
No toggling. No typos.
Swivel automates campaign setup, enforces consistency across platforms, and surfaces errors before campaigns go live, so your operations team stops absorbing problems that should never have reached them.
Get in touch → More resources