Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Collection

Programmatic Troubleshooting

Programmatic problems arrive as symptoms: a line item that has spent a fraction of its flight budget, a Floodlight activity that reads zero after a release, a placement report full of domains nobody chose, a viewable rate well below what the plan promised, and a frequency cap that customers say is not working. Each page here starts from one of those symptoms, separates the causes in a diagnosis table and gives the check, fix, verification and prevention habit for each, linking the platform, measurement and brand safety guides that teach the method.

Looking for how to use this library? Read the guide below the list.

All 7 pages in this collection

How these pages work

Every diagnosis here begins with a reading, not a setting. A delivery problem starts with the funnel of bid requests, bids, wins and impressions, because the stage where the numbers collapse names the layer at fault. A measurement problem starts by deciding which count is missing or in excess: the tag's own firings in the ad server, or one buying platform's attributed share of them. An inventory problem starts with the placement report sorted by share of spend, and a frequency problem with the distribution of exposures rather than the average. Once the fault is visible, a diagnosis table lists what you can see, the check that confirms each cause and the fix. Each cause then gets a short section with the fix, a way to verify it and the habit that prevents it, followed by a routine that keeps the whole set under watch between flights.

The pages assume the methods are taught elsewhere and link to them. The hierarchy of advertiser, insertion order and line item, and where each setting lives, is in the account structure guide; bid strategies, pacing and caps are in the bidding and pacing guide; the conversion tag and its counting methods are in the Floodlight tags guide; and the controls that decide where an ad may appear are in the brand safety and viewability guide. The real-time bidding guide explains what a bid request carries, which is what every targeting filter reads.

Find the page for your symptom

Choose the row that matches what you are seeing. Diagnosis pages in this section are also listed beneath this introduction automatically; the last three rows send faults that belong somewhere else to the page that owns them.

What you are seeingStart withBackground
A line item has spent a fraction of its flight budget and shows no errorLine item not spendingBidding and pacing
A Floodlight activity reads zero, or the buying platform shows none of the conversions the ad server recordsFloodlight not recording conversionsFloodlight tags
A Floodlight activity reads higher than the orders, the conversions credited to media add up to more than it recorded, or display is credited with purchases that would have happened anywayFloodlight conversions counted twiceFloodlight tags
Most of the spend sits on a long tail of domains nobody choseProgrammatic spend going to low-quality sitesBuilding an inclusion list
The viewable rate is well below the figure the plan promisedProgrammatic viewability rate lowViewability and attention
The price per thousand is above what the plan priced and nothing was changed on purposeProgrammatic CPM higher than plannedDSPs, SSPs and data platforms
People say the ad follows them everywhere, although the cap looks rightProgrammatic frequency cap not workingReach and frequency across platforms
An ad has appeared beside content the brand would never chooseResponding to a brand safety incidentWriting a brand suitability policy
A Google Ads display or video campaign, not a DSP line item, will not spendDisplay or video campaign not spendingMoving display from Google Ads
A Tag Manager tag never fires, whatever it was meant to sendTag Manager tag not firingConsent mode

Delivery, measurement or inventory: decide which question you are asking

Nearly every symptom here answers one of three questions, and naming the question first saves most of the diagnosis. A delivery question asks why the money is not going out, or is going out in the wrong shape: filters that leave almost no requests, a bid below what the inventory clears at, creative no exchange has approved, an allowance consumed by a sibling line item. A measurement question asks why a number is missing or disputed: a tag that never fires or fires only when consent is granted, an activity the buying platform cannot read, a viewable rate calculated on a small measurable base, a cap that counts identifiers rather than people. An inventory question asks where the money went: an open auction bought with exclusions only, a goal that rewards cheap impressions, supply paths through resellers, audience targeting that follows people onto any page.

The test is to ask which system would have to change for the complaint to go away. If the funnel inside Google's DSP is short, it is delivery, and the fix lives on the line item or the insertion order. If the DSP delivered but the ad server, the verification vendor or analytics disagrees about what happened, it is measurement, and the programmatic measurement section explains what each system can honestly count. If delivery and measurement agree and the result is still wrong, the question is where the impressions ran, and the placement report is the evidence. The platform's Help Centre is the documentation for the settings themselves; these pages describe the behaviour the site's guides state and name no option, default or threshold, because those change and the documentation is the record.

Two symptoms belong to more than one question. A low Viewability rate can be a measurement artefact, a delivery choice or an inventory problem, and frequency is at once a setting, a measurement and a planning decision, so each page's first reading separates them. Faults inside another platform, such as a Google Ads campaign or a tag that never leaves the browser, belong on that platform's pages.

When to audit instead

A diagnosis page fixes one fault. When several of these symptoms arrive together, or when an account has been inherited and nobody can say why its settings were chosen, an audit is the better tool. The programmatic advertising audit scorecard rates a running buy from what Floodlight counts through safety, deals and frequency to reconciliation with invoices, and ranks the fixes that stop money buying nothing first; the programmatic campaign management playbook joins daily pacing, weekly reviews and monthly reconciliation into one routine that catches most of these faults before a client does; and the campaign launch checklist proves deals, safety settings, caps, creatives and Floodlight activities before the first day. The platform section and the programmatic hub place the guides these pages draw on side by side.