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

Programmatic Advertising › Troubleshooting

DV360 Line Item Not Spending

Why a display line item delivers little or nothing: filters that leave no requests, a bid below the clearing price, unapproved creative and an exhausted order.
Troubleshootingintermediate24 min readUpdated

First, read bid requests, bids, wins and impressions in order for the flight so far

The DV360 bidding and pacing guide describes the line item troubleshooting view as four figures for a date range, the bid requests the line item was eligible for, the bids it submitted, the auctions it won and the impressions it served, alongside any flags the platform raises about the line item itself, and its rule is to read them in order, because the first one that collapses relative to the one before it names the problem area. Set the range to the flight so far rather than to yesterday, since a line item that was starved on three days out of fourteen looks healthy on the fourth. Few requests point at targeting, inventory sources, brand-safety and viewability filters or a deal that is not sending. Requests but few bids point at status and dates, creative approval, an allowance already consumed or a frequency cap reached for most matched users. Bids but few wins point at the price. Wins but few impressions point at a creative rejected after the auction or a rendering failure.

The five whys method carries the first collapse back to a setting someone can change. The line item spent little because it submitted few bids; it bid on few requests because only one creative is approved; only one is approved because the other sizes were uploaded on the morning the flight began and are still in review; they were late because the plan's creative deadline sat after its start date; the deadline sat there because nobody had written down that review takes time. The fix lives at the last why, and the prevention one step further back, which is why the pre-flight at the end of this page is as much a part of the diagnosis as the table.

Two readings make the table quicker to use. First, compare the line item with a sibling under the same insertion order on similar inventory: a sibling with plentiful requests and a healthier win rate says the account, the exchanges and the creatives are fine and the difference lies in this line item's settings. Second, ask whether the line item ever spent. One that spent normally and stopped on a date points at a segment that ended, an allowance that a sibling started taking, a creative paused at advertiser level or a deal whose dates passed; one that never spent points at the build. If the symptom is a Google Ads display or video campaign rather than a DV360 line item, the diagnosis of a display or video campaign not spending reads the equivalent signals inside that platform, where the causes and the settings differ.

Diagnosis table

Diagnosis

SymptomCheckLikely causeFix
Few bid requests for the whole flight, far below a sibling on similar inventoryEach targeting dimension removed one at a time in the troubleshooting view, starting with viewability, content labels and the unlabelled-content settingFew bid requests: targeting, brand safety or viewability filters stacked until almost no inventory matchesKeep the filters the suitability policy requires and widen the rest one at a time, reading requests after each change
Requests and bids are plentiful, wins are few, and a sibling wins more at the same inventoryThe fixed bid against the sibling's win rate; the automated strategy's target or ceiling against what the inventory clears atBids but few wins: a fixed bid below clearing prices or an automated strategy constrained by its targetRaise the fixed bid in steps, or loosen the constraint as a decision about the target rather than a tweak
Requests arrive while status and dates are fine, but few bids followCreative status per exchange, and the sizes assigned against the sizes the inventory sendsRequests but no bids: no approved creative in a size the inventory sends, or a creative rejected by an exchangeFix the rejection reason, re-request review and assign every size the inventory sends
Spend front-loaded then stopped, or stopped on the first day of a monthThe insertion order's pacing view, each line item's budget share and the budget segment's datesThe insertion order's allowance consumed by a hungrier sibling, or a budget segment that endedGive each line item an explicit share, extend or add the segment, and separate deal line items from open-auction ones
A deal line item shows no requests, or requests with no bidsThe deal detail page for status, dates and exchange; the line item's inventory source and filters against the deal's inventoryA deal assigned to the line item that is not sending requests, or whose inventory the line item's filters excludeConfirm acceptance, dates and the exchange with the publisher; move the deal to a dedicated line item without the prospecting defaults
An audience line item sees few requests while a contextual sibling sees plentyThe list's size by environment, its membership duration and how long ago it was createdAn audience list too small to buy against, trimmed by consent, browsers and membership durationWiden with a combined audience or Google audiences, build lists before the campaign needs them, and plan reach for the gap
Few requests and rising prices on a line item governed by an inclusion channelThe channel's domains against past spend on them, and the reach estimate against the budgetAn inclusion list so short that it removes the reach the budget neededWiden the list from the next candidates in the register, or lower the budget the list governs
Wins are healthy against bids, but impressions served are far below winsCreative status per exchange for rejections after the auction, the creative preview and tag test in each assigned size, and the creative report's impressions by size and exchange against winsWins but few impressions: a creative rejected after the auction or a rendering failureFix the asset, tag or file that fails, re-request review where an exchange rejected the creative and re-test every size; do not raise the bid
Requests are plentiful and status, dates and creative approval are fine, but bids are few and fall away as the flight goes onThe frequency caps at campaign, insertion order and line item level, against the size of the audience the line item can reach and the average frequency so farA frequency cap reached for most matched usersWiden the audience, lengthen the cap's window or raise the cap as a decision about frequency, or lower the budget to what the capped reach can absorb

Fix, verify and prevent each cause

Filters stacked until almost no inventory matches

The real-time bidding guide describes how a demand-side platform processes a bid request in stages and discards as early as it can: eligibility first, which asks whether any active line item targets this inventory, geography, format and size, then frequency, pacing, brand safety and verification, and only then a valuation. Each targeting value, content label, sensitive category exclusion, viewability target, environment, authorised-seller option and pre-bid segment is another gate, and the DV360 inventory and deals guide is blunt about the arithmetic: a DSP only bids on requests that pass every one of these filters, and each added filter removes volume. The request itself carries the fields the filters read, which the real-time bidding guide lists from the OpenRTB specification published by IAB Tech Lab: the site or app, the device, the user, the sizes on offer and any deal. A line item whose filters leave almost nothing that passes is not broken; it has been told to want very little.

Fix: in the troubleshooting view, remove one targeting dimension at a time and read the requests figure after each, as the bidding guide's check describes, starting with the filters that remove the most, usually viewability targeting, the stricter content labels, the unlabelled-content setting, environment and the authorised-seller option; keep whatever the written suitability policy requires and widen the rest.

Verify: requests rise in the view the next day and the line item's figure moves towards the sibling's; then confirm that bids follow, because a filter removed to find volume can expose a creative or bid problem the filters were hiding.

Prevent: record the filter stack of every line item on the build sheet beside the reach the plan expects from it, so the person reading under-delivery can see which filters were chosen and which were inherited. Where a viewability filter is the one taking the volume, the page on low programmatic viewability explains what the filter is trading away and when to change the inventory instead.

A bid below the clearing price or a strategy constrained by its target

When requests and bids are plentiful and wins are few, the problem is price, and the bidding and pacing guide separates two versions. A fixed bid sits below what the inventory clears at; the guide's test is the win rate read beside plentiful requests, and the real-time bidding guide adds that in a first-price auction a floor sets the minimum regardless of competition, so a bid just under a common floor wins nothing. An automated strategy is a goal plus a constraint, and a strategy told to beat a target treats the target as fixed and lets spend float, which is why, as the guide puts it, a line item on that setting can under-deliver for weeks with no error shown anywhere: it is doing what it was told. The Display & Video 360 Help Centre is the documentation of the current strategies and their goals; this page describes the behaviour the site's guide states and names no option.

Fix: compare the win rate against a sibling on the same inventory and raise a fixed bid in steps rather than at once, or treat under-delivery on a target strategy as the strategy's verdict on the target and decide whether the plan promised volume or efficiency before loosening it; never change the strategy and the target in the same week, because learning restarts and nothing can be attributed.

Verify: wins rise relative to bids over the following days, and the CPM calculator confirms that the remaining budget still buys the impressions the plan needs at the price the line item now clears at.

Prevent: set the first fixed bid from what the inventory cleared at in its previous flight, write the target down as a decision with the volume it was expected to allow, and read win rate beside requests in the weekly review rather than impressions alone.

No approved creative in a size the inventory sends

The creatives guide states the two rules that turn a good line item into a silent one. A line item serves only the assigned creatives whose size and format match the bid request, and it does not bid where it has nothing to show, so a line item with a single size wins a fraction of the auctions a fuller set would. And every creative passes two reviews, Google's and each exchange's, with a status shown per exchange, so a creative approved by Google can still be pending or rejected on an exchange with stricter rules, and the line item only bids on an exchange where the creative is approved there. Both show in the troubleshooting view as requests without bids while status and dates look fine.

Fix: open the creative status per exchange, read the rejection reason and fix it, which the guide says is most often a missing or malformed click tag, assets loading from a disallowed source, a landing page that redirects unexpectedly or a video that fails the encoding requirements, then request a re-review; then assign every size the inventory sends, checked against the deal's or exchange's typical sizes rather than guessed.

Verify: bids rise against requests, the creative report shows impressions in each size and on each exchange, and no exchange reads zero.

Prevent: upload creatives days before the flight rather than on its first morning, keep rejection reasons in the trafficking log so the same mistake is not rebuilt, and make size coverage and per-exchange approval lines on the DV360 campaign launch checklist that someone ticks against the inventory the line item will actually see.

The insertion order's allowance consumed, or a segment that ended

The account structure guide places the budget on the insertion order and a share of it, or an unlimited share within it, on each line item, and the bidding guide explains what happens when the shares are left unlimited: an insertion order on even flight pacing throttles the combined spend of its line items to its own daily allowance regardless of what the line items ask for, so a deal line item can be starved by an open-auction sibling that reaches the shared allowance first. The same guide names the other quiet stop. An insertion order's budget is split into dated segments, and the segment's end date is also an end date for everything under it, which is the first thing to check when spend stops on the first of the month; the account structure guide adds that an insertion order whose segments end before its line items do stops spending without warning.

Fix: give every line item an explicit budget share rather than an unlimited one, move a deal into its own insertion order where it keeps being starved, and extend or add the segment so that the order's dates cover its line items' dates.

Verify: the pacing view shows each line item taking its share each day, and the starved line item spends from the next day without the sibling's spend falling below its own share.

Prevent: treat every insertion order as a line on the plan someone agreed to spend, record each line item's share on the build sheet, put segment end dates in the team calendar a week ahead, and let the launch checklist confirm budgets and dates against the signed plan.

A deal that is not sending requests, or whose inventory the filters exclude

The inventory and deals guide draws the distinction that explains most dead deal line items: acceptance makes a deal available, and only assignment to an active line item makes it spend. A deal whose dates have passed, or which the publisher has paused, stays visible in DV360 but sends no requests, and a deal identifier entered against the wrong exchange shows no requests and looks like a publisher problem. Where requests do arrive and bids do not, the guide's table points at the line item: not active, deal not assigned, targeting that excludes the deal's inventory, no approved creative in a size the deal sends, or allowance already exhausted. Its sharpest warning is about defaults, because every exclusion applies to deals as well as to the open auction, so prospecting defaults laid onto a deal line item quietly remove most of the publisher's inventory, and a guaranteed deal with a publisher whose pages are largely unlabelled, bought on a line item that excludes unlabelled content, delivers almost nothing.

Fix: read the deal detail page for status, dates and exchange and confirm them with the publisher; check that the deal is selected in the line item's inventory source targeting; then move it to a dedicated line item, as the account structure guide recommends for every deal, with the viewability, content label and audience layers set against the publisher's actual inventory rather than inherited.

Verify: the deal detail page shows requests, then bids, then wins across a full day, and the line item's price matches what the deal types guide says that kind of deal commits the publisher to.

Prevent: give each deal its own row on the build sheet with its identifier, type, dates and status, make the status check and the assignment part of the launch checklist's pre-flight, and agree sizes, formats and environment in the proposal so a mismatch cannot appear later as under-delivery.

An audience list too small to buy against

The audiences guide explains why an audience line item can be built correctly and still see few requests. A list's size is a modelled estimate of matchable identifiers rather than a count of people, it moves as membership durations expire members and as consent and browser changes alter what can be matched, and a list only begins collecting members when it is created, so a remarketing line item launched beside a new list has nobody to reach. The real-time bidding guide adds the shape of the problem from the request side: privacy rules, browser restrictions and publisher choices strip the user object of anything useful on a large share of requests, so audience-based line items see a fraction of the traffic that contextual ones do, and a plan that assumes every impression arrives with an identifier reaches fewer people than its forecast.

Fix: read the list's size by environment and the date it was created; widen a thin list with a combined audience that joins it to related first-party lists or Google audiences, or give the earlier stages of the plan a contextual line item that does not need an identifier at all, and keep the audience line item for the people it can actually match.

Verify: requests rise, the list shows a size after the delay the Help Centre documents for new lists, and the line item's requests sit at a sensible fraction of a contextual sibling's rather than near nothing.

Prevent: create Floodlight lists in the week the tags go live, weeks before the campaign that needs them, set membership durations deliberately, and write every audience size into the plan as an upper bound on the people the line item can reach, with the reach gap planned for rather than discovered.

An inclusion list so short that it removes the reach the budget needed

The tutorial on building a programmatic inclusion list calls the inclusion list the strongest brand safety control and the largest reach cut: every domain left off is inventory the buy can no longer reach, so reach falls and the price per thousand rises as the platform competes for a smaller pool, and its table names the risk of a short list plainly, under-delivery and high prices. The tutorial asks for the reach a list leaves to be estimated before launch, from past spend on the listed domains and the inventory forecasts the platform offers, and says what to do when the estimate falls short: widen the list from the next candidates in the ranking or lower the budget on the governed line items. A line item that under-delivers against a well-built list is usually one whose budget was set before anyone sized the list.

Fix: compare the flight budget with what the listed domains delivered in past placement reports, then widen the channel from the next candidates in the register, with their reasons recorded, or move part of the budget to deals with the publishers the list names, or lower the governed budget to what the list can absorb.

Verify: requests rise, the placement report shows spend spread across the added domains rather than concentrated on a handful, and the CPM calculator shows the budget buying the impressions the plan needs at the price the list clears at.

Prevent: write the reach estimate and the expected CPM into the plan before the list goes live, build the list once as a shared channel so copies cannot drift, and review it by spend on the rhythm the tutorial sets, adding candidates when the list cannot absorb the budget and retiring entries that take money without results.

A win that never becomes an impression

The fourth figure is the one the causes above do not reach. A win shows that the price was enough and that a creative was eligible when the bid was placed; an impression is counted when an ad is actually delivered, so a wide gap between the two means something failed after the auction. Two causes are usual: a creative rejected after the auction, which can happen on an exchange whose review is stricter than Google's, and a rendering failure, where the creative is accepted but does not load or play in the slot. The faults the creatives guide lists for rejection, a missing or malformed click tag, assets loading from a disallowed source or a video that fails the encoding requirements, are the usual suspects for both. Neither responds to a higher bid, because the bid has already won.

Fix: open the creative status per exchange and read for rejections that appeared only once the line item began winning; open each assigned creative's preview and test its tag or file in every size, including the mobile sizes; and where a third-party ad server serves the creative, compare its impression count with DV360's for the same dates, because a gap there can point at the tag, although the two systems also count differently, so compare like for like before blaming either. Correct what fails, request a re-review where an exchange rejected the creative, and re-test every size; the Display & Video 360 Help Centre's current guidance on creative status and rendering lists the statuses and reasons it shows, so check it for the exact wording.

Verify: impressions rise towards wins in the next day's view, the creative report shows impressions in every size and on each exchange, and the ratio of impressions to wins sits close to a healthy sibling's.

Prevent: test every creative in every size before launch, keep rejection reasons in the trafficking log, and put the impressions-to-wins comparison with a sibling into the weekly review beside the other three figures.

A frequency cap reached for most matched users

Frequency caps stack across campaign, insertion order and line item, and a person who has reached any one of them is no longer eligible for a bid. When most of the users the line item can match have reached a cap, requests keep arriving but the line item has little it may bid on, so the view reads as requests without bids while status, dates and creative approval look fine. The cap is doing its job rather than failing, and the cause tends to appear where a small audience list, a tight cap and a budget sized for a larger reach combine.

Fix: read the caps at each level, the audience's size and the average frequency so far, then decide what the plan wants: widen the audience so there are more people to spread the impressions across, lengthen the cap's window or raise the cap if the plan's frequency goal allows it, or lower the budget to what the capped reach can absorb.

Verify: bids rise against requests within a day or two of the change, and average frequency stays inside the plan's range rather than climbing to absorb the money.

Prevent: set each cap from the frequency the plan asks for rather than from a default, check the caps against the audience size and the budget in the pre-flight, and note the level each cap sits on, since caps at all three levels apply at once.

Prevention: a pre-flight that proves requests, creative approval and allowance before the first day

Every cause above would have been visible before the flight to someone who looked, and the DV360 campaign launch checklist is where the looking belongs: it confirms caps, pacing and bid strategy against the plan, compares site and app lists with the inclusion and exclusion plan, checks deal status and assignment, and checks budgets, dates and names against the signed plan. Three proofs make it a pre-flight rather than a tick list. A requests proof: on the first day, open the troubleshooting view and confirm that requests exist in the volume the targeting was expected to supply, using the inventory availability forecasts where the platform offers them for the line item's settings. A creative proof: every assigned creative approved on every exchange the line item buys from, in every size the inventory sends, checked days before launch rather than on the morning of it. An allowance proof: each line item's budget share written down, the segment dates covering the flight, and the deal line items in their own insertion order or with their own share.

The CPM calculator belongs in the same pre-flight. Turn the flight budget into the impressions it has to buy at the price the inventory cleared at last time, and compare that figure with what the targeting, the list or the deal can supply; the bidding guide uses the same calculation to turn a segment's budget into the impression volume a fixed bid implies and from there into whether the targeting can supply it. A line item whose required impressions exceed what its filters leave is under-delivering before it starts, and the plan should widen the targeting, raise the price it is willing to pay or lower the budget while there is still time to choose.

Then make the four figures a habit rather than a rescue. The programmatic campaign management playbook sets a daily pacing read with agreed limits on who may move budget, beside weekly inventory and frequency reviews and a monthly reconciliation; put requests, bids, wins and impressions per line item into that weekly review, with a sibling comparison for any line item behind pace, so a collapse is found in days. Frequency caps stack across campaign, insertion order and line item, and a cap reached for most matched users shows in the view as requests without bids; the page on a frequency cap that does not hold covers the opposite failure, where caps are set and the person sees the ad anyway. The DV360 hub places bidding, inventory, creatives and audiences side by side as the line item decisions this page has read in turn, and the programmatic troubleshooting hub lists the neighbouring symptoms, from low viewability to spend on sites nobody chose, that a line item can show once it is spending again.

Finally, keep the record that makes the next diagnosis a lookup. Every line item should have a row on the build sheet with its filter stack, its creative sizes, its budget share, its deal identifier where one applies and the reach the plan expected, and every change to a bid, a target, a filter or a share should carry a date and a reason, because the commonest finding when a line item stops is a change nobody connected to it.

A caravan dealer finds three throttles on one line item (illustrative)

Illustrative example — not a real result

Ebbingwold Caravans

Ebbingwold Caravans, an invented touring caravan dealer with 4 sales sites, runs a prospecting display line item on a £24,000 six-week flight. After 14 days it has spent £218, against the £8,000 that even pacing would have spent by then. The insertion order is on even flight pacing with a daily allowance of about £570 (£24,000 over 42 days), and a retargeting line item under the same order has an unlimited budget.

The troubleshooting view for the fortnight shows 31 million eligible bid requests, 1.4 million bids, 96,000 wins and 91,000 impressions. The line item uses a fixed bid of £2.40 per thousand impressions, so 91,000 impressions cost about £218, which is what it has spent. Requests are plentiful, so the collapse is between requests and bids. Two causes explain it. The only approved creative is a single leaderboard, while the inventory sends mostly medium rectangles and mobile banners, so the line item has nothing to bid with on most requests; and the retargeting sibling takes most of the order's allowance by mid-morning, after which the prospecting line item is throttled for the rest of the day.

The trader uploads four more sizes, which clear review on every exchange in 2 days, and gives the retargeting line item a 20 per cent budget share, which leaves the prospecting line item 80 per cent of the allowance, about £456 a day. Bids rise to 9.2 million over the next week, but wins reach only 390,000, a win rate of about 4 per cent against 11 per cent for the retargeting sibling on the same exchanges, and the week's spend is about £890 (roughly 370,000 impressions at £2.40 per thousand). The fixed bid sits below the clearing price; raised to £3.10 and then £3.60 over the next 4 days, the win rate reaches 10 per cent and daily spend settles at about £450, just under the line item's share.

The flight still ends well short. The line item has spent £218 in the first fortnight and about £890 in the next week, which is about £127 a day. Over the four days of bid changes, daily spend climbs from that £127 to the settled £450; taking the midpoint, about £290 a day, the four days add about £1,150, so it has spent about £2,260 in all by day 25. Seventeen days at about £450 add £7,650, so it finishes the flight at about £9,900 of the £24,000; even full delivery at the £570 allowance for all 28 remaining days would have reached only about £16,200. The dealer adds two lines to its pre-flight: every creative approved in every size the inventory sends, and every line item under a shared order given an explicit share.

Common mistakes

  • Reading impressions before reading bid requests. The impressions figure says the funnel broke; the four figures say where.
  • Raising the bid to fix every kind of under-delivery. Filtered requests, unapproved creatives, consumed allowances and timeouts do not respond to a higher price.
  • Reading under-delivery on a target strategy as a bug. The strategy is declining to pay above the target; the decision is whether the target was right.
  • One creative size per line item. The line item cannot bid on slots it has nothing to fill.
  • Unlimited line item budgets under one insertion order. The hungriest line item takes the allowance and the others starve.
  • Accepting a deal and assuming it runs. Acceptance makes it available; assignment to an active line item makes it spend.
  • Launching an audience line item beside a new list. The list starts empty and fills only from the day it exists.
  • Changing several settings in one day. Nothing that follows can be attributed, and a second fault stays hidden behind the first.

Frequently asked questions

The troubleshooting view shows plenty of bid requests and plenty of bids, but almost no spend. What does that tell me?

That the auction is being entered and lost, which makes it a price problem rather than a targeting or creative one. Compare the line item's win rate with a sibling buying similar inventory: a much lower rate on the same requests says a fixed bid sits below what the inventory clears at, or that an automated strategy is declining to pay above its target. Raise a fixed bid in steps and read wins again, or decide whether the target was right before loosening it. If wins are fine and impressions are not, look instead at a creative rejected by an exchange after the auction.

Should I just raise the bid to get the line item spending?

Only when the collapse sits between bids and wins. A line item with few requests is being filtered before any price matters, one with requests and no bids has nothing approved to show or no allowance left, and one with wins and no impressions has a creative or rendering problem; none of those responds to a higher bid, and the extra price is paid on every impression that does serve. Read the four figures first, fix the layer that collapsed, and treat the bid as the last lever rather than the first.

The line item spent normally for weeks and then stopped on a particular day. Where do I look first?

At whatever changed on that date rather than at the line item's own settings. The usual candidates are a budget segment on the insertion order that ended, a sibling line item with an unlimited budget that started taking the shared allowance, a creative paused at advertiser level, a deal whose dates passed or that the publisher paused, and an audience list whose members expired. The pacing view, the segment dates, the creative status and the deal detail page answer those in minutes, and a change log that records every edit with a date turns the question into a lookup.

Terms used on this page

Tools to try

Frameworks

Where this page is used

Sources

  1. Display & Video 360 Help — Google
  2. OpenRTB specification — IAB Tech Lab

Related