Dv360 bid strategy
Skill scumunna/programmatic-skills/skills/dv360-bid-strategy
Agent skills for programmatic trading, analytics, and account operations. DV360 first, multi-DSP and multi-runtime (Claude Code and Codex).
npx -y skills add scumunna/programmatic-skills --skill dv360-bid-strategyAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Choose and configure the right Display & Video 360 bid strategy. Use when the user asks which bid strategy to use, fixed vs automated bidding, Target CPA, Target CPM, Target ROAS, maximize conversions or value, minimize CPA or CPC, target viewable CPM, custom bidding, learning periods, bid caps, or why a line item is not winning auctions.
SKILL.md
9.8 KB, ~2.2k tokens by cl100k_base, as published. Nobody here has run it
DV360 bid strategy
Pick the bid strategy for a line item and set it so it actually delivers: fixed when you need control, automated when you have signal and a clear goal, custom when standard goals cannot express the value you care about. The bid strategy is set on the line item and is the lever that turns budget into won impressions against your KPI.
This skill assumes you know CPM, CPA, CPC, ROAS, and viewability. For those definitions and
the KPI math, see the programmatic-foundations skill. For where the line item sits in the
account, see dv360-campaign-architecture.
When to use this skill
- "Which bid strategy should I use?" / "Fixed or automated bidding?"
- "Set up Target CPA / Target CPM / Target ROAS / maximize conversions / maximize value."
- "Should I use minimize CPA, target viewable CPM, optimized reach?"
- "What is custom bidding and when do I need it?"
- "How long is the learning period?" / "Why did performance drop after I changed the bid?"
- "Why is my line item not winning / not spending?"
Boundaries with sibling skills:
- Where to put the line item and how many to create:
dv360-campaign-architecture. - Budget pacing and in-flight optimization:
dv360-pacing-and-optimization. - Building a custom bidding algorithm (rules or script) in depth:
dv360-custom-bidding. - Deal and inventory setup that fixes the price:
dv360-deals-and-inventory. - Full delivery troubleshooting tree:
dv360-troubleshooting.
Quick reference
| Situation | Use | Why |
|---|---|---|
| Little or no conversion data | Fixed bid | Automation has nothing to learn from |
| Reserved / programmatic guaranteed fixed-rate deal | Fixed bid | Price is already fixed by the deal |
| Tight, absolute CPM ceiling you cannot exceed | Fixed bid | Targets are averages, not hard caps |
| Small scale or pure reach/brand goal | Fixed bid | Not enough signal for goal-based automation |
| Want max conversions/value and have signal | Maximize (spend full budget) | Algorithm optimizes per impression |
| Want efficiency at a target | Target CPA / CPC / viewable CPM / ROAS | Optimizes toward the named goal |
| Value differs by impression beyond standard goals | Custom bidding | Score impressions on your own logic |
Two families of automated bidding:
- Maximize a KPI while spending the full budget (volume-first).
- Maximize a KPI while prioritizing a target (efficiency-first), for example a target CPA or target viewable CPM that the algorithm tries to meet or beat.
Core process
- Check conversion data. If the line item (or comparable history) lacks meaningful conversion volume, start fixed or with a non-conversion automated goal (viewable CPM, completed views). Conversion-based automation (tCPA, tROAS, maximize conversions/value) needs conversions to learn from.
- Name the goal. Volume (more conversions/value), efficiency (hit a cost target), or value (ROAS). The goal picks the family: maximize-spend for volume, performance-goal for efficiency, value-based for ROAS.
- Check inventory type. Programmatic guaranteed and reserved fixed-rate deals set the price, so use fixed bidding there. Biddable open auction and non-guaranteed PMP can take automated bidding.
- Set the strategy on the line item. Automated bidding can be set at the line item, or set at the insertion order to apply to all its line items. Add a bid cap only if you must, knowing it can throttle the algorithm.
- Pair with even pacing. Automated strategies optimize across the flight, so even pacing lets them learn and spread delivery. Do not use ASAP with automated bidding (see Decision rules).
- Hold the strategy stable through the learning period (below). If it is starving or missing goal, diagnose with the pitfalls table before changing anything.
Decision rules and thresholds
When fixed bid wins
- Limited conversion data, so automation has nothing to optimize against.
- Reserved or programmatic guaranteed fixed-rate deals, where the price is set by the deal.
- You need tight CPM control. Note that automated targets are averages the system tries to hold, not absolute ceilings, so when you must stay under a hard CPM at all times, fix the bid.
- Small scale or brand/reach goals where there is not enough signal for goal-based automation.
DV360 applies optimized fixed bidding by default to avoid overpaying on a fixed-bid line item; you can opt out if you need the exact bid every time.
When automated bidding wins
Automated bidding calculates a per-impression bid to make the most of the budget. Pick the goal that matches intent:
- Maximize (spend full budget): get the most conversions, value, completed in-view and audible impressions, or viewable impressions for the budget. Volume-first.
- Goal-constrained: maximize clicks or conversions while prioritizing a target KPI, target viewable CPM, target CPA, or Target ROAS. Efficiency-first.
The full list of strategies and the goal each one optimizes is in
references/bid-strategy-map.md.
Custom bidding
When the value of an impression is not captured by a standard goal, use custom bidding: it
scores each impression by importance to your objective and bids accordingly. Build the
algorithm with rules (weighted conversions, no code) or a script (optimize toward
non-conversion signals such as brand lift). For building and validating the algorithm, hand
off to dv360-custom-bidding.
Custom bidding needs enough scored-impression volume to train. As a rule of thumb DV360 looks for meaningful positive-signal volume per advertiser and per line item before it can calibrate, so do not point custom bidding at a brand-new, low-volume line item.
Learning period
Automated and custom strategies need conversion volume and a stabilization window before their performance should be judged. The bidding algorithm can take up to four weeks to learn and calibrate. Value-based strategies (Target ROAS, maximize conversion value) additionally want a minimum conversion history before they qualify, and a no-change window after you switch. Practical rule: change the strategy or its target as little as possible during learning, because edits reset calibration and waste the window.
Interaction with pacing
Pair automated bidding with even pacing so the algorithm can optimize across the whole flight.
ASAP pacing front-loads spend and can overspend early, which fights the optimization and burns
budget before the algorithm has learned. See dv360-pacing-and-optimization for pacing setup.
Reference material
references/bid-strategy-map.md: every fixed, maximize, goal-constrained, value-based, and custom strategy mapped to its goal, the line-item types it applies to, and the matching DV360 API v4 field. Read this when picking a specific strategy or setting it through the API/SDF.
Templates and examples
- Lower-funnel conversion line item, healthy conversion history: Target CPA at the line item, even pacing, no bid cap. Hold it for the full learning window before reading performance.
- Upper-funnel CTV awareness line item, no conversions: maximize completed in-view and audible impressions, or target viewable CPM, even pacing. Conversion goals would have nothing to learn.
- Premium programmatic guaranteed CTV deal at a fixed rate: fixed bid at the deal rate. The price is set, so automation adds nothing.
- Retail line item where a sale's value varies widely: custom bidding scoring impressions by
predicted value, built in
dv360-custom-bidding, then assigned here.
Common pitfalls
- Target set too tight. A tCPA, tROAS, or target viewable CPM below what the inventory clears at starves delivery, because the algorithm cannot find impressions at that price. Loosen the target or widen targeting, then let it relearn.
- Too few conversions for tCPA or tROAS. Without conversion volume the algorithm cannot learn; switch to a maximize or non-conversion goal until volume builds, or consolidate line items.
- Bid cap throttling automated bidding. A cap that is too low blocks the algorithm from bidding to win valuable impressions. Remove or raise the cap if delivery or performance suffers.
- Changing strategy mid-flight. Switching strategy or moving the target resets the learning period and discards calibration. Decide up front and hold it.
- ASAP pacing with automated bidding. Front-loads spend and overspends before the algorithm has learned. Use even pacing.
- "Line item not winning." Before touching the strategy, check the usual causes: bid or target
below the clearing price, targeting too narrow, creative not approved, budget or pacing
capping delivery, or deal eligibility. The full tree is in
dv360-troubleshooting.
Sources
- Automated bid strategies (as of June 2026)
- Value based bidding strategies (as of June 2026)
- Custom bidding overview (as of June 2026)
- Set a fixed CPM bid for a line item (as of June 2026)
- Troubleshoot your deals and line items (as of June 2026)
- BiddingStrategy, DV360 API v4 reference (as of June 2026)
- advertisers.lineItems, DV360 API v4 reference (as of June 2026)
Gives 0 of the 12 instructions most roadmap strategy skills give in ~2.2k tokens
Counted across 591 of the 672 authors here whose files we hold, read 2026-08-07
- read product marketing context before asking questionsin 21 of 591, across 10 files
- base price on perceived value, not costin 15 of 591, across 4 files
- compact after finalizing a planin 14 of 591, across 9 files
- differentiate tiers using features, limits, or supportin 14 of 591, across 3 files
- use Van Westendorp to find acceptable price rangein 13 of 591, across 2 files
- use MaxDiff to identify highly valued featuresin 13 of 591, across 2 files
- map topics to buyer journey stagesin 12 of 591, across 6 files
- Extract domain capabilities and classify subdomainsin 11 of 591, across 1 file
- Define bounded contexts around consistency and ownershipin 11 of 591, across 1 file
- Establish a ubiquitous language glossary and anti-termsin 11 of 591, across 1 file
- Capture context boundaries in ADRs before implementationin 11 of 591, across 1 file
- Open the strategic design template if neededin 11 of 591, across 1 file
Said here and by no other author read
- use fixed bidding without conversion data
- match bidding strategy to goal intent
- set automated bidding at line item or insertion order
- pair automated bidding with even pacing
- hold strategy stable through learning period
- use custom bidding for non-standard impression value
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.