StackAdapt bidding and budgets

Match the bid approach to the objective, size the budget so the bid can learn, and pace it so the flight delivers in full without front-loading. StackAdapt is a self-serve multi-channel demand-side platform (native, display, video, connected TV, audio, in-app, and digital out-of-home). On StackAdapt the ad group is the line item, so bid, budget, and pacing are all set per ad group.

This skill owns the bid, budget, and pacing decision. For where these sit in the build order, see the stackadapt-campaign-setup skill. For the audience the bid competes for, see the stackadapt-targeting-and-audiences skill. For CPM, CPC, CPA, ROAS, and pacing-index math, see the programmatic-foundations skill. For reading whether the bid is hitting goal, the reporting-by-campaign-goal skill gives the by-objective report recipes.

When to use this skill

• "Which bid strategy should I use on StackAdapt?" / "Automated or manual bidding?"
• "Set a target CPA / target CPC." / "Should I let it maximize delivery?"
• "What maximum bid should I set?"
• "Set the budget and pacing for this ad group." / "Daily or lifetime budget?"
• "My campaign is under-pacing / spending too fast / blowing through budget."
• "How do I manage budget in flight?"

Boundary with sibling skills. This skill decides the bid, the budget number, and the pacing mode. It does not decide the audience (stackadapt-targeting-and-audiences) or diagnose deep delivery failures once setup is correct (stackadapt-optimization-and-troubleshooting).

Quick reference

Pick the bid type from the objective and the signal you have:

Objective	Bid type	Optimizes toward	Use when
Awareness / reach	CPM	Impressions delivered	Brand, reach, video, CTV, audio, DOOH
Traffic	CPC	Clicks to site	Drive site visits, lower-funnel native or display
Engagement	CPE	Post-click content engagement	Content, blog, and video where dwell matters
Conversions	Target CPA (automated)	Conversions at a cost target	Pixel events are live and accruing
Conversions / value	Goal-based automated (CPC or CPA goal)	The KPI you set, bid managed by the model	You want the model to chase a defined KPI
Spend the budget	Maximize delivery	Fullest delivery for the budget	Fixed budget, secondary CPA sensitivity

Two control modes sit on top of the bid type:

• Manual or fixed bid: you set the price, you keep hard control, the model does not move it.
• Automated or smart bidding: the model sets each bid in real time toward the goal and KPI you defined. It needs signal (conversions for CPA) and budget room to learn.

A maximum bid caps what any single auction bid can reach. Budget and pacing then govern how that spend is spread across the flight.

Core process

1. Read the objective off the campaign. Awareness, traffic, engagement, or conversions was set at the campaign level (see stackadapt-campaign-setup). The objective is what makes one bid type correct and the others wrong, so start there.
2. Pick the bid type from the table. CPM for awareness, CPC for traffic, CPE for engagement, target CPA for conversions. The campaign automatically optimizes toward the goal you choose, so the bid type is also a statement of what "good" means for this ad group.
3. Choose manual or automated control. Use a fixed or manual bid when you need hard price control, when volume is too low for a model to learn, or at launch before there is conversion signal. Use automated or smart bidding when the objective is conversions, the pixel is firing enough events, and you want the model to set each bid toward the target. Do not put a brand-new conversion ad group straight onto target CPA with no events; it has nothing to learn from.
4. Set the target only when automated. For target CPA, set a CPA the economics can bear and that recent performance suggests is reachable, not an aspirational number the model cannot hit. An unreachable target starves delivery because the model will not bid. If you have no history, start on a fixed bid, gather data, then switch.
5. Set a maximum bid. Cap the per-auction bid so a thin-supply moment cannot overpay. Set it high enough not to choke delivery but low enough to protect efficiency. Too low a max bid is a common cause of under-delivery.
6. Set the budget on the ad group. The ad group is the budget container, so the number lives here, not on the campaign. Use a daily budget for always-on, steady delivery; use a lifetime or total flight budget when the total is fixed and you want the system to spread it. Give automated bidding enough budget to clear the daily auctions it needs to learn; a starved budget keeps a smart bid stuck in learning.
7. Set pacing. Even pacing spreads spend across the flight and is the default for always-on and brand work. Accelerated pacing spends as fast as auctions allow, for short, time-boxed pushes only, and you must watch for early budget exhaustion. Match pacing to whether the goal is steady presence or fast volume.
8. Manage in flight. Check the pacing index (delivered vs expected for the flight to date) and cost vs goal on a regular cadence. The Suggestions Hub recommends optimizations based on the bid type and goal you set, including a concrete suggested bid value when it advises raising or lowering a bid, and can flag domains to exclude. Act on under- or over-delivery early, while there is still flight left to correct.

Decision rules and thresholds

• Bid type follows the objective, not preference. Awareness is CPM, traffic is CPC, engagement is CPE, conversions is target CPA. If the requested bid type contradicts the objective, fix the objective or the bid type before launch.
• Fixed bid at cold launch, automated once signal exists. A conversion ad group with no pixel events cannot run a productive target CPA. Start fixed, let events accrue, then move to automated. Switching too early traps the model in learning.
• Set a target CPA the data can reach. The model only bids when it can plausibly hit the target, so an unreachable CPA reads as under-delivery. Anchor the target to recent actual CPA, then tighten gradually.
• A maximum bid that is too low throttles delivery. If an ad group under-delivers, check the max bid before blaming targeting or budget; raise it if it is clipping winnable auctions.
• Budget lives on the ad group. Set and control spend per ad group. The campaign is not the budget container on StackAdapt.
• Give automated bidding room to learn. Too small a daily budget keeps a smart bid in learning and hurts performance. Fund enough daily auctions, or widen the audience, before concluding the bid strategy failed.
• Even pacing by default, accelerated only for bursts. Even pacing for always-on, brand, and steady delivery. Accelerated for short, time-sensitive pushes, and monitor for early exhaustion.
• One change at a time in flight, then let it settle. Automated bids need a learning window after a target or budget change. Stacking several edits at once makes the result unreadable; change one, give it time, then reassess.
• Read pacing and cost-to-goal together. On pace but over CPA, or under CPA but under-pacing, are different problems with different fixes. Check both before acting.

Templates and examples

Three ad groups in one campaign, each matched to its objective:

Ad group: ACME_CTV_Awareness_2026Q3
  Objective:  Awareness / reach
  Bid type:   CPM, fixed
  Max bid:    set to plan CPM ceiling
  Budget:     $20,000 lifetime over the flight (ad-group level)
  Pacing:     Even
  Why:        Reach goal, so price impressions on a fixed CPM and spread evenly
              across the flight for steady presence.

Ad group: ACME_Native_Traffic_2026Q3
  Objective:  Traffic
  Bid type:   CPC, automated once clicks stabilize
  Max bid:    cap above expected winning CPC, not at it
  Budget:     $300/day (ad-group level)
  Pacing:     Even
  Why:        Goal is site visits; start manual CPC, move to automated CPC goal
              once volume is steady so the model optimizes clicks.

Ad group: ACME_Display_Conversions_2026Q3
  Objective:  Conversions
  Bid type:   Fixed CPC for the first ~1-2 weeks, then Target CPA
  Target CPA: anchored to recent actual CPA, tightened gradually
  Max bid:    set so it does not clip winnable conversion auctions
  Budget:     $500/day, enough to clear daily auctions for learning
  Pacing:     Even
  Why:        Pixel events must accrue before target CPA can learn; launch fixed,
              switch to target CPA once events are flowing, fund learning.

In-flight adjustment example:

Symptom:  Conversion ad group is at 60% of expected pace mid-flight, CPA on goal.
Read:     Delivery problem, not an efficiency problem (CPA is fine).
Action 1: Raise the maximum bid; a low cap is the usual delivery throttle.
Action 2: If still short, widen audience (see targeting skill) or raise budget so
          the smart bid has more auctions to learn from.
Then:     Change one lever, let the learning window pass, recheck pace and CPA.

Common pitfalls

• Putting a brand-new conversion ad group straight on target CPA with no pixel events, so it has nothing to learn from and barely delivers. Launch on a fixed bid first.
• Setting an aspirational target CPA the data cannot reach, then reading the resulting under-delivery as a targeting problem. Anchor the target to actual recent CPA.
• A maximum bid set too low, quietly throttling delivery. Check the cap first when an ad group under-delivers.
• Starving automated bidding of budget, trapping it in learning, then blaming the strategy. Fund enough daily auctions.
• Setting budget or pacing on the campaign and expecting it to control the ad group. Budget and pacing live on the ad group on StackAdapt.
• Leaving accelerated pacing on for an always-on flight, then exhausting the budget early. Use even pacing unless it is a deliberate short burst.
• Stacking several in-flight changes at once so the learning signal is unreadable. One change, let it settle, then reassess. For persistent delivery failures after setup is correct, hand off to stackadapt-optimization-and-troubleshooting.

Sources

• How to Run a Successful Native Advertising Campaign | StackAdapt (as of June 2026)
• Optimize Your Campaigns With Suggestions Hub | StackAdapt (as of June 2026)
• StackAdapt Developer Documentation (REST, GraphQL, Pixel API, Data Taxonomy, MCP Server) (as of June 2026)

The exact in-product bid-strategy menu (the precise labels for automated and goal-based bidding), the maximum-bid field, daily versus lifetime budget controls, and the even versus accelerated pacing toggle are documented in the StackAdapt help center. Some help center articles require a logged-in account, so they are not cited here; where a specific in-product control, label, or default is not publicly documented, the guidance above states the operational best practice rather than a fabricated citation. For account-level help, see the StackAdapt help center at support.stackadapt.com.