Define the decision
Write the objective, accepted outcome and maximum learning loss for ai for marketing.
Business growth, website promotion and AI-powered marketing operations
AI for Marketing: Build a Clear, Measurable Operating Plan
Apply AI for marketing through a portfolio of bounded use cases, governed data, human decision rights, measurable workflow value and ongoing risk review.
ai for marketingai for digital marketing
What does this page explain about AI for Marketing: Apply It to Measurable Paid Growth?
Quick answer: Apply AI for marketing through a portfolio of bounded use cases, governed data, human decision rights, measurable workflow value and ongoing risk review. AI for Marketing is the practical use of AI capabilities to support digital marketing research, creation, activation, optimization and analysis. For ai for marketing, the practical job is to create a phased AI adoption roadmap that starts with valuable, reviewable and reversible workflows. The assigned keyword wording is ai for marketing and ai for digital marketing; those phrases must resolve to one canonical decision boundary rather than competing calculations.
| Section | Distinct excerpt from this page |
|---|---|
| AI in digital marketing and artificial intelligence workflows | The operating model must identify approved data, decision rights, evidence standards, review ownership, customer impact and the mature business result that receives credit. |
Reference for AI for Marketing: Apply It to Measurable Paid Growth: NIST: AI Risk Management Framework.
Key takeaways for AI for Marketing
- Define the accepted business outcome before evaluating ai for marketing.
- Compare business objective and approved use case, input data, permissions and provenance, and model or tool role and human review under the same measurement contract. For AI For Marketing, validate this point against Measurable Paid Growth, Marketing, Paid Growth and keep it separate from the the nearest related FroggyAds topic intent.
- For AI for Marketing, preserve source, placement, audience, creative and change-level evidence in exportable records.
- Use review acceptance rate, factual correction rate, and production time saved as diagnostics, then reconcile accepted value.
- For AI for Marketing, scale only when marginal quality and economics remain inside the approved decision boundary.
What AI for Marketing means in practice
AI for Marketing is the practical use of AI capabilities to support digital marketing research, creation, activation, optimization and analysis. The useful operating definition is narrower than a dictionary label: it states what decision the activity supports, which inputs are allowed, how eligibility is determined and what evidence is required before the result receives credit.
For ai for marketing, the practical job is to create a phased AI adoption roadmap that starts with valuable, reviewable and reversible workflows. That means separating the media action from the business outcome. Delivery, reach, impressions and clicks describe activity; accepted leads, completed purchases, retained customers or another approved business state describe value.
A strong ai for marketing plan begins with a boundary document. Record the accountable owner, target audience or context, approved markets, permitted data, chosen formats, conversion definition, attribution window, maximum learning loss and rollback trigger. The document prevents a platform default from silently becoming the strategy.
Why AI for Marketing matters
The main value of ai for marketing is decision clarity. Teams can compare options only when the comparison uses the same objective, time window, maturity rule and economic definition. Without that contract, a lower reported cost may simply reflect a different event, weaker quality or incomplete conversion maturity.
The strongest plans connect business objective and approved use case, input data, permissions and provenance, and model or tool role and human review with brand, factual and policy controls, workflow integration and fallback, and measurement, risk review and improvement. These elements interact. A useful audience can fail with the wrong creative, a strong format can fail on unsuitable placements, and an apparently efficient campaign can fail after rejected outcomes and reversals are included. A practical ai for marketing brief can operationalize this step with creative variation assistant, while treating mistaking fluent output for verified accuracy as an explicit pre-launch risk.
Use ai for marketing as a controlled learning system. The first launch should be narrow enough to explain, the change log should preserve every material decision, and the reporting should show both the platform result and the accepted business result. Scale is earned by repeated evidence, not by one favorable dashboard interval.
AI for Marketing operating architecture
Build the ai for marketing architecture in layers. Start with the commercial objective and accepted outcome, then define the audience or context, select the format and placement, prepare the offer and landing path, set budget and bid controls, and finish with measurement, exclusions and stop rules. Each layer needs an owner and a validation step.
Use stable names for campaigns, audiences, creatives, placements and test versions. Stable identifiers allow exports from the buying platform, analytics and business systems to be joined later. They also make it possible to distinguish a real improvement from a naming change, copied campaign or altered attribution setting. The ai for marketing review should therefore connect input data, permissions and provenance with policy or brand exception rate, a named owner and a dated change record.
Separate exploration from exploitation. Exploration tests new brief-to-draft workflow, creative variation assistant, and audience-insight synthesis under capped budgets. Exploitation allocates more delivery to combinations that have passed quality and economic checks. Combining both modes in one undifferentiated campaign hides where the learning budget went. For ai for marketing, apply the principle through a bounded test such as brief-to-draft workflow, and require factual correction rate to support the next budget decision.
Connect the guide to live testing
Connect AI for Marketing to a controlled audience test
Use the choices established in “AI for Marketing operating architecture” to define one audience, budget and source set in FroggyAds. Keep the surrounding offer and measurement rule stable so the test adds evidence to ai for marketing instead of mixing several changes at once.
Create My Free Account
AI for Marketing decision scorecard
For AI for Marketing, credit a decision layer only after it has a named owner, an operating control and exportable evidence.
| Decision layer | Operating requirement | Evidence required |
|---|---|---|
| Business Objective And Approved Use Case | Define the decision, input, control and exception path for business objective and approved use case. | Written definition, owner and approval boundary. |
| Input Data, Permissions And Provenance | Define the decision, input, control and exception path for input data, permissions and provenance. | Exportable setup, exclusions and change log. |
| Model Or Tool Role And Human Review | Define the decision, input, control and exception path for model or tool role and human review. | Creative and landing continuity evidence. |
| Brand, Factual And Policy Controls | Define the decision, input, control and exception path for brand, factual and policy controls. | Source or cohort reporting with quality review. |
| Workflow Integration And Fallback | Define the decision, input, control and exception path for workflow integration and fallback. | Reconciled analytics and business outcomes. |
| Measurement, Risk Review And Improvement | Define the decision, input, control and exception path for measurement, risk review and improvement. | Marginal scale result with rollback readiness. |
Special considerations for AI for Marketing
Delivery quality for ai for marketing depends on how the platform identifies users, placements, creative states and measurable events. Record these technical boundaries before interpreting the result. Identity approximation, unavailable signals and unmeasurable inventory should remain visible in reporting.
Evaluate distribution, not only averages. Break results into exposure bands, placements, devices, creative variants, audience stages and time. The distribution often reveals saturation, low-viewability inventory, broken dynamic combinations or a small cohort carrying the entire blended result. For ai for marketing, apply the principle through a bounded test such as landing-page content support, and require policy or brand exception rate to support the next budget decision.
Use automation within guardrails. Approved inputs, fallback creative, caps, exclusions, source review and rollback protect the campaign when a model or delivery system behaves differently from the forecast. Automation should expand controlled decisions, not remove accountability. The ai for marketing review should therefore connect brand, factual and policy controls with factual correction rate, a named owner and a dated change record.
Seven-step implementation workflow
Map eligibility
For AI for Marketing, document the audience, context, placement, GEO, device or prior behavior that makes delivery eligible.
Prepare the experience
For AI for Marketing, build format-specific assets, proof, call to action and a landing path that continues the same promise.
Validate measurement
For AI for Marketing, test delivery, analytics, conversion, acceptance, deduplication and delayed states end to end before campaign decisions depend on reporting.
Launch a bounded test
For AI for Marketing, set explicit test budgets, bid ranges, exclusions, frequency limits and dated review checkpoints before delivery begins.
Diagnose by cohort
For AI for Marketing, compare source, placement, audience, device, creative and exposure-level quality before keep, cap, exclude or retest decisions.
Scale or rollback
For AI for Marketing, expand one controlled dimension when marginal economics pass; otherwise return to the last stable configuration.
Creative, offer and landing continuity
Creative for ai for marketing should make one credible promise to one recognizable audience state. The headline or opening frame identifies the problem or opportunity, the supporting element supplies proof, and the call to action describes the next step. Avoid claims that the landing page cannot substantiate.
Prepare variations around meaningful hypotheses rather than cosmetic changes. Test a different proof point, customer problem, product benefit, objection, offer structure or format adaptation. Preserve enough consistency that the team can identify which idea changed response quality. For ai for marketing, apply the principle through a bounded test such as landing-page content support, and require policy or brand exception rate to support the next budget decision.
Landing continuity is part of the creative system. The destination should repeat the same terminology, offer and expectation introduced in the ad. If ai for marketing produces clicks but the landing page changes the promise, hides the action or loads poorly on the target device, the campaign is not ready for scale.
Choose the execution format
Choose a paid-media format that supports AI for Marketing
Use the criteria around “Creative, offer and landing continuity” to decide whether push, native, display or pop fits the message and destination. Set format, targeting and spend as campaign controls in FroggyAds while the ai for marketing decision remains the standard for judging the result.
Create My Free Account
Measurement contract and reconciliation
Measure ai for marketing through a chain rather than a single rate: eligible delivery, measurable exposure, qualified interaction, landing completion, primary conversion, accepted outcome and realized value. The chain reveals where volume becomes unusable and prevents a strong top-line metric from masking downstream weakness.
The core reporting set includes review acceptance rate, factual correction rate, production time saved, cost per approved asset, conversion or engagement lift, and policy or brand exception rate. Define each metric's numerator, denominator, data source, time zone, currency, attribution rule and maturity window. Where a platform metric cannot be reproduced from exportable evidence, label the limitation instead of presenting false precision. In a ai for marketing workflow, this control is most valuable when using confidential or unlicensed inputs could otherwise make the reported result look stronger than the accepted business outcome.
Reconcile platform, analytics and business records on a regular schedule. Differences are expected because systems use different identity, attribution and validation rules. Unexplained differences should block aggressive scale until the team knows whether the variance comes from tracking, delayed events, duplicates, rejected outcomes or reversals. The ai for marketing review should therefore connect input data, permissions and provenance with policy or brand exception rate, a named owner and a dated change record.
Metrics, definitions and diagnostic risks
For AI for Marketing, define every decision metric with a numerator, denominator, source, reporting window, currency, attribution rule and maturity condition.
| Metric | Definition requirement | Diagnostic check |
|---|---|---|
| Review Acceptance Rate | State numerator, denominator, source, time window, currency and maturity rule. | Check for adopting AI without a defined decision or workflow before the metric receives decision credit. |
| Factual Correction Rate | State numerator, denominator, source, time window, currency and maturity rule. | Check for using confidential or unlicensed inputs before the metric receives decision credit. |
| Production Time Saved | State numerator, denominator, source, time window, currency and maturity rule. | Check for publishing unreviewed claims before the metric receives decision credit. |
| Cost Per Approved Asset | State numerator, denominator, source, time window, currency and maturity rule. | Check for mistaking fluent output for verified accuracy before the metric receives decision credit. |
| Conversion Or Engagement Lift | State numerator, denominator, source, time window, currency and maturity rule. | Check for automating bias or weak brand patterns before the metric receives decision credit. |
| Policy Or Brand Exception Rate | State numerator, denominator, source, time window, currency and maturity rule. | Check for scaling low-value content or creative volume before the metric receives decision credit. |
Budget, economics and break-even control
Set the economic boundary for ai for marketing before launch. Estimate expected value per accepted outcome, gross margin, operating capacity, refund or rejection risk and the maximum loss allowed for learning. The budget becomes a controlled experiment only when the team knows what would make the test financially acceptable or unacceptable.
Use a break-even relationship that the business can audit: maximum acquisition cost equals expected contribution per accepted outcome multiplied by the probability that the measured event becomes that accepted outcome. Replace broad platform conversion counts with the state that actually creates value. The ai for marketing review should therefore connect measurement, risk review and improvement with cost per approved asset, a named owner and a dated change record.
Evaluate marginal performance when scaling. Average cost can remain attractive while the newest spend enters weaker audiences, placements or frequency bands. Compare the next budget increment with the approved threshold and keep the prior configuration available for rollback. For ai for marketing, apply the principle through a bounded test such as landing-page content support, and require policy or brand exception rate to support the next budget decision.
Quality, privacy, accessibility and governance
Quality control for ai for marketing includes inventory review, placement evidence, invalid-activity monitoring, creative compliance, landing integrity and outcome acceptance. No single vendor label proves quality. The buyer needs source-level or cohort-level evidence that can be connected to business results.
Privacy and governance are design inputs, not final checkboxes. Use only permitted data, minimize unnecessary identifiers, document membership and deletion rules, and avoid inferring sensitive personal characteristics. A targeting or retargeting feature should be rejected when the business purpose does not justify the data use. A practical ai for marketing brief can operationalize this step with creative variation assistant, while treating mistaking fluent output for verified accuracy as an explicit pre-launch risk.
Accessibility supports both user value and campaign reliability. Text, contrast, motion, controls and landing forms should remain understandable across devices and assistive technologies. Deceptive interaction patterns may increase accidental clicks while reducing trust and accepted outcomes. In a ai for marketing workflow, this control is most valuable when scaling low-value content or creative volume could otherwise make the reported result look stronger than the accepted business outcome.
Common failure modes and diagnostic order
The common failure modes for ai for marketing include adopting AI without a defined decision or workflow, using confidential or unlicensed inputs, and publishing unreviewed claims. These failures often look like media problems but originate in planning, data or measurement. Diagnose the earliest broken stage before changing bids or increasing creative volume.
A second group of risks includes mistaking fluent output for verified accuracy, automating bias or weak brand patterns, and scaling low-value content or creative volume. Protect the campaign with exclusions, budget limits, named owners, change logs and predefined stop conditions. The goal is not to eliminate uncertainty; it is to keep uncertainty visible and financially bounded. For ai for marketing, apply the principle through a bounded test such as landing-page content support, and require policy or brand exception rate to support the next budget decision.
When results weaken, compare the current period with a stable cohort. Check tracking, audience or placement mix, frequency distribution, creative age, landing performance, conversion lag and accepted-outcome rules. A disciplined diagnostic sequence prevents a team from solving the wrong problem. The ai for marketing review should therefore connect measurement, risk review and improvement with cost per approved asset, a named owner and a dated change record.
Put the guide into practice
Turn AI for Marketing into a bounded campaign test
With “Common failure modes and diagnostic order” documented, launch only the next reversible test. Set a spending limit, preserve the baseline and use source-level and audience controls so the next step depends on qualified outcomes for ai for marketing, not activity volume.
Create My Free Account
Failure-mode response cards
Adopting Ai Without A Defined Decision Or Workflow
For ai for marketing, this failure weakens evidence or business quality. Record the earliest observable signal, the accountable owner, the corrective action and the condition that confirms recovery before spend is expanded.
Using Confidential Or Unlicensed Inputs
Publishing Unreviewed Claims
Mistaking Fluent Output For Verified Accuracy
Automating Bias Or Weak Brand Patterns
Scaling Low-Value Content Or Creative Volume
30-day controlled rollout
Days 1–4: contract and instrumentation
Freeze the ai for marketing definition, outcome state, conversion map, source naming, exclusions and initial budget. Test events from impression or eligibility through accepted business outcome.
Days 5–10: controlled delivery
Launch a narrow ai for marketing test with a stable control. Review pacing, placements, audience overlap, creative rendering, landing performance and early quality signals without overreacting to small samples.
Days 11–20: diagnostic tests
For AI for Marketing, diagnose one issue at a time with a meaningful creative, targeting, placement, bid or landing hypothesis while preserving a control and waiting for conversion maturity.
Days 21–30: marginal scale decision
For AI for Marketing, reconcile accepted outcomes before each budget increase; expand one dimension only when the evidence is reproducible and operating capacity can support it.
Scaling without losing evidence
Scale ai for marketing one controlled dimension at a time. Expand budget, audience, geography, format, placement or creative inventory separately enough that the effect can be observed. Preserve a control and compare marginal outcomes, not only the blended account average.
A valid scale decision requires capacity as well as media efficiency. Confirm that sales, fulfillment, support, inventory, payment and compliance systems can absorb the expected outcome volume. Media that exceeds operational capacity may create lower-quality service, refunds or rejected leads that erase the apparent gain. A practical ai for marketing brief can operationalize this step with campaign anomaly triage, while treating scaling low-value content or creative volume as an explicit pre-launch risk.
Keep rollback simple. Store the last stable settings, creative set, audience rules and exclusions. If marginal cost, quality, tracking variance or operational load crosses the approved threshold, return to the stable configuration and investigate before another expansion. In a ai for marketing workflow, this control is most valuable when using confidential or unlicensed inputs could otherwise make the reported result look stronger than the accepted business outcome.
Where FroggyAds fits
FroggyAds can support ai for marketing when the plan benefits from self-serve access to multiple paid formats, source controls and campaign-level optimization. The platform connects advertisers with inventory from 750+ SSP integrations and lets buyers manage targeting, bids, budgets, source IDs and creative tests from one account.
Use FroggyAds as the execution layer, not as a substitute for the operating contract. Bring a defined objective, approved creative, landing page, tracking plan, exclusions and accepted outcome. Start with a bounded test, review source-level evidence and expand only after the business result is reconciled. For ai for marketing, apply the principle through a bounded test such as audience-insight synthesis, and require cost per approved asset to support the next budget decision.
The minimum deposit is $50, while a useful learning budget depends on format, market, bid level, conversion rate and the evidence needed for a decision. Avoid treating a minimum funding amount as a recommendation or a guarantee of statistically stable results. The ai for marketing review should therefore connect input data, permissions and provenance with policy or brand exception rate, a named owner and a dated change record.
Frequently asked questions
Where can AI save a marketing team time first?
Start with a repetitive, reviewable task such as draft classification, transcript summaries or campaign naming, then measure the hours saved.
Which marketing decisions still need human ownership?
People should remain accountable for positioning, customer promises, sensitive targeting, budgets and the approval of published work. Set those decision rights before a pilot starts.
How should marketers evaluate an AI-generated draft?
Check every fact, source, claim, tone, exclusion and customer instruction before the material reaches an audience. Check that the destination supports the same customer promise.
What customer data is appropriate for an AI tool?
Use only authorised, necessary information under documented retention, access and vendor terms that fit the business's obligations. Exclude unnecessary identifiers and resolve unclear permissions before use.
Can AI choose the best campaign audience automatically?
No. Models can support analysis, but biased inputs, weak labels and changing markets can still produce poor choices. Use the worksheet to convert the guidance into a documented, reversible and auditable process.
How does a team test AI marketing automation safely?
Run a small offline sample with known outcomes, named reviewers, failure cases and a clear route back to manual work.
What spending categories make up the full cost of AI-assisted marketing?
Include model usage, integrations, data preparation, review time, monitoring, security work and correction of inaccurate output. Assess the cost of an approved asset, not just model access.
What makes an AI vendor's performance claim credible?
A useful claim names the task, dataset, comparison, measurement period, error rate and conditions where performance weakens. Check that the underlying evidence can be reproduced.
How can brand voice survive AI-assisted production?
Give the tool approved examples and boundaries, then let a trained editor make the final wording and judgment. Review for biased or weak brand patterns before publication.
When should an AI marketing workflow be stopped?
Pause it after recurring factual, privacy, policy or customer-experience failures exceed the team's written tolerance. Retain the previous stable settings and investigate before restarting.
Official sources used for this guide
For AI for Marketing, use current primary platform, industry-standard and accessibility documentation; verify interfaces, policy terms, implementation steps and terminology before launch.
- NIST: AI Risk Management Framework
- NIST: AI RMF Playbook
- Federal Trade Commission: Artificial Intelligence Guidance and Enforcement
- Federal Trade Commission: Advertising and Marketing Basics
- Google Ads: How AI Max for Search Campaigns Works
- Google Ads: Build a Performance Max Asset Group Using Generative AI
- Google Search Central: Helpful, Reliable, People-First Content
- W3C: Web Content Accessibility Guidelines 2.2
AI for Marketing operating worksheet
Use the AI for Marketing worksheet to turn guidance into a documented process with a named owner, evidence requirement, decision rule, rollback point and review date.
Definition and measurement rules
Write the operational definition for ai for marketing before choosing a dashboard. Name the event, denominator, eligibility rule, attribution scope, time zone, currency and data owner. The assigned keyword wording is ai for marketing and ai for digital marketing; those phrases must resolve to one canonical decision boundary rather than competing calculations.
For AI for Marketing, keep evidence exportable, reproducible and clear enough for a reviewer who did not configure the campaign.
Audience, context and exclusion map
Document why each signal is relevant to ai for marketing, how it is collected or inferred, how long it remains valid and which exclusions prevent waste or policy risk. Mark overlap between prospecting, retargeting, customer and suppression groups so the same user state is not purchased repeatedly without intent.
Creative and landing contract
List every approved promise, proof source, format adaptation, call to action and landing destination for ai for marketing. Include size or device constraints, fallback creative, accessibility checks and the owner who can withdraw a claim or asset when the underlying evidence changes.
Forecast and failure scenario
Model conservative, expected and upside cases for ai for marketing using transparent assumptions for eligible reach, price, response quality, conversion maturity and accepted value. Add a failure case with the maximum learning loss, earliest reliable signal and conditions that stop delivery.
Source and cohort evidence
Preserve campaign, audience, placement, publisher or source, device, geography, creative and time identifiers where the buying environment allows it. When a dimension is unavailable, record the limitation and avoid quality claims that require evidence the platform does not provide. In a ai for marketing workflow, this control is most valuable when scaling low-value content or creative volume could otherwise make the reported result look stronger than the accepted business outcome.
Measurement reconciliation
Create a reconciliation table for ai for marketing with platform delivery, analytics events, business outcomes, variance, known cause, unresolved amount and accountable owner. Use the same time zone, currency and maturity window before comparing systems.
Change log and experiment record
For every material change to ai for marketing, record the observed problem, hypothesis, exact change, start time, expected signal, minimum evidence, result and rollback decision. This record protects learning across operators, agencies and copied campaigns.
Scale and rollback checklist
Before expanding ai for marketing, confirm that marginal economics pass, inventory or audience quality remains stable, frequency is controlled, creative coverage is sufficient, operations can absorb outcomes and the previous stable configuration can be restored quickly.
AI in digital marketing and artificial intelligence workflows
AI in digital marketing is not a separate discipline from AI for marketing. It is the application of governed AI assistance across research, planning, creation, buying, optimization and measurement within digital channels.
Marketing with artificial intelligence should remain outcome-led. The operating model must identify approved data, decision rights, evidence standards, review ownership, customer impact and the mature business result that receives credit.
Launch a controlled paid-media test
For the paid-acquisition side of AI for Marketing, FroggyAds provides self-serve campaign controls, source-level reporting, conversion tracking and budget ownership.
Create My Free Account
Search intent and buyer decision
AI for Marketing: Build a Clear, Measurable Operating Plan: the buyer decision this guide supports
Use AI for Marketing: Build a Clear, Measurable Operating Plan when the immediate task is to evaluate AI for Marketing: Build a Clear, Measurable Operating Plan as its own acquisition decision with explicit scope, controllable variables, source evidence and accepted outcomes. For performance-focused advertisers, the useful output is a documented media decision rather than another broad advertising overview. The nearest related FroggyAds page is Ai Marketing; this URL keeps ownership of the distinct task to evaluate AI for Marketing: Build a Clear, Measurable Operating Plan as its own acquisition decision with explicit scope, controllable variables, source evidence and accepted outcomes.
For the AI for Marketing: Build a Clear, Measurable Operating Plan decision, campaign objective, source quality, audience and market fit, ad format are the useful operating concepts. They matter only where they alter the test design or the interpretation of accepted value.
| Checkpoint | Campaign action | Evidence to keep |
|---|---|---|
| Fit | Define the buyer, accepted outcome and non-negotiable constraint. | Retain evidence specific to AI for Marketing: Build a Clear, Measurable Operating Plan and its accepted outcome. |
| Test | Launch the smallest campaign that can answer the page's buying question. | Retain evidence specific to AI for Marketing: Build a Clear, Measurable Operating Plan and its accepted outcome. |
| Decision | Keep, cap, exclude or expand from accepted-outcome evidence. | Retain evidence specific to AI for Marketing: Build a Clear, Measurable Operating Plan and its accepted outcome. |
Hypothetical calculation: if a controlled campaign for ai for marketing: build a clear, measurable operating plan spends USD 100 and produces 7 accepted conversions, accepted CPA is USD 100 / 7 = USD 14.29. Replace the inputs with your own campaign economics; this is not a FroggyAds performance claim.
Use FroggyAds as the execution layer for AI for Marketing: Build a Clear, Measurable Operating Plan: keep the offer and conversion definition stable, apply the needed media controls and let advertiser-side accepted value decide whether more spend is justified. Create your free FroggyAds account.
Direct answer
AI for Marketing: Build a Clear, Measurable Operating Plan — what matters first
Direct answer: This page helps you evaluate AI for Marketing: Build a Clear, Measurable Operating Plan as its own acquisition decision with explicit scope, controllable variables, source evidence and accepted outcomes. Keep the comparison or test inside that scope, then use FroggyAds campaign controls only where paid traffic is part of the decision.