By Ashley KaysPlatform references reviewed October 13, 202617 channel and buying options
01 / THE FOUNDATION
Start with the outcome. Then choose the media.
PPC means pay per click. It is a charging model, not a channel. Some ads are bought by impressions or other optimization objectives. An RSA is a search-ad format. Advantage+ is an automation family. Sponsored posts are a distribution arrangement.
One modest budget should answer one focused question. Choose one offer, one meaningful conversion and one primary channel before expanding. More platforms can create more noise when each receives too little data.
Your one-page campaign brief
Buyer: role, situation, need and trigger.
Offer: benefit, price, proof and main objection.
Action: purchase, qualified inquiry, booked call or another meaningful event.
Economics: margin, acceptable acquisition cost and buying cycle.
Experiment: hypothesis, budget cap, sample plan and decision date.
A form submission, completed assessment and paid customer are different outcomes. Name the event precisely.
02 / BUYER INTENT
Meet the decision the buyer is making.
Intent
Observable signal
Useful message
Measure next
Problem-aware
“Why does my website get no leads?”
Explain the likely friction and offer a useful diagnostic.
Relevant engagement, diagnostic completion, progression to a qualified action.
Solution-aware
“Improve website conversion rate”
Show the approach, the benefit and evidence that answers an objection.
Qualified inquiry or product evaluation.
Comparing
“Conversion agency pricing”
Clarify fit, scope, price and why this option is useful.
Accepted lead, booked meeting or cart progression.
Ready to act
Specific service, product, buy, book or quote terms.
Match the exact offer and remove friction from the next step.
Purchase, qualified booking and net customer revenue.
Intent is a hypothesis inferred from behavior. Confirm it with outcomes: query and landing-page cohort → primary conversion → qualified opportunity → customer. Report brand, nonbrand, prospecting and remarketing separately. A return visit or long session does not prove someone is ready to buy.
03 / CHOOSE THE CHANNEL
Seventeen options. A clear job for each.
Open a channel for its audience fit, creative needs, test approach and decision criteria. The spend ranges are illustrative USD media-only planning allowances, generally for 30 days. They are not promises, published CPC benchmarks or universal minimums. Production, management, tools and taxes are additional.
Showing 17 options
01Capture existing demand
Google Search Ads
$1,000–$3,000+
The useful budget depends on query cost and conversion economics; this range is a planning scenario.
What it is
Text ads that can appear when someone searches. Google Ads is the larger buying platform: Search, Shopping, Performance Max, Display, YouTube and Demand Gen serve different jobs.
When to use it
People already look for your service, product or the problem you solve. Start with tightly related commercial queries and a matching destination.
Audience & demographics
Search intent, service area and purchase context lead the plan. Age is a test dimension, not a substitute for a relevant query.
Buyer intent
High for pricing, buy, quote, book and specific product searches; lower for broad how-to research. Separate brand from nonbrand.
What to create
A focused keyword map, scoped negatives, responsive search ads, useful sitelinks, a fast landing page and a verified conversion event. Shopping also needs accurate product data.
How to test
Compare one offer or landing-page hypothesis within a stable intent group. Start with a manageable exact/phrase structure, then assess broader matching with reliable conversion feedback.
What to monitor
Search terms, qualified conversion rate, cost per qualified lead, impression share constrained by budget, device/location quality and closed revenue.
When to pause or scale
Exclude irrelevant queries immediately. Expand proven intent groups when additional leads remain qualified and marginal acquisition cost fits the business.
Planning range, not a platform minimum. Check demand in your target market before allocating the full amount.
What it is
Search advertising through Microsoft Advertising, including Bing search inventory. Other Microsoft audience placements are a separate distribution choice.
When to use it
A search offer already works, or Keyword Planner shows relevant demand that your current campaign is not reaching.
Audience & demographics
Use the actual market, query and device mix. Do not assume every Bing user is older, affluent or a business decision-maker.
Buyer intent
Often high when queries show a purchase decision. Search-partner and audience inventory should be reviewed separately.
What to create
Search ad copy, conversion tracking, negative lists and an import review. Check locations, bids, URLs, goals and network settings after importing Google campaigns.
How to test
Run a separately measured search test with its own budget. Compare qualified outcomes, not a blanket assumption that clicks will be cheaper.
What to monitor
Query relevance, partner distribution, duplicate goals, cost per qualified outcome and new-customer revenue.
When to pause or scale
Reduce inventory with persistently poor lead quality after conversion lag. Scale the query groups that hold acquisition economics.
An automation approach, not an additional channel or separate compulsory budget. Available controls vary by campaign and account.
What it is
Meta’s family of automated advertising tools for tasks such as audience expansion, placements and creative variations. Review the exact features enabled in the current setup.
When to use it
Your offer is clear, conversion events are reliable and you have varied creative. Automation needs useful feedback about the business outcome.
Audience & demographics
Broader delivery can discover buyers outside your first audience hypothesis. Use actual buyer data and eligible controls; avoid assuming a narrow persona describes everyone who will convert.
Buyer intent
Can span prospecting and warmer audiences. Inspect new versus existing customers and the conversions the system is learning from.
What to create
Accurate conversion values, catalog data where relevant, distinct creative concepts and a review of any automated image or text enhancements.
How to test
Where available, use a controlled experiment against a comparable setup. Keep objective, offer and measurement window aligned.
LinkedIn lists a $10 daily minimum. That buying floor is not evidence that $10/day will generate a useful B2B sample.
What it is
Sponsored professional-feed content and other LinkedIn formats, with targeting based on professional attributes.
When to use it
The buyer’s role, industry or company matters, contract value supports higher acquisition costs, and sales can follow up properly.
Audience & demographics
Job function, seniority, company size and industry are usually more useful than age. Distinguish the champion, evaluator and budget owner.
Buyer intent
Often problem-aware but not actively shopping. A relevant diagnostic, point of view or case study can start the relationship.
What to create
One specific business problem, credible proof, an image or document, a focused lead form or page, and CRM qualification and follow-up rules.
How to test
Test a narrow role/offer combination against a second message. Avoid changing the audience, offer and format simultaneously.
What to monitor
Cost per qualified account, sales acceptance, opportunity creation, pipeline aging and closed-won revenue by cohort.
When to pause or scale
Pause if leads fail qualification or sales cannot follow up. Scale when the cohort creates opportunities at an acceptable acquisition cost, not merely inexpensive forms.
TikTok documents campaign budgets above $50 and ad-group daily budgets above $20. These are different hierarchy levels, not amounts to add together.
What it is
Paid video and other ad formats within TikTok, using a feed-native creative style.
When to use it
The product can be demonstrated quickly, a problem is instantly recognizable and you can produce fresh video concepts regularly.
Audience & demographics
Skews younger in U.S. adult usage, with meaningful use beyond ages 18–29. Validate product demand and purchase behavior in your specific market.
Buyer intent
Usually discovery-led. A vivid demonstration can move someone toward consideration; views alone are a weak business signal.
What to create
Vertical video with an early hook, visible product use, captions, a truthful benefit and a simple CTA. Obtain rights for creator content and music.
How to test
Start with a few genuinely different hooks and proof angles. Compare conversion quality after enough time for the purchase cycle.
What to monitor
Early retention, click-to-page drop-off, qualified conversion rate, new-customer cost and fatigue.
When to pause or scale
If attention is strong but sales are weak, inspect the offer and page. Scale repeatable converting concepts and maintain a creative replacement pipeline.
Gmail is an eligible Demand Gen placement, not a standalone paid email blast. Do not double-count it inside a wider Demand Gen allocation.
What it is
Visual ads delivered in eligible Gmail inventory through Google Demand Gen. Demand Gen also supports other Google channels, with controls and eligibility depending on setup.
When to use it
Your offer is visual and understandable outside a search moment, and you can measure the quality of the people who visit.
Audience & demographics
Audience and context matter more than an assumed “Gmail demographic.” Your entire customer email list is not automatically the reachable ad audience.
Buyer intent
Generally discovery or returning interest. Opening or expanding an ad does not establish purchase intent.
What to create
High-quality images, concise benefit copy, brand assets, an offer-matched landing page and reliable conversion tracking.
How to test
Use available channel controls to isolate a Gmail question when the volume supports it; otherwise evaluate the broader Demand Gen plan honestly.
What to monitor
Placement-level outcomes where available, engaged visits, qualified conversions and click-versus-view attribution.
When to pause or scale
Remove weak placements or creative after a useful sample. Scale when additional spend produces valuable customers, not just ad interactions.
Negotiated planning allowance, not a platform minimum. A major creator or publisher may cost much more; production and usage rights change the fee.
What it is
Paid distribution or endorsement through a creator, publisher or newsletter. This is a commercial arrangement, not one ad network.
When to use it
A trusted voice reaches a specific buyer group and can explain the product credibly in its normal editorial format.
Audience & demographics
Request current audience geography, relevant interests and prior campaign evidence. Followers alone do not reveal reachable buyers.
Buyer intent
Depends on the audience relationship, topic and offer. A narrow, relevant audience can be more useful than broad exposure.
What to create
A brief, approved claims, deliverables, usage rights, tracked links, a landing page, disclosure requirements and reporting expectations.
How to test
Use a small partnership batch with comparable offers and tracked destinations. Compare qualified outcomes and incremental reach where possible.
What to monitor
Actual delivered reach, engaged visits, assisted conversions, customer quality and the full fee including production and rights.
When to pause or scale
Do not renew solely on impressions. Rebook when the audience and customer economics hold. Paid connections should be clearly disclosed near the endorsement.
Current OpenAI guidance lists a $25 daily minimum. Availability, objectives and account eligibility remain subject to the current Ads Manager rollout.
What it is
Paid ad placements in eligible ChatGPT experiences. Ads are separate from answers; buying an ad does not buy a recommendation inside an answer.
When to use it
You have access, a useful offer fits a decision someone is exploring, and you can run a contained test with accurate conversion measurement.
Audience & demographics
Use the eligible audience and location controls in your account. Do not invent a fixed age profile or assume advertisers can read individual conversations.
Buyer intent
A conversation can signal research or consideration, but it is not identical to a Google exact-match search. Validate the quality of resulting visits.
What to create
Clear offer copy, a credible destination, accurate tracking and a defined success event. Inspect current creative and account requirements before activation.
How to test
Begin with one offer and a bounded budget. Keep results separate from other search channels and allow for reporting and conversion lag.
What to monitor
Eligible delivery, spend, qualified conversions, attributed net revenue and the difference between engagement and completed business outcomes.
When to pause or scale
Fix tracking or poor offer fit first. Scale only after observed economics support it; treat early small samples as learning.
Google placements can overlap across campaign types. Facebook, Instagram and Advantage+ can share a single Meta plan. Count the actual campaign allocation once.
04 / AUDIENCE REALITY
Demographics help you find people. Their needs explain the purchase.
Use demographics alongside situation, motivation, buying role and observed behavior. The table describes U.S. adult platform usage in 2025, not ad reach, customer composition or conversion probability.
Share of each age group who say they use the platform
“A founder with a working site and a small sales pipeline, who needs to understand why inquiries stall before purchase.” The trigger is lost opportunities. The objection is paying for another vague marketing report. The proof needed is a clear scope and a tangible improvement.
How to validate it
Interview customers, review sales objections and analyze qualified conversions by cohort. Treat demographic ranges as test hypotheses until customer data supports them. Keep targeting appropriate for the product and applicable platform restrictions.
05 / KEYWORD RESEARCH + NEGATIVES
Find useful demand. Exclude the wrong job.
Start with the offer. List customer language, problems, alternatives, products, locations and purchase modifiers from interviews, sales notes and site search.
Expand and validate. Use Keyword Planner with a consistent country, language, network and date range. Record volume and CPC estimates with their settings. Organic Search Console impressions are evidence of queries, not paid-search volume or CPC.
Group by decision. Keep brand, category, competitor and exploratory research separate. Map each cluster to one offer and a relevant page.
Choose matching deliberately. Exact and phrase are useful starting structures, but can still match related intent. Broad matching needs strong goals, useful feedback and close query review.
Review actual search terms. Add relevant winners, investigate poor lead quality and make scoped exclusions. Record the reason and date for each negative.
Candidate negative
Potential scope
Why it needs judgment
“free template”
Paid consulting ad group; phrase candidate.
May exclude a template seeker. Do not add “free” globally if you also promote a free assessment.
“agency jobs”
Service acquisition campaign; phrase candidate.
Recruitment intent may be irrelevant to hiring an agency. “Jobs” can be valuable for an education or career campaign.
[brand support login]
New-customer campaign; exact candidate.
Existing-customer navigation may belong elsewhere; inspect actual terms first.
“how to”
Review list, not an automatic exclusion.
Could be early research from a future buyer. Excluding it can remove useful demand.
Use account-level exclusions only for terms irrelevant everywhere. Campaign and ad-group negatives can protect offer boundaries. Google’s negative matching behaves differently from positive matching and does not automatically cover every close variant; check relevant plurals and synonyms.
Google RSAs can contain up to 15 headlines of 30 characters and four descriptions of 90 characters; display paths allow 15 characters each. The system varies the combination, order and amount of text shown.
Cover distinct jobs: relevant search language, audience, benefit, offer, proof, objection and action. Avoid 15 versions of the same sentence. Each asset should make sense alone and alongside the others.
Pin only when a message must appear in a supported position. Ad Strength can diagnose asset variety; it is not a measure of profitability or a reason to weaken a clear offer.
A strong ad makes it easy to recognize the problem, understand the improvement and choose the next step. Conversion comes from the whole experience: message, offer, proof, destination and follow-up.
Build the message in six parts
Person: who is this useful for?
Problem: what friction are they experiencing?
Benefit: what becomes easier or better?
Mechanism: how does the offer help?
Proof: what real evidence supports it?
Action: what happens when they click?
Example: a site that loses leads
“Website visits. Leads coming in. Still too few sales?”
The hook names the problem. The $997 Growth Fix gives it a concrete next step: review the journey, make one agreed improvement and create a measurement plan. “Find my growth leak” is useful because the destination explains exactly what the buyer receives.
Use a real demonstration, customer quote with permission or documented outcome when available. Never invent a testimonial, revenue number or guaranteed result to make an ad more persuasive.
The creative production checklist
Prepare a small set of distinct concepts, readable feed images, suitable vertical versions, captions, clear branding and a mobile page. Keep important text inside placement-safe areas. Match the ad’s offer, price and CTA to the destination. Review image accuracy, rights and automated creative variations before launch.
Start with three hooks and two proof approaches as a creative backlog. Run only the comparisons your budget can support at once. A new background color is a weaker initial question than whether the buyer cares about the benefit at all.
08 / AD SPEND + EXPECTED RESULTS
Model the economics. Then replace assumptions with evidence.
No amount guarantees impact. A useful budget depends on auction cost, conversion rate, margin, sales cycle and how precise the decision needs to be. For example, a $50 target cost per acquisition implies $1,500 for 30 acquisitions if the target holds. Thirty observations do not automatically make an experiment statistically reliable.
EDITABLE 30-DAY PLANNING SCENARIO
What would this test need to earn?
Assumptions, not recorded results
Two alternative interpretations: the purchase model treats conversions as orders; the lead model treats them as inquiries. Do not add those outcomes together. Fractional outputs are mathematical expected values, not a guarantee or confidence interval.
At a 40% contribution margin before advertising, break-even ROAS is 1 ÷ 0.40 = 2.5×, before fixed overhead. Deduct product cost and variable fulfillment, payment and return costs when estimating the margin.
Lead-generation economics
Qualified CPL = spend ÷ qualified leads Media CAC = lead CPL ÷ lead-to-customer rate Fully loaded CAC also includes relevant sales and acquisition costs.
A $50 lead with a 20% close rate implies $250 in media per customer. Use the same lead definition and cohort at every step. Assigning an arbitrary dollar value to a form does not establish revenue ROAS.
Measure revenue consistently
Use tagged links, analytics, platform pixels or eligible server events, deduplication and CRM/order IDs as appropriate. Test one real conversion through the complete flow. Respect consent choices. Record the attribution window, timezone, currency and revenue definition; subtract refunds and use consistent treatment of tax and shipping.
Platform dashboards can each claim the same sale. Reconcile against the order system or CRM and compare new versus existing customers. ROAS measures attributed revenue, not profit or causality. Blended revenue ÷ total ad spend is a useful business ratio, but it does not prove incremental revenue.
09 / A TEST YOU CAN LEARN FROM
Define the decision before buying the clicks.
Step
Make it concrete
Hypothesis
“Explaining the deliverable will increase qualified bookings compared with a vague growth promise.”
Primary outcome
Choose one: purchase, qualified inquiry or accepted sales opportunity. Track CTR and video retention as diagnostics.
Comparison
Change one major variable. Use a randomized platform experiment when available; ordinary ad delivery is not an equal random split.
Sample and timing
Set a baseline, the smallest worthwhile effect and a sample plan. Observe complete conversion cycles and seasonal patterns. A two-week test is not automatically enough for a six-week sales cycle.
Guardrails
Budget cap, lead quality, margin, refund rate, frequency and the ability to fulfill or follow up.
Decision
Predefine continue, revise or stop criteria. If the sample is too small, label the result inconclusive instead of inventing a winner.
Check delivery frequently
Catch broken links, tracking failures, overspend, irrelevant queries and offer errors promptly. Monitor spend and pacing daily during launch. This is operational quality control, not a reason to change every ad every day.
Review performance on a useful cycle
Review search terms and creative diagnostics regularly, then judge customer outcomes after conversion lag. Repeatedly checking for a winner and stopping at the first favorable fluctuation can produce false confidence.
10 / PAUSE, REPAIR OR SCALE
Act on the part that is failing.
PAUSE / REPAIR
The experience is broken.
Stop a broken destination, wrong price, irrelevant location, faulty conversion event or offer you cannot deliver. Fix the issue and verify the full journey before resuming.
REVIEW / LEARN
The economics are unclear.
If spend reaches a pre-agreed review point without qualified outcomes, inspect intent, creative, page and follow-up. “Three times target CPA” can be a review trigger, not a universal proof that an ad cannot work. Account for lag and small samples.
SCALE / MONITOR
The customer outcome holds.
Increase spend when tracking is sound, qualified outcomes are repeatable and marginal acquisition cost fits the economics. A small 10–20% budget step is one cautious operating heuristic, not a platform rule. Watch quality, frequency, inventory, capacity and conversion lag after each change.
High CTR with weak conversions points toward audience mismatch, an unclear offer or landing-page friction. Cheap leads with weak sales point toward qualification or follow-up. A strong average ROAS can hide expensive new customers behind returning customers. Diagnose before deciding which lever to change.
SEE THE STRATEGY IN PRACTICE
From research to the ad to the next customer.
Explore seven client campaign stories with full RSA copy, image ads, personas, buyer intent, keyword plans and editable budgets.