Skip to content

Paid Advertising

First PublishedLast UpdatedByAtif Alam

The decision this page enables: which paid channels to run, what creative to ship, what budget framework to follow, and how to tell what’s actually working.

What paid advertising is — and what it isn’t

Section titled “What paid advertising is — and what it isn’t”

Paid advertising is renting attention on someone else’s platform. Search engines, social networks, programmatic display, video, audio, out-of-home — anywhere you hand over money in exchange for guaranteed placement in front of a defined audience.

It is not the same as:

  • Influencer marketing — paying a person for their audience (see Influencer Marketing).
  • Affiliate marketing — paying on outcome (see Affiliate Marketing).
  • Earned media / PR — coverage you don’t pay for (see PR).

The distinction matters because the economics, measurement, and creative approach are different. This page is exclusively about paid placement.

Every paid ad you run does one of three jobs. Knowing which job your ad is doing tells you what creative to make, where to run it, and what to measure.

Catching people who already know they want something like your product and are actively searching. Search ads are the canonical example.

  • Channels: Google Search, Bing Ads, retargeting (warm cookies), Apple Search Ads, marketplace ads (Amazon Sponsored, App Store Search Ads).
  • Creative: short, intent-matched, comparison-or-feature-led. The user has done their thinking; you just need to show up.
  • Measure on: last-click CAC, conversion rate, ROAS.
  • Typical role in mix: 40–60% of paid budget for products with established category demand.

Creating awareness or shifting consideration among people who aren’t actively searching. Social, video, display, audio, OOH.

  • Channels: Meta (Facebook + Instagram), TikTok, YouTube, LinkedIn (for B2B), display/programmatic, podcast ads, Twitch, Spotify, OOH.
  • Creative: story-led, problem-then-solution, creator-style. The user wasn’t looking; you have to earn the click.
  • Measure on: incrementality (geo holdouts, conversion lift tests), assisted contribution, branded-search lift, attention metrics.
  • Typical role in mix: 30–50% of paid budget for products with a creative-able value prop.

Re-engaging people who already visited / signed up / lapsed. Warm lists, custom audiences, lookalikes off your customers.

  • Channels: Meta + Google retargeting; email-list-uploaded audiences on every platform; in-app ad networks.
  • Creative: reminder, social proof, discount, recovery, win-back.
  • Measure on: incrementality (against a holdout — retargeting often double-counts conversions that would have happened anyway).
  • Typical role in mix: 10–25% of paid budget.

Channel quick-pick (which job, which audience, which platform)

Section titled “Channel quick-pick (which job, which audience, which platform)”
JobIf your audience is…Try firstNotes
Demand captureHigh-intent searchersGoogle SearchDefault starting point for established categories.
Demand captureB2B SaaS buyers in-marketLinkedIn (targeted) + GoogleLinkedIn precision; Google volume.
Demand captureConsumer app installersApple Search Ads + Google App CampaignsASO + paid in tandem.
Demand creationConsumer broad audienceMeta Reels + TikTokCreator-style UGC outperforms agency creative 2–3×.
Demand creationB2B mid-market & enterpriseLinkedIn + YouTubeLinkedIn for targeting; YouTube for proof + depth.
Demand creationLocal / regional consumerMeta + OOH + local programmaticPair digital with physical for the “feel everywhere” effect.
Retargeting / lifecycleSite visitors / trial usersMeta + Google + email-driven custom audiencesCap frequency to ≤3–5 impressions/week per user.

Until ~2022, paid-ad performance was driven mostly by targeting precision. After Apple’s iOS 14 changes and the evolution of ad-platform ML, the lever shifted: modern ad platforms find your audience for you; you bring the creative.

What this means in practice:

  • Stop micro-segmenting — give the algorithm a broad audience and let it find the converters.
  • Spend creative time and money — the 80/20 rule: in 2026, creative is 80% of paid-ad performance. Targeting is the 20%.
  • Refresh creative every 2–4 weeks — even your winners decay. Plan a continuous creative-production pipeline, not a quarterly “campaign launch.”
  • Test creative concepts, not headline tweaks — Variant A and Variant B should represent different ideas, not different word choices.

A practical way to split paid budget so you keep performance steady while still exploring:

70% Proven — channels and creative that are working today; scale + maintain.
20% Experimental — channels you've validated to v1 but haven't fully scaled; or new
creative concepts in proven channels.
10% Wild — speculative bets; new platforms, new creative formats, new audiences.
Most won't work; one in five will.

Two related rules:

  • Never scale on a hot week. A channel that suddenly performed 3× last week is more likely to revert than to keep climbing. Wait 14 days.
  • Never kill on a cold week. A channel that suddenly underperformed last week may just be statistical noise. Wait 14 days.

One brief per concept; ship 3–5 briefs per channel per month at minimum.

Concept name: [e.g. "Cold-DM survival kit hero"]
Channel: [Meta / TikTok / LinkedIn / Google / YouTube]
Job: [Demand creation / Demand capture / Retargeting]
Hypothesis: [e.g. "Showing the 'before / after' inbox shot will outperform
the current product-screenshot ad by 1.4× on CTR among
SMB SaaS founders age 28–44."]
Creative concept: [one-sentence description, not a script]
Reference / mood board: [link]
Length / aspect ratio: [e.g. 9:16 vertical, 0–30s]
Hook (first 3 sec): [e.g. "Your inbox at 9 AM. Now your inbox after one week."]
Audience: [as wide as algorithm will accept; do not over-narrow]
Geos: [e.g. US + UK + Canada]
Daily budget: [e.g. $300/day]
Duration: [e.g. 14 days minimum]
Success metric (primary): CAC ≤ $X (vs current channel CAC of $Y)
Success metric (secondary): hook retention >25% at 3s (TikTok / Reels)
Kill metric: [e.g. CAC > $1.5X for 5 consecutive days]
Kill timeline: [decide by day 7; kill or scale by day 14]
Owner: [name]
Reports to: weekly creative review

Ad-account naming convention (sanity rule)

Section titled “Ad-account naming convention (sanity rule)”

Pick a structure and never deviate. Future-you will thank present-you.

{Channel}-{Campaign objective}-{Audience}-{Concept}-{Variation}-{Date launched}
Examples:
META-DemandCreation-SMBFounders-BeforeAfterInbox-V1-2026-05-10
META-DemandCreation-SMBFounders-BeforeAfterInbox-V2-2026-05-10
LINKEDIN-DemandCapture-MidMarketPMs-PerSeatComparison-V1-2026-05-15
GOOGLESEARCH-DemandCapture-NotionAlternative-StandardSearch-V1-2026-04-30

Without a naming convention, your ad-account becomes unreadable within 90 days and post-hoc analysis becomes impossible.

  • CAC by channel — the headline number. Segment by job (capture / creation / retargeting) because the absolute values differ ~3×.
  • ROAS — revenue per dollar spent. Useful for e-commerce; less useful for B2B subscription unless you bake in expected LTV.
  • CPM (cost per 1,000 impressions) — efficiency of placement. Rising CPMs in the same channel are usually a creative-fatigue signal.
  • CTR (click-through rate) — how compelling the ad is at first glance. Healthy: ≥1% for display, ≥2% for social, ≥3% for search.
  • CVR (conversion rate after click) — how well the landing page matches the ad’s promise. Healthy: ≥5% for self-serve SaaS landing pages; ≥3% for high-consideration B2B landing pages.
  • Frequency — impressions per unique user per week. Above 5–7/week, you’re wasting money on diminishing returns.
  • Brand-lift — does running this ad increase branded search? Run brand-lift studies on demand-creation campaigns ≥$50k.
  • Incremental-lift (geo holdout / conversion lift) — the only honest measure of demand-creation effectiveness. Run quarterly.
  • CAC payback — for B2B SaaS, ≤12 months is healthy; ≤18 months is acceptable; >24 months is a money-loser hiding in the average.
  • Creative refresh rate — how often you ship new creative concepts. Floor: every 2–4 weeks per channel.
  • Channel concentration — % of paid budget in your largest channel. Above 60% is fragile.

SaaS workspace — paid mix at growth stage

Section titled “SaaS workspace — paid mix at growth stage”

Annual paid-ad budget: $1.4M. Split:

Google Search 60% ($840k)
- "notion alternative", "team docs tool", "[competitor] alternative", brand defense
- CAC ~$42 self-serve; ~$280 for Team-tier qualified
- 4× ROAS on annualized contribution
LinkedIn 25% ($350k)
- Demand capture: in-market product researchers (job titles + company-size filters)
- Demand creation: retargeting + lookalike off Team-tier customers
- CAC ~$180; ACV ~4× self-serve = strong ROI
YouTube 15% ($210k)
- Founder-led explainer (90 seconds, "what's different about flat-team pricing")
- Pre-roll on B2B SaaS / productivity / remote-work channels
- Measure: brand-lift (+18%), branded-search lift (+24% in test geos)

Within Google: 70/20/10 split is “branded + 5 proven non-brand queries” (proven), “20 newer non-brand queries” (experimental), “competitor terms + display + YouTube discovery” (wild).

After 12 months: total CAC blended $58 → $64 (slightly up due to scaling), payback 9 months (healthy), branded-search volume +110% YoY (suggesting demand-creation work is compounding).

Consumer fitness app — paid mix at growth stage

Section titled “Consumer fitness app — paid mix at growth stage”

Annual paid budget: $3.2M. Split:

Meta (Reels-style UGC) 70% ($2.24M)
- Demand creation: creator-style UGC ads (60+ creators on retainer)
- CAC ~$28 (vs $52 from polished agency creative); CPI ~$4.20
- Refresh creative every 14 days
TikTok 20% ($640k)
- Demand creation: experimental; ramping creator partnerships
- CAC ~$34; growing volume MoM
YouTube 10% ($320k)
- Demand creation + lifecycle: pre-roll on fitness / wellness / nutrition content
- Higher-LTV cohort (~1.3× average LTV)

The team has an in-house creator studio (2 PMMs + 2 producers + 6 freelance creators) that ships ~30 creative variants per month across the three channels. Creative spend is ~12% of media spend — well within industry norms but on the higher end because creator-style content is the leverage.

The 10% “wild” bucket has tested Audible/Spotify audio ads (modest), pre-roll on health podcasts (good — moved to experimental), and broadcast TV (failed — killed at day 21).

  • Scaling before payback is proven. “It’s growing fast” can mean “we’re acquiring expensive customers fast.” Don’t 2× your daily budget until the existing budget has hit payback.
  • One ad in a channel and calling it a test. A channel needs 5–10 creative concepts to know if the channel itself is the issue. Don’t generalize from a single-creative result.
  • Ignoring creative refresh rate. Every winning ad decays. Plan replacements before the decay starts, not after.
  • Attribution narcissism. Each ad platform claims credit for the same conversion. Last-click attribution undercounts demand creation; ad-platform attribution overcounts each channel by 2–4×. Use an independent measurement layer (incrementality, geo holdouts, MMM).
  • Confusing volume with quality. A channel that drives lots of trials but no paid conversions is a money pit. Track downstream metrics, not just top-of-funnel.
  • Over-narrowing audiences. In 2026, the platform ML is better than your manual segmentation. Go broad; let the algorithm find converters.
  • Killing demand-creation campaigns by last-click attribution. A YouTube campaign that “doesn’t drive direct conversions” may be driving a 30% lift in branded search. Measure both.
  • Running paid without a creative-production pipeline. Hiring an agency to ship 1 ad/month and wondering why scale stalled.
  • No frequency cap on retargeting. Showing the same ad 12× / week to your existing audience trains them to ignore you (and erodes brand equity).
  • Letting the lowest CPM win. Cheap CPMs in low-value placements are still bad spend. Track conversion downstream, not impression cost.
  • Google Ads + Google Analytics 4 — Search + YouTube + display.
  • Meta Ads Manager + Meta CAPI — Facebook + Instagram + Audience Network. Conversion-API server-side tracking is essential post-iOS 14.
  • LinkedIn Campaign Manager — B2B targeting + Insight Tag.
  • TikTok Ads Manager + Spark Ads — TikTok and creator-amplified UGC.
  • Apple Search Ads + Google App Campaigns + AppsFlyer / Branch — mobile measurement.
  • Triple Whale / Northbeam / Rockerbox — independent attribution layers.
  • Haus / Measured / Recast — incrementality + MMM measurement.
  • Foreplay / Motion / Atria — creative inspiration and ad-library analysis.
  • Hacking Growth (Sean Ellis) — the canonical growth/paid playbook.
  • Performance Branding (Mark Pollard) — the brand + performance integration argument.

See also: Martech Stack & Automation for the AI tooling for ad creative production, the experimentation discipline behind incrementality testing, and the attribution philosophy for measuring paid impact across the funnel.