Your first paid campaign has one job: to measure whether paid acquisition can work for your funnel at a price you can afford — and that job requires a fixed budget you can lose, one channel, one offer, and a kill threshold written before launch. The failure mode is optimizing: founders launch on three platforms with four creatives, get ambiguous results everywhere, and can't tell whether the funnel or the ads failed. The discipline is fewer variables, more attention per dollar, and a decision at the end.
This is a media-buying guide, not financial advice; ad platforms' policies and your industry's advertising rules apply.
Are you actually ready for paid?
Three preconditions, all cheap to verify. The funnel converts organically: your landing page turns a realistic share of qualified visitors into signups or demos — if organic and referral traffic doesn't convert, paid traffic won't either; ads amplify funnels, they don't fix them. Unit economics leave room: you know your CAC ceiling (LTV divided by 3 is the working heuristic) and your payback tolerance. Tracking works: conversion events fire correctly, attribution is configured, and you can see a signup's source — without this, you're buying fog. If any precondition fails, fix it first; this is usually a week of work and it's the highest-return week in paid's entire lifecycle.
How do you structure the first campaign?
| Decision | First-campaign answer | Why |
|---|---|---|
| Budget | Fixed, losable — e.g., 10× CAC target total | Enough for signal, capped for sanity |
| Channel | One, where your buyer demonstrably is | Channel-choice is the biggest variable; test one at a time |
| Audience | Narrow — your proven segment, intent or lookalike | Broad matching burns budget on curiosity traffic |
| Creative | 2–3 variants of one message, not 6 themes | Message test, not creativity contest |
| Offer | Your best-converting organic offer, unchanged | Don't test two moving parts at once |
| Duration | 2–4 weeks, no mid-flight panic edits | Algorithms need stability; humans need data |
The pre-written rules are what make it a measurement: kill threshold (spend equal to 3–5× CAC target with zero conversions → stop, the channel or funnel is broken) and success threshold (CPL within 50% of target with credible conversion quality → scale gradually). Write both on the campaign doc before the first dollar.
What do you actually measure?
The chain, every link: CPC → landing conversion → qualified signup → activated customer → CAC. The chain localizes failure precisely — cheap clicks with no signups is a landing-page problem (fix the page, not the ads); signups that never activate is a fit or targeting problem (narrow the audience or the promise); healthy CAC but no retention is the worst finding: the ads found customers you shouldn't want, and the budget bought future churn. Watch conversion quality, not volume: a first campaign that delivers 40 signups of which 12 activate is a better result than 300 of which 5 do — and the blend that ignores activation will tell you the opposite.
What do first campaigns cost, honestly?
Budget expectations matter: B2B CPCs in professional categories commonly run $3–10+ on search, $1–4 on social, and costs have been pressured upward as AI-era auction density increased — per Crunchbase News, AI-related companies absorbed roughly 90% of a record $189B funding month in February 2026, and heavily funded companies bid up professional-audience inventory. Compute your own ceiling instead of benchmarking vibes: if LTV is $6,000 and the 3:1 rule applies, CAC must land under $2,000 — which supports real bids and patient optimization. If the arithmetic leaves no room for realistic CPCs × conversion rates, paid isn't your channel yet; that's a finding, not a failure.
How do you scale what works?
- Gradually: raise budgets 20–30% at a time; doubling overnight resets the algorithm's learning and your data's meaning.
- Vertically first: more budget on the winning audience/message pair before adding new ones.
- Then horizontally: a second channel starts as its own single-channel test with its own kill threshold — never as a bundle.
And keep the organic engine running throughout: per the U.S. Small Business Administration's guidance, paid and organic channels work best as complements, with paid providing the data (which messages pull) that organic compounds for free. One channel, one offer, both thresholds written, the chain measured link by link, and a decision — scale, fix, or stop — at the end of the month. Then, and only then, add the second variable.
For more context, read When to Hire a Growth Team (and What It Needs Before It Can Work).
For more context, read The 4 Numbers That Decide Whether You Raise or Bootstrap.
For more context, read growth loops.