How first campaigns earn learning

Pick one objective that matches money#

Choose the Sales objective optimised to the event that pays you: purchase for shops, qualified lead for services, booking for local firms. Everything upstream, from audience to creative, inherits its discipline from this one choice, so vanity objectives like reach or video views never belong in a first sales campaign.

Confirm the choice against reporting: if finance recognises purchase revenue, optimise to purchase, not to add to cart. Meta own objectives guide maps each objective to its events. Start at the Meta Ads hub, and read our objectives guide before you commit.

Write the objective on the brief with the event name beside it, because ad sets drift within a quarter and audits need the original intent. One campaign, one money event, no exceptions in month one. Standard events list the vocabulary to use.

Signal before spend#

Install Pixel plus Conversions API before spending a pound: browser pixel captures on-site behaviour, server events repair what browsers lose, and matching quality decides how much of it Meta can use. Fire and test the exact optimising event with a live transaction, then confirm it in Events Manager.

Verify domain ownership and set event priorities where platform prompts require it, so measurement survives privacy prompts intact. Meta Pixel setup walks the install, and our Pixel guide adds the test checklist. Pair with Conversions API setup and dedup rules so server and browser events count once.

Delay launch by a day rather than launch blind: a week of spend on broken signal teaches nothing and costs plenty. Clean signal is the cheapest performance lever in the account. Domain verification closes the loop.

Structure: one campaign, one ad set#

Launch one campaign with a single broad ad set: no interest stacks, no five-way splits, no lookalikes on day one. Consolidation gives the single ad set every conversion signal, which is the only way a new account exits learning on a modest budget.

Keep Advantage+ audience suggestions broad and controls minimal: location, minimum age and language only. Consolidation explains the maths, and Advantage+ audience separates steering from binding. Resist the urge to mirror old account structures that were built for different signal levels.

Name everything for the future: campaign for the objective, ad set for the audience and date, ads for the concept and hook. Midnight debugging thanks tidy namers. Budget placement follows the same less-is-more logic.

Budget for learning, not for comfort#

Fund the single ad set to earn roughly 50 optimised events in seven days: divide your typical cost per event into that target before launch, and do not launch if the maths needs a miracle. Thin budgets produce Learning Limited ad sets that never stabilise, and every later read inherits the wobble.

Prefer campaign budget over ad set budgets at this stage so delivery can route spend to the working ad without manual moves. Meta learning phase guidance defines the thresholds, and our learning guide turns them into budget sums. Hold the budget steady for the first full week regardless of early nerves.

Record the budget, the event target and the date in a launch log; week one feelings fade but the log keeps the next test honest. Ad set budgets return later, once scale demands control.

Creative minimum that survives review#

Ship three to five materially different ads: distinct hooks, distinct proof, distinct offers, not colour swaps of one idea. One image control, one creator-style vertical video and one carousel or Reels cut give delivery real choice across Feed, Stories and Reels from day one.

Write primary text under the visible window with the payoff first, keep in-image text to the words that must survive every crop, and caption every video for muted viewing. Diversification shows why concepts beat variants, hooks time the opening, and overlay rules keep words legible.

Check ratios before launch: 4:5 for Feed impact, 1:1 for universal coverage, 9:16 for Reels and Stories. Ratios guide maps the coverage in full.

Week one reads and the next move#

Read in order: delivery first, creative second, offer last. Delivery questions ask whether the ad set exits learning; creative questions compare thumb-stop and outbound click rate across ads; offer questions wait until clicks arrive but conversions do not. Kill nothing before each ad has meaningful impressions except clear policy or tracking faults.

Promote the winning concept into ratio variants and fresh hooks in week two, and only then consider a prospecting and retargeting split. Clean test rules protect the read, and prospecting splits structure what follows. Talk to our team for a free launch audit when the numbers puzzle you.

Log every decision with spend, events and reason: the log becomes the media plan by month three. Patience in week one compounds all year. Testing rhythm schedules what comes next.

Frequently asked questions#