Skip to main content
Use this guide to run a campaign with FloKit as the payback and recommendation layer. The goal is not just to launch spend. The goal is to run a controlled campaign where acquisition, conversion, revenue, and guardrails are visible from the start. This guide assumes your subscription source, attribution source, and spend data are connected.

Campaign goals

Start by choosing one clear campaign goal:
GoalExample targetFloKit focus
Validate a new channelTikTok annual-plan acquisitionCAC, trial conversion, payback
Scale a proven campaignIncrease budget on strongest cohortSpend, ROAS, payback window
Test creative qualityCompare UGC hooksConversion, renewal quality, CAC
Test offer or paywall signalAnnual plan emphasisTrial-to-paid, refund rate, payback
Reduce wasteCut weak cohortsBudget shifts, creative pauses, exclusions
Avoid launching with too many goals. One campaign should have one primary decision.

Prerequisites

  • Production workspace is active.
  • Revenue source is connected.
  • Attribution source is connected.
  • Spend data is available.
  • Campaign naming convention is documented.
  • Target market, platform, channel, and offer are defined.
  • Guardrails are configured.
  • Review owner is assigned.

Campaign setup

1

Define the campaign brief

Document objective, channel, audience, market, budget, offer, creative set, target payback window, and decision date.
2

Confirm tracking

Make sure campaign, ad set, creative, country, platform, and offer fields are available in the attribution source and can be joined to revenue.
3

Set guardrails

Configure campaign-level spend caps, excluded campaigns, approval rules, and rollback triggers before launch.
4

Launch in the ad platform

Launch the campaign using your normal ad platform process. FloKit reads the resulting spend and attribution data.
5

Monitor early data quality

Confirm installs, spend, attribution, trial starts, and subscription events are arriving. Do not make budget decisions from incomplete data.
6

Review payback signal

Once enough data is available, compare CAC, ROAS, trial-to-paid conversion, renewal quality, and payback against the target.
7

Review recommendations

Use the action queue to approve, reject, defer, or manually test recommended budget shifts, creative pauses, or audience changes.
8

Close the campaign review

Document the outcome, next action, and whether the campaign should scale, continue, pause, or become a new test.

Campaign brief template

FieldExample
ObjectiveValidate annual-plan acquisition on TikTok
ChannelTikTok
MarketUS
AudienceBroad interest + lookalike
OfferAnnual plan with trial
Budget$2,000/day initial cap
Primary metric60-day payback projection
Secondary metricsCAC, trial conversion, refund rate
Decision date14 days after launch
OwnerGrowth lead
GuardrailsMax $2,000/day, rollback if CAC rises 20%

What to monitor

PhaseSignals
Day 0-2Spend delivery, attribution mapping, event arrival, obvious tracking gaps
Day 3-7CAC, trial starts, paywall conversion, creative fatigue, early ROAS
Day 7-14Trial-to-paid conversion, cohort size, refund risk, projected payback
Day 14+Renewal quality, payback trend, scale or pause decision

Using FloKit recommendations

FloKit may recommend:
  • Moving budget toward a stronger cohort.
  • Pausing creative with declining conversion or weak payback.
  • Excluding a low-LTV audience.
  • Testing a different offer or paywall emphasis.
  • Changing bids or target CPA where payback supports scaling.
Do not approve recommendations automatically during the first campaign. Review each recommendation against the campaign brief and guardrails.

Decision framework

ResultDecision
Payback is strong and data quality is trustedIncrease budget within guardrails
Conversion is strong but payback is weakReview offer, retention, refunds, and audience quality
CAC is high but renewals are strongExtend observation window before cutting
CAC is high and revenue quality is weakPause, reduce budget, or change targeting
Data is incompleteFix tracking before deciding

Campaign closeout

At the end of the first review window, document:
  • What was launched.
  • What data was trusted.
  • What data was missing.
  • Which recommendations were approved or rejected.
  • Whether the campaign scaled, paused, or became a new test.
  • What guardrails should change before the next campaign.