All answers

The Prevailing
Answers for
Business model:
Business model, pick up to two

How long does AppLovin's learning period last?

The PrevailingPublished Oct 28, 2026

Our answer

Expect about 5 to 7 days of learning on a new AppLovin campaign, longer on a Day 7 goal or under about 10 purchases a day. Change nothing that week. Give it 2 to 4 weeks before judging it.

Confidence
Medium
Strongest proof
3 AppLovin rules
5 practitioners, 8 opinions or claims
Sources
6 sources
16 statements

New here? Take the 1-minute tour

Why this is the answer

3 reasons and our judgment

Most agencies that run AppLovin put learning at about a week. Sweat Pants Agency gives 5 to 7 days. Jetfuel Agency gives 3 to 5 days inside a 7-day no-touch window, and Pilothouse says to leave a new campaign a week, maybe two.

AppLovin publishes no length, but it says what slows learning. Day 7 campaigns take longer to exit learning, it says. It asks for a budget that buys at least 10 purchases a day so the model learns efficiently.

Judging a campaign takes longer than its learning period. Sweat Pants Agency gives a test 2 to 4 weeks because early swings are noise. Common Thread Collective (CTC) argues learning itself takes 3 to 4 weeks.

Our judgment

  • Plan on about 5 to 7 days for a Day 0 campaign funded for about 10 purchases a day. We took the range Sweat Pants Agency gives, bracketed by Jetfuel Agency's shorter phase and Pilothouse's week, maybe two.
  • Expect longer on a Day 7 goal or with fewer purchases. That follows AppLovin's statement on Day 7 campaigns, its budget advice and Jetfuel Agency's view that the model needs about ten conversions a day.
  • Separate a campaign's learning period from the time it takes to judge it. Read a campaign after the first week but judge it at 2 to 4 weeks, a window that also covers CTC's estimate.
  • Hold budget, target and creative still for the first 7 days, and allow 48 hours after any later bid or target change. This draws on Jetfuel Agency's no-touch window and relearning time and Sweat Pants Agency's warning on target changes.

Our judgment, built from Sweat Pants Agency, Jetfuel Agency, Pilothouse, AppLovin and Common Thread Collective: our reading of these sources, not a finding from any one of them.

What the sources say

16 statements from 6 sources
  • AppLovinOct 28, 2026Supports

    AppLovin says CPM volatility in the first few days is normal while its AI learns. In the first 3 days check that pixel events fire cleanly, and always judge performance at the campaign level.

    “CPM volatility in the first few days is normal.”

    Read it on applovin.com

  • AppLovinOct 28, 2026Supports

    AppLovin lets you start at $15 a day but advises a budget that buys at least 10 purchases a day at your typical cost per purchase. Eligible new advertisers who spend $5,000 within 60 days get $5,000 in credit.

    “for the model to learn efficiently, aim for a budget that generates at least 10 purchases a day at your typical CPA, and scale from there as you see results.”

    Read it on support.applovin.com

  • AppLovinOct 28, 2026Supports

    AppLovin says Day 7 campaigns take longer to exit the learning phase, and strong Day 1-2 performance is an early sign one is scaling safely. It says its model shifts spend toward the best-performing creatives.

    “Day 7 campaigns typically take longer to exit the learning phase, since cohorts need time to mature.”

    Read it on support.applovin.com

  • SuperScaleOct 13, 2026Caveat

    SuperScale, which buys mobile game ads, says on AppLovin moving the goal, about 5% at a time, moved spend where the daily cap rarely did, and it gave thin campaigns two to three weeks before judging.

    “Moving the goal did, and we moved it about 5% at a time.”

    Read it on superscale.com

  • Common Thread CollectiveOct 6, 2026Caveat

    CTC relays AppLovin's own 2025 data: new-customer ROAS grew 29% (Universal), 63% (Prospecting) and 78% (Discovery) from day 0 to day 14, so CTC judges new-customer campaigns at day 14.

    “do not judge Prospecting or Discovery campaigns on early ROAS.”

    Read it on commonthreadco.com

  • Common Thread Collective2026-10Disagrees

    CTC, which manages AppLovin ads, says launch before mid-October since learning takes 3-4 weeks and needs 50+ conversions a week, about $500 to $2,000 weekly for most stores. No data given.

    “Brands starting after October 15 will still be in the learning phase during Cyber Week.”

    Read it on commonthreadco.com

  • DTC Newsletter / DTC PodcastJul 24, 2026Supports

    Pilothouse's Jacob says Meta campaigns give a good read in three to five days, but AppLovin may get no sales the first three days and takes off around days four to seven, so leave it a week or two.

    “AppLoven, you might not get any sales the first three days.”

    Read it on podcasters.spotify.com

  • DTC Newsletter / DTC PodcastJul 24, 2026Supports

    Pilothouse, which sells AppLovin management, says the AppLovin pixel is simple to install on Shopify, and once tracking and UTMs check out, a new campaign should be left alone for one to two weeks.

    “Just leave it for a week, maybe even two weeks.”

    Read it on podcasters.spotify.com

  • Sweat Pants Agency2026-07Supports

    Sweat Pants Agency, which runs AppLovin tests, says to give AppLovin two to four weeks to learn, scale in steps without loosening target ROAS, and recheck incrementality at each new spend level.

    “The step most brands skip is re-checking incrementality at each new spend level.”

    Read it on sweatpantsagency.com

  • Sweat Pants Agency2026-07Supports

    Sweat Pants Agency, which sells paid media management, advises two to four weeks before judging an AppLovin test, since early swings are noise, and fixing the window and budget up front.

    “The trap is calling the test in week one.”

    Read it on sweatpantsagency.com

  • AppLovinJun 16, 2026Caveat

    AppLovin says in its own geo holdout tests, campaigns need to be out of the learning phase and delivering steadily before a test starts. It gives no length for the learning phase.

    “campaigns need to be out of the learning phase and delivering steadily before a test starts”

    Read it on applovin.com

  • Jetfuel AgencyJun 4, 2026Supports

    Jetfuel, an agency that audits AppLovin accounts, tests AppLovin with a 14 to 21 day geo holdout against a synthetic control, holding spend steady, and says an inconclusive result is not a negative one.

    “an inconclusive result is not the same as a negative one”

    Read it on jetfuel.agency

  • Jetfuel AgencyJun 4, 2026Supports

    Jetfuel, an agency that runs AppLovin accounts, says leave a campaign alone for seven days, then raise budget 30 to 40% every two or three days; bid changes trigger relearning, budget changes generally don't.

    “Budget changes at that pace generally don't trigger relearning, while bid changes always do”

    Read it on jetfuel.agency

  • Jetfuel AgencyJun 4, 2026Supports

    Jetfuel, an agency that audits AppLovin accounts, says the most common failure on new accounts is changing bids or creative around day three, resetting learning. It says Axon needs about ten conversions a day. No data.

    “Axon is a value-based bidder that needs around ten conversions per day before its lookalike model fully engages”

    Read it on jetfuel.agency

  • Jetfuel AgencyJun 4, 2026Supports

    Jetfuel says AppLovin's Axon is a value bidder with no manual model controls that needs around ten conversions a day before its lookalike model fully engages, and advises launching with Universal audience.

    “Axon is a value-based bidder that needs around ten conversions per day before its lookalike model fully engages”

    Read it on jetfuel.agency

  • Sweat Pants AgencyJun 1, 2026Supports

    Sweat Pants Agency says AppLovin campaigns usually launch on cost per purchase with a target about half your average order value, then move to ROAS, and the first 5 to 7 days are too unstable to judge.

    “Performance in this window is unreliable. Scaling decisions should wait until the algorithm has enough signal to stabilize.”

    Read it on sweatpantsagency.com

In the full answer

For readers who sign in
  • What to do this week3 steps
  • The numbers behind it3 figures from cited sources
  • Where each source standsA chart placing each source on the question
  • The strongest disagreementThe best case against the answer, from a cited source
  • How sure we areWhat its confidence rests on, and what would change the answer

Read next

More in Setup