Goal first, numbers second
Every Instagram ad campaign has one objective: leads, sales, website traffic or awareness. Meta's algorithm looks for people who fit the objective you choose. If you want leads but set the campaign up for likes, the algorithm will bring you people who like but never message.
Before you split the budget, answer one question: what are we measuring this month? If the answer is one word, you're on the right track.
Fewer campaigns, more creatives
Meta's algorithm puts every ad set through a learning phase. To exit it, an ad set needs roughly 50 results (leads, for example) in a week. Split a small budget across five ad sets and none of them reaches that threshold; they all get stuck in learning.
The takeaway: on a small budget, cut the number of campaigns and increase the number of creatives. Testing 3-5 different visuals and texts inside one ad set tells you more than five separate campaigns.
Cold and warm audiences
A cold audience is people who don't know you yet. A warm audience is people who have already come across your page: those who watched your video, visited your profile, or came to your site and left.
The warm audience is small, but the cost per lead there is usually several times lower. In practice, putting 15-25% of the budget into retargeting, meaning the warm audience, works well.
Example split: 600 AZN a month
Here's a starting split we use often. Main campaign, cold audience: 330 AZN (55%). Testing new creatives and audiences: 150 AZN (25%). Retargeting, warm audience: 120 AZN (20%).
After the first two weeks, we look at the numbers. The creative that wins the test moves into the main campaign, and whatever isn't working is switched off. This isn't a law, it's a starting point: it varies by industry.
When to scale up, when to stop
The signal to increase the budget: your cost per lead has held steady for at least a week and the leads are turning into sales. Don't make sharp jumps when you scale. Doubling the budget at once can push the algorithm back into the learning phase, so raising it by about 20% at a time is safer.
The signal to stop: your cost per lead is twice your target and hasn't come down in three or four days. In that case the problem isn't the budget, it's the creative or the offer. Adding more budget won't fix it.
