+1 vote

I have a subscription-driven app that I'd like to run FB ads on. The goal is get a user to download, convert to free trial, and then convert to paid user. I'd categorize the app as casual entertainment.

This will be my first time running FB ads for it (though I have a FB pixel with some website users that used it in the past)

Given bad economic times, I suspect conversion to paid will be lower than in the past and we really can't afford to burn too much money. My total ad spend is around $10k and I really want to maximize the spend and not light it on fire. I'm okay with extremely low volumes to start.

I want to find the safest, most conservative ROAS driven way to get started. We are extremely risk-averse here especially given the bad economy.

I have several ad sets already made, all of which with a source file I can easily modify myself (we don't have to use all of them if this spend isn't high enough to test that many ad sets).

Any advice on how to get started? How much spend to start with? What type of campaign / bid / optimizations to run? How many creatives to test at first? What can I study or what tutorials / step by step guides should I be following?

by (150 points)

2 Answers

+2 votes
The answer and strategy will be a function of your target CPA(ie - you can run more ads/ad sets/tests & test faster with lower CPA).

Broadly though -> IM(H)O, this answer should address the best overall approach for managing low-budget campaigns capital-efficiently(and I think the answer there holds for the most part irrespective of the crisis):
https://quantmar.com/380/Effective-mobile-acquisition-strategy-developer-limited
by (2.3k points)
This answer is super helpful. Thank you Shamanth. Can you walk me through the target CPA point? Obviously, the lower the CPA the better, so wouldn't it be my goal to start that as low as possible? Is that something I can just fill in within FB's ad manager (I assume this is what manual bids are)? I suppose my initial target action would be an install, and eventually with enough volume, it'd be subscriptions.

What would you say is the minimum spend to establish a baseline ROAS? I'm sure that also changes based on CPA and also conversion rates. The more data points of conversion, the more stat sig with fewer results needed to be confident on a ROAS, but would you suggest is typical? Playing it safe, would $100 a day with 3 creatives to start be too low to establish a baseline D7 ROAS?
@totalnoob -> of course. re target CPA I mean that the lower your CPA, the more conversion events and ad sets you can test/run.

in other words, let's say you have a CPA of $20 and a monthly budget of $9k. you can potentially run 3 ad sets/audiences with a daily budget of ~$100 each - and have each of these ad sets hit 5 conversions per day.

however if you had a CPA of ~$10, you could set up 6 ad sets and have them hit 5 conversions per day.

so lower CPA = you can test more ad sets/audiences.
+1 vote

This is a really good / important question. I address this in the recent webinar I presented about advertising strategy during a recession (the relevant portion starts at 24:42)

by (12.4k points)
This is awesome! I just watched the entire thing. I love the "PANIC" in all caps. It was a nice touch and comforting to know I'm not alone in the panic.

I love the idea of starting bids really low and creeping it higher until we reach a targeted ROAS. For Facebook, would this be achieved by using manual bids on an install action if I'm doing this for the first time? Do you have thoughts on how much I should put as a manual bid on FB or how to calculate that? I assume I don't want to go too low otherwise it'd never get shown? And how much minimum spend per day do I need to start getting some accurate numbers on ROAS/if I have the appropriate bid