+1 vote
This is a two-part question mainly revolving around Facebook app install advertising.

The first being, how does the size of your daily budget and corresponding scale affect the CPIs your able to achieve and sustain within a campaign? For example,

Say your able to naturally achieve a CPI for your mobile game of $0.50 from a Broad - Tier 1 - Lowest Cost campaign on Facebook Ads. If your achieving this CPI within a daily budget of $200/day and therefore ~400 installs/day, what change can you expect to occur from running the same ad but with a daily budget of $2,000/day instead? Will you continue to sustain $0.50 CPI and obtain 4,000 installs/day or will the CPIs rise after a certain scale, if so, by how much?

Part two of this question is basically a similar scenario only:

You are still achieving $0.50 CPI for your mobile game within a Broad - Tier 1 - Lowest Cost Facebook ad campaign running at $200/day daily budget.

But now, you determine your LTV for this game is $0.35 so you need to acquire users for maximum $0.30 CPI.

So the question is, if you change your Facebook ad campaign with $0.50 CPI inside a "Lowest Cost" bid type and change it to a Lowest Cost - Bid Cap of $0.30 limit, what affect will this have on your campaign?
Mainly, how many installs will you likely be able to acquire within your $200/day budget? If you increase this budget to $2,000/day will that help you acquire more users at your max bid of $0.30 CPI?

I understand there is a lot of variables not accounted for and this question may be hard to answer without "it depends" but any insight based off your experience would be greatly appreciated!
by (350 points)

2 Answers

+1 vote

The short answer is that it really depends on your eCPM and that will dictate the CPI you are able to sustain.  However I should mention CPIs rise not because they just magically do, it's because there's no low cost inventory left and you have to increase your bid and hence your CPI to continue to scale.

There's a bit of preface you have to understand in order for this to make sense.

What are you really doing on Facebook?

Think about what you're really doing on Facebook. At the end of the day you're really just paying for the ad space on Facebook to show your ads. Also bear in mind that thousands of other companies are paying to get access to that ad space too. To figure out who gets shown which ads, Facebook turns it into an automated auction for that ad space.

Why your eCPM dictates your CPI

However Facebook has a problem. You might be doing a CPI campaign, but there are other consumer brands that are doing CPC and CPM campaigns. Afterall a chocolate bar brand doesn't care about CPIs. To solve this, Facebook unifies all the types of campaigns into a eCPM and uses the eCPM to auction out their ad space. Remember at the end of the day you're really just paying for ad space and ad space at its most simpliest form is sold for the impressions hence why it uses an eCPM. For a CPI campaign your eCPM is CTRxCVRxBidx1000.

Your eCPM is what you're really using to bid for ad space in Facebook (and really just about any other network). The higher your eCPM the more likely you are to win that ad spot and your ad is shown there.

 

So lets see how this plays out in your two scenarios:

Scenario 1: $0.50 bid with $200 daily budget to $2000 daily budget

The answer is you will maintain a $0.50 CPI as long as your eCPM stays competitive enough to keep winning the ad auction.  However what's likely to happen is that the broader targeting likely means your CTR and CVR will decrease as you scale.

Meaning  with CTRxCVRxBidx1000=eCPM your overall eCPM will decrease, and you'll see that your volume does not scale linearly. 

If you want to hit your daily budget so that you can continue to scale your options are: increase bid, create better ads (increasing CTR) or create a better appstore page(increases CVR). Then you you'll see eCPM rise and be able to spend to your cap.

 

Scenario 2: $0.50 bid with $200 daily budget to $2000 daily budget. Also $0.50 bid to $0.30

Same situation as Scenario 1 except you get there faster. This is because CTRxCVRxBidx1000= eCPM with bid going down so does your eCPM. You just win less ad space because your eCPM is no longer as competitive.

 

So how does this play out in your two scenarios? Well for one thing your CPI's will largely be unchanged but your volume will dry out with time.

by (290 points)
edited by
+1 vote

With Facebook's algorithm (or any advertising platform's, really), you have to understand that getting paid is the first priority: the algorithm will surface ads that produce revenue and fill inventory on that basis. Some background on Facebook's ad algorithm here: How does Facebook’s advertising targeting algorithm work?

So with that out of the way, you can think about all of the parameters of an ad set that you described as playing a role in Facebook's algorithm coming up with an expected value (to Facebook, not you, the advertiser) for your ads, but you're missing a critical piece: conversion rate. If you're not buying on CPM, Facebook only gets to charge you when some objective happens (an install, an in-app event, etc.), so they use that to determine how frequently your ad gets placed and to whom. If no one clicks on your ad and converts, Facebook won't fill inventory with it, because they won't get paid.

Everything else is kind of secondary: daily budget, target cost, etc. You can imagine different scenarios that showcase this:

image

All this is simply to make the point that the best thing you can do to improve delivery is to focus on conversion: Facebook will favor your ads if they get paid as a result of filling their impressions with them. You can move the needle on delivery much more effectively by improving conversion than by increasing bids.

by (6.2k points)