In this article you'll learn about:

There have been a lot of recent Facebook optimization changes that Preflect has been detecting over the past several months that have become significant enough to warrant an adjustment to our recommended best practices.

The main Facebook change that we’ve been monitoring is the speed at which Facebook determines how to allocate your ad spend at the ad level. The time it takes Facebook to make budget allocation decisions has greatly reduced and requires far less performance data.

This is not surprising considering that it is a commonly known fact that Facebook can no longer accurately track conversion and sales data, however this leads us to believe that Facebook now places greater importance on their own analysis of your images, videos, and ad texts to estimate the likelihood of your ad performing well.

This shift in how Facebook allocates ad spend is a great thing for merchants. Facebook has lots of data on what their users like and don’t like and can make accurate predictions most of the time. Making faster decisions about where your ad dollars should go also means that less ad spend is wasted testing new ads for multiple weeks at a time.

Generally speaking, you can spend as little or as much as you like. It all depends on what you can afford to spend, how much you’re comfortable with spending, and what your sales goals are.

Setting reasonable budgets also depends on your Average Order Value.

To run your ads most proficiently, we recommend that users set their budget to 2x-5x their AOV to produce better and faster results.

At a minimum:

For stores with AOV between $0 and $250, spend no less than $50/day.

For stores with AOV that is $250+, spend no less than $100/day.

When you’re happy with your ROAS or when you’re seeing your ROAS increase over time.

Once you know that your ROAS is climbing and your ads are optimizing, it’s time to scale your budget! Keeping it small and steady is key. You want to increase your budget by 10-20% every 4-7 days when results are increasingly good. Making budget changes too drastic or too frequently, will reset the learning phase and severely impact performance. This is why we recommend increasing your daily budget by ≤20% every Friday.

When you have ads that have been running for less than 7 days, it is unwise to make budget adjustments as this resets your ads’ learning phase.

When you’re spending more than $50/day (or $100/day if your AOV is $250+) and you’re not happy with your ROAS, you should reduce your budget. Smaller budgets are spent more efficiently by Facebook in most cases. Reducing your budget increases your ROAS in the short-term.

As stated in the overview to this guide, Facebook changes that have been occurring over the past months have made the answer to this question much simpler.

Facebook now tests ads for a much shorter period of time than in the past, meaning that you can now run a greater volume of ads. This means that there is no true right or wrong answer to this question. However, this is still tied into your daily budget, so the more you spend - the more ads you should be running. We always recommend following Preflect’s recommendation at the top of your ads page. Due to recent changes in Facebook’s optimization process, the maximum number of ads you are able to run is 50 ads.

Preflect runs your Facebook, Instagram, and Google ads via API, not in a traditional ad account like you’re used to, so you won’t see the Meta ads in your FB account. This is completely normal. We create a Preflect ad account under our business manager so that's why you don't see it in yours. We do this to help you navigate the ads manager headaches, security & to free up your time to focus on creatives!

By default, Preflect’s AI groups your ads into just two ads sets, one for prospecting and one for retargeting/remarketing.

Anytime that you add, edit, adjust ads or change your daily budget, this will reset the learning phase of these ads. This means that you will experience a drop in performance for a short period of time following a new change. This is a necessary sacrifice for running cost-efficient and fast tests.

This structure is fantastic for testing ads quickly without wasting ad dollars. Because you test new ads with ads that are proven to work, you give Facebook’s AI the opportunity to see if a new ad will beat some of your best ads. If Facebook decides that a new ad you are testing is not as good as your best existing ads, it will decrease ad spend dramatically on the newly tested ad and instead spend more on the ads that are proven to work. This process happens quickly and prevents ad dollars from being wasted because Facebook automatically moves the budget back into your best ads.

To ensure that you don’t have to dig through tons of confusing Facebook data to identify which ads Facebook is currently being favorited in your account, Preflect has made it very easy for you by introducing our individual ad suggestions to either keep or replace the ad.

Ads which are suggested to be kept might not be ads with the highest ROAS or the lowest CPC, but they will always be able to drive the most revenue, earn you the most clicks, the most traffic, and the most sales. Facebook will always push ads that they predict will do the best, even before analyzing that much real performance data.

That type of testing is bad for two reasons:

On top of that, you’ll have no clue how long to wait before you decide that an ad is bad enough to shut off. By testing winning ads with new ads, Facebook will determine when an ad has been tested long enough to determine if it’s a winner or a loser, and swiftly move your budget.

Making frequent changes is bad for performance because each major edit will reset your ads learning phase and this will severely impact optimization!

Major edits are:

You should make edits when:

To conclude, Preflect users who see the best results make edits once every 7-10 days. This allows your ads to optimize and collect data.

Archiving an ad deletes the ad inside of Facebook. When you want to make a major change to an ad and you want to preserve the performance data of the original ad, archive the original ad, duplicate it, and then make edits to the duplicate. Now, you’ll have performance data for the new version and the old (archived) version so you can compare the two versions.

Pausing an ad does not delete the ad inside of Facebook. It just shuts it off.

Pause an ad when you intend to run it at a later time “as is” without making any changes.

Do this only if an ad was launched within 24 hours. You can edit it to fix typos or errors. We don’t recommend editing an ad that has been running live for multiple days as you will not have the ability to go back and review performance before and after making the change.

If your ROAS (Return on Ad Spend = revenue ➗ ad spend) is low, there could be a variety of reasons. If you are following Preflect’s Getting Started Guide and all best practices for creating and editing ads, then it is time to look at some other metrics such as: website conversion, CTR scores, average order value (AOV), etc. Follow this link to find out how to improve in these specific areas and more: My ROAS (Return on Ad Spend) is Low

First sort by the “Highest ROAS” so you can see which ads are performing the best. Next, ask yourself: what is special about this ad?

Now, create more ads that look very similar, but not identical to your winning ad.

Next, archive ALL ads that Preflect recommends to replace.

Lastly, upload and deploy all of your new creatives that look very similar, but not identical to your winning ad.

Look at the individual data of your ads and determine the best course of action.

If your weakest metric is Click Through Rate: you need to test new, different creatives. These creatives should be a different theme and different style than the ads that you tested before that did not perform well. Do not be afraid to make radically different ads in the spirit of testing.

If you are receiving a high CTR but low ROAS: you need to re-evaluate your web design and merchandising strategy. This is something that is more complex than simply changing creatives. It is also entirely out of the scope of the work that Preflect does, so if you’re having issues with your Store Conversion Rate, we recommend contacting a web design expert who can better assist you.

Ad creatives, just like art, can be widely subjective in the ideation and creation process. However, with Facebook ad creatives, beauty is not in the eye of the beholder, but instead is in the eye of Facebook and Instagram users! For advice on how to improve your ad creatives, click here!

Your ad copy supplements your ad creative. This means that many customers think about them as interconnected. To learn about what makes a good or bad ad copy, click here.

Sometimes an ad works great with one creative paired with one version of ad copy, but that does not necessarily mean that the creative will perform well with other ad copy variants. This is why testing is important. If you want to learn how to test, click here!

Facebook and Instagram ads can appear in either regular feed format (square), carousel feed format (multiple square images), or story format (vertical rectangle).

The ideal dimensions for feed/square format ads are: 1080 x 1080 px (1:1)

The ideal dimensions for story/vertical format ads are: 1080 x 1920 px (9:16)

For more in-depth information on formatting within Preflect, click here!