How to Optimize a Facebook Ads Account

Daily, in Facebook Ads Rockstars, we get questions related to the correct way of optimizing Facebook ads accounts.

What KPIs should be looked at and how often? When can we start to scale a campaign? When should we stop a campaign/adset/ad?

Overall, the process of account optimization is rather simple if you have a proven methodical algorithm of doing it. Otherwise, it could be a complete mess…

In this article, I will be reviewing my standard day as a Facebook account manager.

1. Check the nighttime data.

Once I’m back at my working place, the first thing I do is check how my accounts’ KPIs have changed overnight.

For that, I need to know what KPIs are vital for this particular campaign and what goals we have agreed on with my client. You always want to have your major KPIs on the same screen so you do not have to scroll for them.

The KPIs to always check first thing in the morning for me are:

– ROAS – of course! this is the conversion rate of your campaigns and is almost always the main indicator of the accounts good/poor performance (the level of ROAS always depends on the margins you are working with, but, of course, you always want it to be as high as possible);

– Amount spent – very often, clients want me to spend a certain amount daily/monthly, so I need to make sure that I have enough of active adsets to reach this goal (if I spent less than needed yet overwent the minimum required ROAS, I can scale up today to increase the daily spend);

– Cost per result – sometimes, ROAS is not as important for the client as CPR, especially for non-ecommerce products. They might get the most sales on their backend and are not interested in high ROAS on their TOF.

– Unique CTR and CPC (cost per link click) – click-through rate and cost per click tell you if the ads are still performing on the same level as they are used to (I usually want to have min CTR >= 3% and CPC=< $1)

– If the results are not good – consider checking such KPIs as cost per add to cart and cost per initiate checkout to find where exactly in your funnel there is a performance drop.

2. Start turning on/off adsets/campaigns.

Okay, we got this! The KPis have been checked. What do we do next?

First thing in the morning, we want to turn on/off all the adsets/campaigns that should run throughout the rest of the day.

– Check historical performance of active adsets – filter out all adsets except for active ones and check their performance lately.

I always do the review of active adsets based on three time slots – last 30 days, last 7 days, and last 3 days (plus today).

If my main KPI is ROAS (which is the case for most accounts), I sort all the accounts by ROAS and start with the longest time-slot – 30 days.

Say my ROAS for this account should be at least 2.0. Then, I check ALL adsets below 2.0 for the last 30 days and filter out the rest. After this, I check if these adsets have been working fine for the last 7 and 3 days.

If yes, I keep them running for now. If not, I turn them off.

– Check historical performance of inactive adsets – often, your data gets updated on Facebook with a delay, and your task is to identify and turn back on these “late bloomers”.

The process is similar – filter out everything except inactive adsets and sort them by ROAS.

Use the same time slots – 30, 7, and 3 days (plus today).

Say you have found 12 adsets that are inactive, have ROAS > 2.0 but are turned off for whatever reason.

Check their performance for the last 7 and 3 days. Have they been running at all? No? Turn them back on.

Have they been running in the last 7 or 3 days and still have high ROAS? Turn them back on.

Otherwise, leave them off and move forward.

3. Check for updates from your client.

Often, clients will be giving you updates with new content for creatives, new promotion campaigns, ask you for updates/to jump on a call with them.

– If you get the updates early in the morning – immediately create new ads/campaigns/adsets or whatever else your client wants to start testing, as you want them running all day long.

– If you get updates in the middle of the day – take your time, create the required tests, but schedule them all to start the next morning from 8 am.

– If you get updates in the morning, but they are huge – create all the required tests, but schedule them for tomorrow, to not rush the process.

– If the client wants a call – make sure that you have finished parts 1 and 2 before arranging the call.

4. Start optimizing. 

So you are an account manager and you need to check its performance regularly. I check every account at least once/every 2 hours.

During this time, I:

– Turn off poorly-performingadsets

A popular question is “when should I stop a poorly-performing adset?”

Well, that always depends on the product you sell.

Of course, when optimizing, I sort adsets by “Amount spent”.However, for me, I always check 2 things:

1. Have I spent more than 1/2 of the daily budget? If yes, and ROAS is not good enough, I stop this adset for today.

2. Do any of the ads in this adset have high enough ROAS? If yes, I turn off poorly performing ads and let the adset try to catch up with the winning ad.

– Scale up winners

For the winning adsets that have spent more than 1/2 of the budget and have ROAS>my goal, I scale them using the following methods:

1. Increase the budget by 30% – for adsets that have a limited audience size/high ads frequency (for cold audiences – <1m in size, >1.5 in ads frequency).

2. Duplicate the adset – for all adsets that have large enough audiences, hence still have a lot of space for scaling.

– Check KPIs

Of course! Never stop checking your KPIs! You want to predict how your campaigns will behave, and numbers will always tell you whether you have a positive or negative tendency.

3. Scale your CBOs – more information in the article we have recently published – click here.

5. If you stopped many adsets during the day, create new tests for tomorrow (compensation).

Most of our clients are not interested in spending $300-500/day. They come to use and want to go for 20, 100, or even 500k monthly! So, we cannot afford letting our daily spend go below the required margin.

Say I managed to scale up by 1k, while turning off adsets with the 3k daily spend in total. This is where I start thinking over the strategies of where I can increase the spend.

– I start testing new regions

Say I have a number of proven audiences for the US. I try to implement the same audiences in the United Kingdom, Canada, and other English-speaking countries.

If the product is one needed everyone, in large amount, I can even go worldwide.

– I test lookalikes from custom audiences

Some very good audiences you can create are the LLAs from your 95%-video viewers, users engaged with your FB page, your website traffic, your email subscribers, and, of course, LTV customers (for different regions).

– I test new creatives

You know that you target the proven audience, but your ads frequency is over the limit already. Make new ads – use same texts and headlines, but tweak the thumbnail/content and see how it will perform tomorrow.

– I test different placements

I always start running campaigns with automatic placement. Then, I check the statistics and see which mediums work the best.

Say you have 100 conversions for the campaign, 85 of which come from Instagram. Try to create an adsets where you would target clients through Instagram only.

– I test manual bidding 

This is a very complex matter and needs a whole lotta new post to talk about. But to start with, you have your product’s target cost – try to use manual bidding with the same cost and see if you can win good clients on the auction with this bid.

6. Before leaving the account for night.

Personally, I can’t just go and let the account run by itself. I worked SO HARD all day long to make it work! So, I:

– Optimize the adsets one last time before tomorrow to check if all my winners are running;

– Turn on my automated rules – for minimum ROAS for adsets, for minimum ROAS for ads, for scaling, etc;

– Leave running ONLY the winners for today – the rest can be restarted tomorrow based on the 30, 7, 3-day time slot checks.

7. Repeat next day.

Well, technically, you rarely repeat everything in the same manner! I mean, every campaign is unique, every test is unique, every audience behaves in a unique manner. BUT that’s exactly what I love about Facebook ads management – I learn something new every day!

To wrap it up!

Overall, that’s the life of a Facebook account manager – turn on/off, optimize, repeat next day. Sounds simple, is NOT simple at all! But if you have a proven methodical approach towards optimization, there ain’t no mountain high enough for you!

What do you think? Is your ads management day similar to mine? Yes? Well, then, you are no the right track to success! No? Tell me how you approach optimizing. We live and learn, and I am sure you can teach me whole lotta stuff as well! So feel free to post your ideas/questions/suggestions below.

Let’s share our ideas and boost our sales together! United we scale!

Every month I choose a few entrepreneurs to personally help them scale their Facebook ads and businesses.

Watch the free training and apply here: https://www.ecommercescalingsecrets.com/home

?