1.) Use The Same Ads Per Lookalike.
When you launch multiple lookalikes into testing make sure each ad set has the exact same number of ads and that they each have the same ads.
This ensure during testing that results don't get skewed because you are using different ads per each ad set.
Having the same ads in each ad set gives allows you to identify which lookalike audience is performing best and which ad performs best per each lookalike.
2.) Define Your Golden Metric.
If you want to increase your sales online, optimize for purchases.
If you want to increase the volume of leads, optimize for leads.
It's important to optimize for the exact conversion event you want because this allows the algorithm to help get more of those events that you've chosen.
This one should be obvious but so many people select link clicks, landing page views and add to carts and they never get as many leads or purchases as they want.
3.) Increase Your Daily Spend On Winners.
If you have an audience lookalike that is generating results, increase the budget on it. Don't increase your budget on underperforming lookalikes just cut them by turning them off.
If you are using campaign budget optimization (CBO) then you should turn off the ad sets that don't generate results and the CBO will automatically shift the larger portion of spend into the ones that do get results.
This is going to help you lower your cost per result, increase the frequency of your results and add more consistency to your results.
4.) When Creating Lookalikes At Least Have 100+ Audience Size.
When you create a lookalike audience you're going to want more than 100+ people in the custom audience otherwise you are not going to have enough data for the algorithm to build an accurate enough audience for you.
Your custom audience must contain at least 100 people from a single country in order to use it as the basis for a Lookalike Audience.
Custom audience source quality matters too. For example, if a custom audience is made up of your best customers rather than all your customers, that could lead to better results.
The larger the custom audience based on the specific best action they took the more accurate the lookalike will be and the better the results you'll get.
5.) Start With 1% First Then Move Up.
Creating a lookalike 1% audience is best to keep the audience accurate.
If you are scaling results but hitting all wall on diminished returns with your 1% lookalike then start expanding it's size by making it a 2% lookalike and keep going up from there if the results are great and you are scaling with ease.
6.) Exclude The Custom Audience
When you create your lookalike exclude the custom audience you created it with. This will ensure you don't target the same people that you've already reached.
Otherwise if you don't exclude it it will raise your retargeting audience cost and lower the your first time impression ratio. Resulting in less reach, fewer conversions, higher costs, bidding against yourself and a downward spiral over the life of the audience.
So always exclude the custom audience. But you'll want to make sure you have a retargeting campaign otherwise you'll be missing out on a lot of low cost conversions.
7.) Advanced Lookalike Opportunities.
This one is more of a fun fact than anything...
You can create up to 500 Lookalike Audience variations from a single source audience.
Whether you create a 1%, 2%, 3%, 4%, or 10% or do a 1-5%, 6%-10% the variations seem unlimited.
You can do it based on frequency of how often a user takes the action, what device they are using, how long ago they converted, the time period you set for the audience, etc. This list goes on awhile.
It's important to identify what metrics are relevant to your business so you can decide what lookalike variations are worth testing.
I recommend starting with a basic 1% lookalike and if it performs well then start testing more variations of that specific source audience to create the lookalike. That way if you do more testing it's far more likely to get results.
8.) Stacking Lookalikes.
You should only stack lookalike audiences once you've found lookalike audience that are proven and you are ready to scale.
This ensures you build up an audience of lookalikes that have shown to get great results.
If you stack lookalikes before you know if they work it's likely one of them will underperform which can drag down your audience. The worst part about that is not knowing which audience is causing the poor result.
So only stack proven lookalikes in an ad set. This will ensure you can scale results and it also helps with exiting learning phase since you have more budget going towards one large audience allowing you to get more conversions.
9.) Value Based Lookalikes.
If you want to attract more customers similar to the customers that spend the most money with your business.
Than you'll want to create a value-based lookalike audience by uploading an audience with the customers and the total number of dollars they've spent with you're business over their lifetime.
That lookalike audience will then be made up of the people most similar to your highest value customers. Allowing you to acquire the customers who spend more money than your average customer with at your business.
A regular lookalike audience (one that doesn't take customer lifetime value into account) can only find people similar to all your customers.
10.) Lookalike Country Targeting.
If you are wanting to target multiple countries it's best to do an entire area rather than piecing out each individual country.
For example, targeting one ad campaign in, 'Asia', can result in higher performance than running individual campaigns for each country in Asia.
In most other cases, it's best to keep one country per lookalike by having one audience for the US and another one for Australia.
This will ensure you keep your data separated and the algorithm can learn faster to optimize and generate you more results.
If you follow just some of these tips with lookalikes you'll be able to generate more results, lower costs, and scale faster...
Want to get consistent Facebook Ad Results...
Schedule a 1-on-1 coaching session with Chase here.
Or...
Get started with the Facebook Ads Expert Mastery Class.
To your success.
- Chase
This site is not part of the Facebook website or Facebook Inc. as well Additionally, this site is NOT endorsed by Facebook in any way. FACEBOOK is a trademark of FACEBOOK, Inc. This site is not part of the TikTok website or TikTok Inc. or ByteDance as well Additionally, this site is NOT endorsed by TikTok in any way. TIKTOK is a trademark of TIKTOK, Inc.