Subscribe to our newsletter to receive business tips, learn about new funding programmes, join upcoming events, take e-learning courses, and more.
Take the first step to successfully starting and growing your business.
Join for free)
Posted: Tue 9th Jan 2024
5 min read
Before you begin your online advertising efforts with Meta ads, make sure you're clear on your campaign goals.
Most of the time, our clients are looking to attract new customers or drive sales on their site. But this aim is often restricted by a limited budget and a maximum allowable cost for acquiring a customer or an order.
For Meta to "work", it needs to be able to achieve a CPA (cost per action) or ROAS (return on ad spend) at or below the target ROAS or CPA target you set.
Depending on your business, you may have other campaign objectives. These could include driving traffic to your website, getting more visitors to your physical store or encouraging people to sign up your email list.
There are many ways to run ads campaigns that can all have a significant effect on performance:
Ad content (video, images, copy, carousels, collection ads, catalogue ads, instant experience)
Campaign types (Advantage+ Shopping campaigns, conversion, traffic and so on)
Conversion goals (add to cart, purchase)
Audiences (broad or lookalike audiences)
Locations targeted
Landing page of the ad traffic (home page, collection page, product page)
The best possible formula and account structure often look very different across businesses. As a result, the combination with which you decide to run an initial test is unlikely to bring about your most successful campaign.
There will be room for you to optimise and scale significantly.
On new accounts, this is compounded on new accounts by the learning that Meta can do over time.
The algorithm that drives campaign performance starts with no data and so will be at its absolute worst in the beginning. However, that performance will improve quite rapidly over the first few weeks of running.
The more budget you have and testing you do, the more you can adjust your performance to get better results.
VIDEO: Getting started with Meta Business Suite
Learn how to create and manage posts, review comments and messages and understand your results across Facebook and Instagram:
We would always recommend committing enough initial budget to be able to establish a CPA or ROAS, which is reliable when first testing Meta ads.
As a minimum, you should plan for a budget allocation of 20 times your target CPA and spend it over two weeks.
If you're using a target ROAS, calculate your campaign budget as:
Budget = (AOV (average order value) / ROAS) x 20
This ensures that once you spend this budget, you'll have established a reliable CPA that allows you to conclude on the initial test.
Based on the results you get, we recommend you proceed as follows:
You're acquiring orders profitably (a positive CM3, or contribution margin 3).
You can scale the campaigns to start getting the highest possible number of orders while monitoring CPA/ROAS to maintain it at your target.
Start testing variations of your set-up (content, audiences, conversion objectives and so on) to try and further improve performance.
You're able to acquire new customers but not quite profitably.
You work to start testing variations of your set-up (content, audiences, conversion objectives etc.) to try and further improve performance to reach the target CPA.
Keep your advertising spend at a level you're comfortable with, knowing that you're acquiring orders at a loss. Depending on your LTV (lifetime value), this might not be a problem.
In this instance, it seems like the CPA is extremely unlikely to average out at your target.
You should identify the biggest areas that seem to be underperforming.
Benchmark against industry data.
Check platform engagement metrics, CPC (cost per click), CPM (cost per mille) etc.
Investigate your site for aspects that might be underperforming.
Are there large drop-offs on certain pages or steps of your funnel?
Is there poor UX or any broken links?
Only continue spending if you're testing specific aspects of your campaigns that you believe could have a significant impact on performance.
