Solved: Understanding Meta Ads Funnel Placement in Andromeda

Hi there,

Thanks for reaching out!

That's a great question, and one that gets asked a lot. This whole 'new way' of doing Meta ads is causing a fair bit of confusion. I'm happy to give you some of my thoughts on it, cos tbh, I think a lot of the advice out there is a bit misguided and oversimplifies things to a dangerous degree. It's not as simple as just throwing a bunch of creatives into one campaign and hoping for the best.

Let's be honest about Meta's "magic" algorithm...

First off, let's talk about this idea that Meta's algorithm (Andromeda or whatever they're calling it this week) can now perfectly figure out your entire sales funnel just from looking at your creatives. It's a nice idea, but in my experience, it's a load of codswallop.

When you put everything into one CPO campaign, you're essentially telling Meta, "Here's my money, go find me some sales." The algorithm, being a machine that follows instructions literally, will go for the lowest-hanging fruit. It will find the people who were already close to buying and show them your ads. This looks great on the surface – you get some sales! But what's actually happening? It's neglecting the crucial work of filling the top of your funnel. It's not building an audience of new potential customers because that's harder and more expensive work. So you get a short-term boost, but a month or two down the line, your performance dries up because you haven't been prospecting.

I see this all the time. An account looks like it's doing well, then suddenly performance falls off a cliff. The agency or marketer is left scratching their head, but the reason is simple: they stopped feeding the top of the funnel because they handed over all control to the algorithm.

Think about what happens when you set a campaign objective to "Reach" or "Brand Awareness." You are giving the algorithm a specific command: "Find me the largest number of people for the lowest possible price." The algorithm does exactly that. It finds users who are cheap to reach because they're not in demand—they don't click, they don't engage, and they certainly don't buy. You are literally paying Facebook to find you non-customers. The simplified CPO structure can have a similar effect if you're not careful. It'll prioritise showing ads to your existing warm audience (retargeting) because they're cheaper to convert, and ignore the vital, more expensive work of finding new people.

Your concern about removing an ad and killing its "connection" to other ads is understandable, but it's based on a misunderstanding of how it works. There's no magical synergy between creatives. The algorithm is simply testing everything all the time. If one ad is getting all the spend and delivering results, it's because it's working. If another ad is getting no spend and has a terrible CPA, it's a dud. Turning it off isn't going to break the good one. It's just good housekeeping. You should be ruthless about this. An ad that has spent 2-3 times your target CPA without a conversion needs to be switched off. End of story. Don't let fear of a non-existant "connection" stop you from making sound financial decisions.

I'd say you need to take back control with a proper funnel structure...

So if the 'one campaign to rule them all' approach is flawed, what's the alternative? It's not about going back to the old days of dozens of complicated campaigns. It's about a simple, logical structure that respects the sales funnel and gives the algorithm clear directions. You need to be the strategist, not just a passenger.

I almost always recomend a structure based on temperature: Cold, Warm, and Hot audiences. This usually means three core campaigns, each with a distinct job.

This structure works because it's logical. You're telling Meta exactly who to go after in each campaign. The Prospecting campaign's job is to find new people. The Retargeting campaign's job is to convert people who already know you. The Retention campaign is to get more value from people who have already bought from you. There's no ambiguity. The algorithm can then get to work optimising within those clear boundaries, rather than trying to guess what you want it to do across your entire audience pool.

You mentioned using one ad set. In this structure, I'd still advocate for multiple ad sets within your Prospecting campaign. Why? Because you need to be testing audiences. You'd have one ad set for your best Lookalike, another for a different Lookalike, and maybe one or two for your best interest-based audiences. This allows you to clearly see which cold audience is performing best and allocate budget accordingly, either manually or with Campaign Budget Optimisation (CBO). Lumping them all into one ad set just muddies the water.

You'll need to think about your customer's nightmare, not their demographic...

When it comes to picking those ToFu audiences, especially with detailed targeting, you need to go deeper than surface-level demographics. "Women aged 25-45 who like yoga" is not an audience strategy. It's a cliché.

You have to define your customer by their pain. Their specific, urgent, expensive nightmare. What keeps them up at night? For one of our B2B SaaS clients in the recruitment space, their target audience wasn't just "HR Managers." It was "Heads of Talent at medical facilities who are terrified of losing their best nurses to competitors offering better shifts." The nightmare wasn't 'needing recruitment software'; it was 'running a ward understaffed and facing a crisis.' We helped them reduce their Cost Per User Acquisition from a crippling £100 down to just £7 by focusing on this specific pain point in our ads and targeting.

Once you've identified that nightmare, you can find them. What podcasts do they listen to? What industry newsletters do they read? What software do they already use? Who are the influencers they follow? These are your interests. This is how you find your people. Targeting "small business owners" is useless. Targeting "users of Xero accounting software" who also follow "Shopify" and are admins of a "Facebook Business Page" is getting somewhere. You're targeting a problem state, not a person.

I'd prioritise your audiences like this:

1. Hot (BoFu): People who have added to cart, initiated checkout. They are ready to buy. Retarget them hard and fast.
2. Warm (MoFu): Website visitors, video viewers, social media engagers. They know you but need a nudge.
3. Warmish (ToFu): Lookalikes of your best customers. These are statistically the most likely cold prospects to convert. Start with a 1% Lookalike of your purchasers list.
4. Cold (ToFu): Interest and behaviour-based audiences built around the 'nightmare' scenario.

Test them in that order. Don't even think about broad targeting until your pixel has thousands of purchases on it and you're spending a serious amount of money each month.

You probably should calculate how much a customer is actually worth...

This all brings us to the most important question: how do you know if an ad is "underperforming"? The answer has nothing to do with what Meta's dashboard says is a "good" CPA. It has everything to do with your own business maths.

You need to know your Customer Lifetime Value (LTV). If you don't know this number, you're flying blind. The question isn't "How low can my CPA go?" but "How high a CPA can I afford to acquire a great customer?"

Let's break it down. You need three numbers:

• Average Revenue Per Account (ARPA): What's the average a customer spends with you per month (or year)?
• Gross Margin %: What's your profit margin on that revenue?
• Monthly Churn Rate %: What percentage of customers do you lose each month?

The calculation is simple: LTV = (ARPA * Gross Margin %) / Monthly Churn Rate

So, if a customer pays you £100 a month, your gross margin is 70%, and you lose 5% of your customers each month, your LTV is (£100 * 0.70) / 0.05 = £1,400. Each customer is worth £1,400 in gross profit to you.

A healthy business model aims for at least a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. So, with an LTV of £1,400, you can afford to spend up to £466 to acquire a single customer and still have a very healthy business. Suddenly that £50 CPA from your prospecting campaign doesn't look so bad, does it? It looks like a bargain.

This is the maths that lets you scale aggressively. When you know you can afford to spend £466 to get a customer, you can outbid competitors who are trying to get their CPA down to £20. You're playing a different game. Use the calculator below to figure out your numbers.

We'll need to look at your creative and messaging...

Your point about creative being more important now is absolutly correct. But it's not about just having a diverse batch of ads. It's about having the right message for the right person at the right time. This is where our funnel structure comes back into play.

ToFu (Prospecting) Creative: These people don't know you. You can't just shove a product in their face and say "buy now". Your ad needs to earn their attention. You do this by talking about their 'nightmare'. Use the Problem-Agitate-Solve framework.
"Are your cash flow projections just a guess? Worried one bad month could mean you can't make payroll? Stop guessing. We give you a clear financial dashboard so you can grow confidently."
See? We haven't even mentioned the service yet, we've just described their pain perfectly.

MoFu/BoFu (Retargeting) Creative: These people know you. They've been to your site. Now is the time to build trust and overcome objections. Use social proof. Testimonials, case studies, reviews, user-generated content (UGC). For an eCommerce store, this is where you show a carousel of your best reviews or a video of a happy customer unboxing their order. For a SaaS business, a short video case study showing how you achieved a 10x ROAS for a client like them. For one of our eCommerce clients selling cleaning products, we used testimonial-heavy ads in retargeting and saw a 633% return and a 190% increase in revenue.

Here's a quick cheat sheet for how to think about it:

When you have this structured approach, your 'diverse batch of creatives' suddenly has a purpose. You're not just throwing spaghetti at the wall to see what sticks; you're having a logical conversation with your customers as they move through the buying journey.

This is the main advice I have for you:

To wrap this all up, here are the main recommendations I'd make based on your question. This is the process we follow for our clients, from software companies where we've driven over 5,000 trial signups, to eCommerce brands where we've achieved a 1000% return on ad spend.

As you can probably tell, there's a lot more to running Meta ads effectively than the latest guru playbook suggests. It's about combining a deep understanding of marketing fundamentals (like the sales funnel) with a technical mastery of the platform. Getting this structure right is the difference between an account that burns cash and one that becomes a predictable, scalable engine for growth.

It can be a lot to implement, and making a mistake can be costly. If you get the structure wrong, you can end up with audience overlap that drives up your costs, or you might misinterpret the data and turn off a potentially winning campaign too early. This is often where getting some expert help can make a huge difference.

We offer a completely free, no-obligation initial consultation where we can go through your ad account together. I can show you exactly how we'd apply this structure to your specific business and what kind of opportunities you might be missing. It's often a real eye-opener for people and gives them a clear, actionable plan to move forward with, whether they decide to work with us or not.

Hope this helps!

Regards,

Team @ Lukas Holschuh