Solved: Engagement Campaign ( WhatsApp), offline conversions?

Hi there,

Thanks for getting in touch. I saw your post and it sounds like a classic, and very frustrating, problem. Getting loads of chats but none of them turning into actual business is a real drain on time and money. I'm happy to give you some of my initial thoughts and guidance based on what we've seen with other clients who've faced similar issues. It's a solvable problem, but it requires a bit of a shift in strategy away from what you're currently doing.

We'll need to look at your campaign objective first...

Okay, the very first thing that jumps out at me is you're running an 'Engagement' campaign. On the surface, it makes sense – you want people to engage with you on WhatsApp. But here's the thing about Meta's algorithm: it's incredibly literal. It will do exactly what you tell it to do, and nothing more.

When you choose 'Engagement' as your objective, you're telling Meta "find me people who are most likely to click, like, comment, or start a message thread". These are often called 'clicky' people. They're cheap to reach because they engage with lots of content, but they are very rarely the people who are actually serious about making a purchase, especially for a high-value or 'affluent' market. The algorithm isn't looking for people with buying intent; it's just looking for people who will perform the engagement action for the lowest possible cost. This is almost certainly the main reason you're getting so many bad leads. You're fishing in a pond full of people who like to chat, not a pond full of people who are ready to buy.

The fundamental fix here is to change your campaign objective. You should be optimising for the actual result you want. I'd strongly recomend you test a 'Leads' campaign or a 'Sales' campaign with a conversion objective. This tells Meta to find people who are more likely to take a higher-intent action, like filling out a form or completing a purchase. Yes, the cost per initial result (like a click or even a lead) will almost certainly be higher than what you're paying now, but the quality of those leads should be miles better. You're paying a bit more to get in front of a much better audience.

I'd say you need to rethink the lead capture process...

Directing people straight to WhatsApp is quick and easy, which is precisely the problem. It creates a very low barrier to entry. There's no friction. This means you get everyone and anyone dropping you a message, with no way to filter them beforehand. You then have to spend your own valuable time manually qualifying each and every person, which as you've found, is massively inefficient.

You need a better filter. Instead of a direct chat, I'd suggest two alternative routes to test:

-> Meta's Instant Forms (Lead Gen Objective): This is probably the best place to start. When someone clicks your ad, instead of going to WhatsApp, a form pops up right there within Facebook or Instagram. The beauty of this is that you can add custom qualifying questions. You can ask about their budget, their specific needs, their timeline, or any other question that helps you identify a "good lead". This does two things: it forces the prospect to think about what they actually want, and it automatically weeds out the people who aren't a good fit. We recently worked with a software company and by using Meta lead gen forms with qualifying questions, we achieved 4,622 registrations at a cost of $2.38 per registration.

-> A Dedicated Landing Page (Sales/Conversions Objective): This is the next level up. You send ad traffic to a specific page on your website designed for one thing only: converting visitors into leads. This page would have persuasive copy, clear details about your offer, and a contact form. It's a higher-friction approach, which is a good thing. Someone who takes the time to visit a page, read the information, and fill out a form is inherently more qualified and more interested than someone who just clicks a 'Message Us' button. Your cost per lead will be higher, but the return on that lead should be much, much greater. This is a standard aproach for most B2B or high-ticket offers.

Both of these methods give you a structured way to capture information and qualify leads before you ever have to speak to them.

You probably should use offline conversions, but let's get the frequency right...

You're absolutely right that feeding data back to Meta via offline conversions is a powerful tool. But your question about frequency is important. Uploading daily is probably overkill unless you're getting hundreds of qualified leads every single day. For most businesses, this would be a lot of manual work for very little extra benefit.

I'd say a weekly upload is a perfect starting point. It's frequent enough to keep the algorithm fed with fresh data, but not so frequent that it becomes a massive chore. The most important part isnt the frequency, though. It's the quality of the data you upload.

Don't just upload a list of everyone who contacted you. You need to create a simple process to score your leads. It can be a simple spreadsheet with columns for 'Name', 'Email', 'Phone', and a 'Status' column. Mark each lead as 'Qualified', 'Unqualified', or 'Customer'. Then, you only upload the lists of 'Qualified' leads and 'Customers' to Meta as separate offline conversion events. This gives the algorithm a crystal-clear signal: "These are the types of people I want more of. Go find them." Over time, this will make your targeting far more effective than any interest-based targeting ever could.

You'll need more specific targeting than just 'affluent'...

This brings us to targeting. Relying on Meta's pre-defined "affluent" demographics is a bit of a gamble. The signals Meta uses for this are varied and often not that accurate. A better approach is to think about the behaviours and interests that correlate with affluence and, crucially, with an interest in what you sell.

Think about your ideal customer. What magazines do they read (e.g., The Economist, Financial Times)? What brands do they buy (e.g., luxury car brands, high-end fashion, specific hotel chains)? What are their hobbies (e.g., golf, sailing, skiing)? Targeting layers of these specific interests is far more effective than a broad demographic bucket.

However, the real power comes when you combine this with the data you're now collecting. Once you have a list of at least 100 qualified leads or customers from your offline conversion uploads, you can create the most powerful audience in Meta's toolkit: a Lookalike Audience.

You would create a Lookalike Audience based on your 'Qualified Leads' custom audience. This tells Meta to analyse the thousands of data points associated with your best leads and go find a million new people who share those exact characteristics. A 1% Lookalike of your best customers is lightyears more effective than trying to guess interests. This is exactly how we scale campaigns for clients. We start with some initial interest testing to gather data, and as soon as we have enough quality conversions, we switch to Lookalike audiences built from those converters. It's a proven path to finding higher-quality prospects.

This is the main advice I have for you:

As you can see, getting this right involves several moving parts. It's a process of changing your entire funnel, from the ad's objective right through to how you handle the data on the backend. It takes time, consistent testing, and a clear understanding of how the different elements work together.

Implementing this kind of strategic overhaul can feel a bit daunting, and it's easy to waste budget on tests that aren't structured correctly. Working with someone who has done this many times before can obviously speed up the process and help you avoid common pitfalls.

If you'd like to go through your setup in more detail and get a clearer plan of action, we offer a free, no-obligation consultation where we can review your account together. It might be helpful to have a second pair of expert eyes on it.

Regards,
Team @ Lukas Holschuh