Why Facebook Ads Aren’t For Indie Projects

Following up on my previous posts about my project of cards and desk mats for programmers today I’d like to dig deeper into the topic of ad networks and why these aren’t for me and other indie creators.

Since I had moderately biggest successes with Facebook ads (incl. Instagram) I’ll focus on this network. But all of these observations apply to Google Ads, LinkedIn ads, Reddit ads, TikTok ads, and all other ad networks.

Ad networks seem like a nice way to reach your audience to inform them about something they like. For the last three years I was trying to set up campaigns to scale my business. Eventually it never happened.

Once again my last campaign gave a cost of sale around 100 dollars, what means it’s two times more than product price and ten times more than reasonable ad cost.

Before you try using ad networks I encourage you to read this article, whatever SaaS or other product you want to advertise.

Self serving ads seem great. You can optimise campaign for sales, narrow down audience to interests, and theoretically you should just scale campaign to improve sales. But it’s not how it works. Specifically due to several factors I’ll describe here.

First of all let’s talk about money. The general idea of indie, bootstrapped projects is that you build something awesome, set up campaign and gather profit.

But if you will calculate several factors it becomes really dubious.

You have to factor in the cost of manufacturing, cost of delivery, insurance, taxes, Shopify fees, Stripe fees. What’s left is for ads and your profit.

Let’s say your product costs 50 dollars. After subtracting costs you’re left with, say, 10 dollars. Now let’s assume you split it with ad network. 50/50.

They get 5 dollars per sale and you get 5 dollars. Sounds fair right?

Let’s go with this example. If you invest 1000 dollars in ads, you’ll sell 200 products right? So your profit will be 1000 dollars.

But the problem is that 5 dollars for ads to sell a product is impossible. A realistic cost is 20-30 dollars. And to reach it you have to heavily optimise every aspect from product to website to ads.

But let’s assume you manage to become Phd in ads, conversions, web design, UX.

You still have to have money for the ads. But you can’t up the price because it’s already high and if it will be more people won’t buy at all.

So you have to lower cost. Let’s say you want a profit of 5 dollars, ads per sale will be 20 dollars, delivery, insurance, fees, tax 15 dollars. So you have to manufacture for 50-40=10 dollars.

You now sell a product manufactured for 10 dollars, paying Facebook 20 dollars and earning 5 dollars. Makes no sense at all.

But let’s say you find it a nice deal. You invest 1000 dollars in ads in first month. You get around 1300 back right?

But it doesn’t work like that. Before you reach 20 dollars of cost per sale you have optimise everything heavily. And to do so you have to modify campaign.

But Facebook needs 50 conversions to really learn. So you have to spend at least 1000 dollars to know if the algorithm reached your goal. At least because the initial cost per sale will be higher.

After that you change settings to optimise campaign, Facebook resets it and learns from start.

Let’s assume you’ll have to optimise campaign 10 times before you reach 20 dollars per sale. You already lost 10 thousand dollars and can at the end discover that the goal can’t be reached.

Why not? Because there might be high competition for that audience. Moreover ad networks use historical data to establish minimum cost per click. It means that if people paid 50 dollars for cost per sale, even if no one advertises to your audience now, you’ll still have to pay 50 dollars even without any current competition.

Of course for indie project shelling out 10 thousand dollars just to check if they can set up profitable business is ridicolous. Especially since you don’t get anything in return. As long as you run the campaign you can or can not see replicate results. When you stop the campaign and start later you may have to start teaching algorithm again what costs 1000 dollars.

One might think he could just scale the campaign with budget. You spend little one and increase it gradually. That sounds great. But again, with smaller budget learning algorithm takes 1000 dollars but in longer period of time.

But let’s assume you can assess profitability before on lower budget. You can’t increase the budget once you do it, because Facebook will reset the algorithm with every major budget change.

It means that you have to start with substantial part of your target budget to scale it in reasonable time. Starting with 10 dollars would give results in a week, and then increasing the budget by 5% per week means you’d reach budget of 1000 per week in almost two years.

All in all it means that you’d have to spend 10 thousand dollars to optimise the campaign and there’s no other way.

So to sum it up, if you lower the production cost, optimise your website, product, everything, pay more for ads then the product manufacturing cost and have 10 thousand dollars then maybe, just maybe, it will be all profitable.

I dont even mention that the future sales would have to cover the optimisation costs of the ads. I mean the period where the costs would be higher than the income.

What brings the conclusion that Facebook ads make no sense for my decks for programmers for several reasons:

• I won’t pay 20 dollars to Facebook to sell a theoretical product I would manufacture for ten dollars because it conflicts with my sense of what is ethical
• I won’t shell out 10 thousand dollars to maybe earn max. 3 thousand or loose 10 thousand and end up with nothing (not because I don’t believe in my products, I don’t believe in ad networks based on my experience)
• I dont even have 10 thousand if I had, I’d put it in a bank. Even gambling sounds better than Facebook ads in this scenario

I hope these calculations show why I don’t think Facebook ads are for indie developers like me. Whatever angle you take it makes completely no sense to advertise. I brought up Facebook ads, but for other ad networks out there including Google Ads, the outcomes from my experience are even worse.

Maybe I could bundle and sell bundles, but as I wrote in my previous article it isn’t trivial.

Two things worth noticing from this is that first of all, maybe there’s a sense to build decks that have less cards and without UV coating and linen texture like the main decks. I’m working on a Clean Code deck for that matter and you can already preorder it.

Secondly, maybe it makes sense to offer printables. And I work on such printable too. It’s actually not a deck, but a role playing game for kids 8+ in a fantasy world. It is a mix or standard kids game with a RPG with DM.

But compared to already existing games of this genre it isn’t 1000 page long, but 20 pages, so people can play immediately. The name is Quest of The Dragon’s Lair and is loosely inspired by Hobbit by Tolkien. You can preorder it too.

That’s it for now. If you didn’t so far, check out the new gift guide for programmers and language deck mats too. Did I mention t-shirts?

cheers!