Thoughts on CAC, LTV, Fees and Metrics

"We don't need these, that's very Silicon Valley capitalistic thinking!"

Whenever you mention CAC, LTV, Fees or any other financial metrics. Very crypto-aligned ideological die hard people will fight back against you that we shouldn’t encourage that line of thinking because it’s “evil” and “capitalistic” like the “web2 world of Silicon Valley”.

This is so wrong and misguided on so many levels. I’ll explain why.

Let’s imagine for a second rather than building blockchains and blockchain based apps, we’re building a new internet native society. In this society/civilisation, each address is a citizen.

Now we know each citizen is different and subscribes to different values for existing. Some believe that farming and dumping tokens is completely okay since those are the rules of the protocol, others believe in holding and being active participants in the networks they have a financial stake in.

When we talk about CAC (cost of acquiring a customer), really what we’re saying is:

• Given how much value the product intrinsically has, how much money do I need to spent to communicate that value to a customer. If the product is great and has a fabulous brand/marketing, its CAC costs will reduce over time. If its new and still has to prove value it will have higher CAC costs (maybe in the form of the CEO doing all manual outreach to onboard customers).

There’s nothing Silicon Valley or wrong about this framing, it simply reflects the matter of fact of reality. Now when we relate it in the context of addresses, all we’re saying is how much $ does this protocol need to spend to bridge the value it currently has versus how much extra is needed to communicate that to the right addresses. If we can’t measure this, how do we know what’s really value additive versus what isn’t? CAC costs are a reflection of value created in reality.

Let’s talk about LTV and Fees Paid. This is one that also gets a bad “moral” rep because crypto is all about “cutting out the middle-man and making no money in the process” to certain classes of people. This just feels flat out wrong and wishful thinking. The point of crypto is to remove extractive, value destroying middle-men and replace it with efficient code. There is nothing wrong in capturing value that gets created. Anyone who believes so should send their crypto to 0x0 to stand by their belief. There’s another special property about fees and LTV that isn’t talked about enough: It’s not game-able since it costs a user.


It’s the whole reason why fee markets exist in Layer 1s! Whenever you introduce some sort of fee it means the user has to think whether they think the service of product is good enough for them to spend their energy (time/money) on it. Similarly, if you charge no fees for your product, how do you know if people actually value it? As a logical step, if you don’t know how much your users have paid in fees since the beginning of time, how do you know which ones value your service the most? As a derivative question, how do you know where to spend capital to retain those users and attract ones similar to their characteristics? The answer is you can’t! You’re flying blind.

Pulling all of this together, over time once you can quantify CAC and LTV for addresses a lot easier, you can then start to understand:

• Which users require how much capital to use products

• How much do they actually pay for the products they do like

Knowing those two things will enable much more smarter and less ponzi-like incentives that are currently prevalent in the space. Addresses that don’t require a lot to try new things and will stick around for the ones that are good will accrue more rewards than those that require a lot of capital to try something new and pay the least in fees (airdrop hunters). What we’re really talking about at a higher order abstraction is address based reputation and the improving the incentives for playing long term games rather than short term extractive ones.

If you’re still on the fence, I’d ask you what the alternative is? Treating addresses that have no history the same as those that have been long term ecosystem participants? Would you really pay the same for Vitalik to use your product versus 1 of 5000 airdrop hunter addresses? Ask yourself that and you’ll find the truth.

It’s going to be interesting as on-chain reputation becomes more real in the coming months and years and ultimately, quantified through financial metrics.