How Much Marketing Budget Does A Startup Need?

A founder asked me recently if spending 15% of revenue on marketing was enough.

I asked what their customer lifetime value (LTV) was. They didn't know.

I asked what they could afford to pay to acquire a customer (CAC). They weren't sure.

They just knew they "should" spend 12-20% on marketing because that's what the articles say.

This is completely understandable and a totally comprehensible way of solving the very hard problem of setting a marketing budget when you don’t know where to begin.

It’s also backwards.

When you pull figures online telling you how much companies should spend on their marketing budgets, what goes unsaid is this: that figure is downstream of unit economics. Or put another way, if your average customer is worth $300 and it costs $200 to deliver on your promises to them, then you cannot profitably spend more than $100 to acquire them.

Don’t get me wrong—there’s wisdom in the convention advice that startups typically allocate 12-20% of gross revenue to marketing. But if your company has $0 revenue, then 12-20% of that is…$0.

So what do you do when you’re first getting started?

You can’t reliably use “percent of revenue” to know how much marketing budget a startup needs. And there are a few reasons for this.

Percentage of revenue is good for more established companies that need to determine whether their marketing budgets are really low or really high. But if you’re really early on, it’s not going to be very useful. In fact, even if your company is more established, there are far better ways to set a budget as I’ll discuss below.

If you’re serious about proving ROI for marketing, then you need to figure out your cost to acquire a customer (CAC) and the lifetime value of that customer (LTV). Because those two figures are going to help you figure out how much you can spend on marketing and make a profit.

Let's say your average customer is projected to be worth $500 in lifetime value. If it costs $300 to deliver on your promises to them, you must spend no more than $200 to acquire the customer.

Or if you put it another way, if you take LTV and subtract CAC and cost of goods sold (COGS), and the number is positive, then your business model is sustainable.

A lot of times, marketers will use LTV:CAC ratios to determine an acceptable amount to spend on CAC. For example, in a lot of industries, the commonly accepted benchmark is to maintain an LTV:CAC ratio of at least 3:1 or 4:1, meaning for every dollar spent on acquisition, your business should earn $3-$4 in return. So if your average customer has a lifetime value of $500, that gives you a CAC ceiling of $125-$165.

It’s at this point that you can start looking at what is a normal LTV or CAC in your industry. If you use broad, general stats, you’ll find that the average CAC for eCommerce sits around $78. But be careful with this figure and anything like it, because they have a nasty tendency to hide massive variation in the name of neat averages. Especially if you sell low-margin items.

For now, all you really need to know is that if your LTV is $500, your COGS is $300, and the cheapest you can acquire a customer is $200, then you don't have a marketing budget problem. You have a product or pricing problem. No amount of "15% of revenue" will fix negative unit economics.

There’s no single right way to start a business, which means there’s no clean categories or neat boxes when it comes to budgeting. But there are general stages which businesses go through, and stage determines the logic that goes into coming up with a budget. And it all depends upon how far along a business is on the explore-exploit tradeoff (the need to try new things vs. doubling down on what works).

I’m painting with broad strokes here. Read what’s below with that in mind.

Budget: As little as possible. Ideally under $2,000/month.

You aren't "doing marketing" yet. You are doing customer development, and more broadly, you’re trying to figure out if marketing can work for your business model at all.

Your goal here isn’t volume, but rather validation. You need to answer three questions:

Most of your "marketing budget" at this stage should actually be spent on time, not ads. Talk to as many potential customers as you can, and focus on testing offers manually through direct outreach.

The common mistake here is running ads to "learn about the market." While a small amount of ad testing can be a good way to learn about the market, it’s easy to take this way too far. And if you do, it just becomes an incredibly expensive way to learn what a phone call could have told you for free.

Do qualitative research first. Only after you have strong hypotheses about who buys and why should you spend serious money testing channels.

Budget: 15-25% of revenue (or uncapped).

You've found a channel where you can acquire customers at your target CAC consistently. Now you move from explore mode to exploit mode.

Your goal is to scale this channel until it saturates or your CAC rises above your threshold.

In this stage, the "percent of revenue" rule often breaks down in the best way possible. If you put a dollar into a machine and it gives you four dollars back—and you have the cash flow to support it—why cap your budget at 15%?

You should spend as much as your operations and cash flow can handle.

Budget: 25-50% temporarily, settling back to 10-20%.

This is where things get messy. At some point, that first successful channel is going to saturate. Or you’ll just get sick of your revenue depending entirely on one channel and you’ll want to de-risk the business. Either way, you have to find Channel #2 and Channel #3.

The budget spikes here because you are doing two expensive things at once: you are paying to maintain your old channel (exploitation) while simultaneously burning cash to test new ones (exploration).

This is why venture-backed SaaS startups often spend significantly more on marketing than bootstrapped ones. When the business offer is good and at least one channel is proven to work, venture capital can temporarily remove the cash flow constraint, and make it easier to force growth across multiple channels simultaneously by throwing money at the problem. (Which is absolutely a valid way to go about things if you can do this.)

Now if you’re on Main Street and you’re paying for your business out of pocket, your situation is different. You have to be careful and you can’t burn your runway. So you may choose to surge your spending by less in order to burn your cash at a lower rate, knowing it’ll take longer to get out of the so-called “Valley of Death.” This is an equally valid approach to budgeting.

Budget: 5-15% of Revenue.

You know what works. You have brand equity. You are focusing on efficiency and retention. This is where the standard "10-15%" advice (or your industry’s equivalent benchmarks) finally becomes accurate, as you shift from pure acquisition to efficiency.

My framework above is neat, but reality is messy. Good unit economics can ward off many problems, but there’s still a few common ways your budget can break if you’re not careful.

I’ll walk through them with you so you don’t find yourself shaking your fist at the sky six months later.

If you spend $10,000 on ads today, but the customer doesn't pay you for 90 days, you will go bankrupt even if you are "profitable" on paper. It’s said that 82% of startups fail due to cash flow, not lack of profit. Your budget is often dictated by your bank balance, not your LTV.

Every channel has a ceiling. In a small market, saturation happens quickly. In a large market with poor targeting, you start reaching the wrong customers. The right spend is: scale up to the saturation point, then stop or shift to testing new channels.

Most companies can't reliably tell you which marketing drove which revenue. They know they spent money, and they know revenue went up, but the causal link is messy.

The lower your confidence in attribution, the more conservative your budget must be. If you know Channel X drives revenue, spend aggressively. If you are guessing based on last-click data, keep the purse strings tight.

Sometimes, marketing isn't the constraint. If your churn is high, or your sales team can't close the leads you already have, spending more on marketing is lighting money on fire. You need marketing triage—fix the hole in the bucket before you turn on the hose.

If you forced me to give you a framework, it wouldn't be a percentage. It would be this process:

Take your LTV and subtract COGS. Then subtract a bit more to give yourself a margin of safety.

That is your ceiling—the absolute maximum amount of money you can spend to acquire a customer. If you can’t beat that number, fix your offer or business model before spending on marketing.

Don't average them. Budget them separately.

You probably don’t know which half of your marketing works. And, honestly, that’s OK and is even normal. But the more uncertainty there is, the more conservative you should be about how fast you scale.

If you haven't validated that people want what you're selling at the price you're charging, spending money on marketing is just funding expensive research into why your offer doesn't work.

Start with one clear hypothesis. Budget enough to test it conclusively—usually $1,000 to $3,000. If it works, scale it until it breaks. If it doesn't, turn it off and move on.

The expensive mistake isn't spending too much or too little. The expensive mistake is continuing to spend on channels that don't work because "we need to spend 15% on marketing."

No, you don't. You need to spend whatever it takes to acquire customers profitably and test alternatives to hedge against saturation. The percentage is an output of your unit economics, not an input you choose first.

If you’re anxious, fixating on a specific benchmark like "15% of revenue" can make you feel better. After all, it gives you a hard number to put in a spreadsheet so you can tell your investors (or yourself) that you are doing things "by the book."

But that certainty is an illusion. And illusions bankrupt businesses if left unchecked.

The reality is that your marketing budget is not a fixed pie to be sliced up. It is a variable engine that changes based on how well your machine is running.

In the early days, your budget is near-zero because you are trading time for knowledge. In the growth phase, your budget should be practically uncapped because you are trading quarters for dollar bills. And in the mature phase, your budget settles down because you are optimizing for efficiency.

So, stop asking "how much marketing budget does a startup need?"

Instead, ask more specific questions like "what is the max I can afford to spend to win a customer?” And “do I have a channel that can do that?"

If the answer is yes, spend until you break it. If the answer is no, put your credit card back in your wallet. Then focus on fixing your product or your offer.