How to Plan a Marketing Budget for Your App Launch

Building is easy. Selling is the job.

Let me be blunt. In the AI era, anyone can get an app built. What decides whether your business lives is whether you can sell: bring customers in, get them to pay, and get them to come back, at a cost lower than what they bring you.

Most first-time founders plan the app in detail and the marketing in one line: "We will run some Facebook ads." Then launch day arrives, the budget is gone, and the app sits in the store with a handful of downloads.

So before anything else, ask yourself:

• How will my first 100 paying customers hear about me? Name the channels.
• How much can I afford to spend to win one customer, and how do I know?
• If the first month brings half the orders I expect, can I keep going for six more?

This guide shows how to answer those questions for a two-sided business such as ride-hailing, food delivery or home services.

You are marketing to two sides

A marketplace has two customer groups, and you must win both:

Spending on demand while supply is thin is the most common way to waste a launch budget. Every customer who opens the app and finds no driver or no open restaurant is expensive to win back.

Start with the harder side

NFX's well-known list of 19 marketplace tactics puts this first: launch with the side of the market that is harder to get, and test your sales and onboarding process to find out which side that is. In many local service markets it is supply, but do not assume. Your manual pilot will tell you.

Other tactics from the same list that suit small-budget founders: target a tight niche first, limit your launch to one geographic area, match both sides manually at the start, and, where it makes sense, pay the most valuable side to join. NFX notes that Uber initially paid drivers in key cities to be on its app so that riders could always find a car.

Channels, from cheapest to most expensive

1. Referrals

Your existing customers and providers are your cheapest salespeople. A simple "give one, get one" credit, paid only when the new user completes a first order, keeps cost tied to results. Make sure referral rewards cannot be abused with fake accounts: pay out on completed and paid orders, not sign-ups.

2. Promotions and launch offers

First-order discounts, free delivery for a week, or a fixed-price ride during launch can drive trial. Set an end date and a budget cap before you start. Promotions that never end become your real price.

3. Local partnerships

Offices, universities, residential compounds, hotels, malls, clinics and community groups can bring you concentrated demand. For food delivery, restaurants themselves can promote your app to their walk-in customers. For services, property managers and real estate agents are natural partners.

4. Community and direct outreach

Paul Graham's essay "Do Things That Don't Scale" argues that founders must recruit early users personally. In practice this means flyers in the right buildings, WhatsApp and Facebook groups, stalls at events, calling businesses one by one, and answering every early customer's message yourself.

5. App store optimisation (ASO)

Your store listing is your shop window. Both stores give you free testing tools:

• Apple lets you run product page optimisation tests on a live app with up to three treatments, varying the icon, screenshots and app previews against your original page.
• Google Play offers store listing experiments for the icon, feature graphic, screenshots and, for localised experiments, descriptions, measured on installs and opens.

Test one change at a time, and localise your screenshots and text for each language you serve.

6. Performance ads

Paid search, social and app-install ads scale quickly and can burn money just as quickly. Google's App campaigns, for example, can show your app across Search, Google Play, YouTube and the Display Network, and can be optimised for installs or for specific in-app actions. Optimise for completed orders, not installs: an install that never orders is worth nothing.

Use paid ads after your cheaper channels show you what converts, not before.

How to set the budget

There is no correct percentage of revenue for a new marketplace. Build the budget from your own numbers instead.

Step 1: Work out what a customer is worth

From your manual pilot, estimate:

• Average order value
• Your take (commission or fee) per order
• Variable costs per order (payment fees, support, refunds, incentives)
• Orders per customer per month, and how many customers are still ordering after one, two and three months

Contribution per order times the number of orders a typical customer makes over a period gives you a rough lifetime value (LTV) for that period. Be conservative. Early data is noisy.

Step 2: Set a CAC target

Customer acquisition cost (CAC) is everything you spend on marketing and promotions divided by the number of new paying customers. David Skok's widely used SaaS metrics guidance suggests healthy businesses have LTV more than three times CAC and recover CAC within about 12 months. Those are subscription-software rules of thumb, not marketplace laws, and Skok himself says they can be broken in some situations. Use them as a sanity check, then set your own target from your margins and cash position. With limited capital, shorter payback is safer.

Do the same for supply: what does it cost to recruit and activate one reliable driver or merchant, and how long do they stay?

Step 3: Work backwards from targets

If you want 500 paying customers in your first three months and your target CAC is, say, US$6, your demand budget is about US$3,000 before supply costs. These numbers are examples only. Replace them with your own.

Step 4: Keep a reserve

Hold back part of the budget for what works. You will not know your best channel until you test it.

Phasing: pre-launch, launch, growth

An example budget allocation

The table below is an illustrative example, not industry data or a recommendation. It shows how a founder launching a food delivery service in one city might split a first-90-day marketing budget. Your split should come from your own pilot results.

What to measure every week

• Supply: active providers, hours online, acceptance rate, provider churn
• Demand: new customers by channel, first-order conversion, repeat rate after 30 days
• Marketplace health: fill rate (orders that found a provider), wait or delivery time, cancellations
• Money: CAC by channel, promotion cost per order, contribution per order, cash remaining
• Store: listing conversion rate, ratings and reviews

If you cannot measure a channel, cut it or fix the tracking before you spend more.

Common budget mistakes

• Spending the whole budget in launch week. A big splash with nothing behind it brings customers once. Spread spending so you can keep winning them back.
• Counting installs as customers. Only completed, paid orders count. Downloads are a vanity number.
• Copying a big competitor's discounts. Large platforms can fund losses for years. You probably cannot, and customers trained on deep discounts often leave when the discounts stop.
• Ignoring supply until complaints arrive. By the time customers complain about waiting times, you have already paid to acquire people who will not come back.
• No owner. If nobody is responsible for marketing results each week, the budget will drift. Name one person.

Your marketing budget checklist

• I know which side of my marketplace is harder to win, from real pilot data.
• I have a supply plan with recruitment targets and incentives for the first 90 days.
• I have estimated LTV conservatively from real orders, not hopes.
• I have set a CAC target and a payback limit I can afford.
• Every promotion has an end date and a budget cap.
• Referral rewards pay out only on completed, paid orders.
• Store listings are localised and I have planned listing experiments.
• Paid ads optimise for orders, not installs.
• I review channel results every week and move money to what works.
• I have enough cash to survive at least six months even if growth is slow.

A word on who we build for

Joldisoft builds ride, delivery and services platforms, so you might expect us to tell you to spend more on the app. We will not. We do not take every project. We prefer founders who come with validated demand, a sales and marketing plan like the one above, a budget for operations and runway, and a named person who runs the business every day. We want partners whose businesses are still running years later, not one-off builds where the client's money disappears. If your marketing plan is ready, talk to us about the app. If it is not, finish the plan first.