Shipping the product is about 10% of running a software business.
The other 90% is getting people to find it, convincing them to try it, optimizing the page they land on, running ads that pay back, emailing journalists who may or may not write about you, tracking competitors, and figuring out why half your signups bail at the pricing page.
That's the business. Most indie builders aren't doing any of it.
The 10% / 90% problem
Developers do some work well. They write the code, ship the product, fix the bugs, add the features. That's the 10%.
Developers tend to skip the rest. They don't publish weekly SEO articles on competitor keywords. They don't run 10 concurrent A/B tests on the landing page. They don't manage a three-platform ad budget or a journalist outreach list. They don't read competitor changelogs every Monday.
The 10% is the product. The 90% is what turns a product into a company.
The spreadsheet era of growth
A serious indie operator runs growth today with about eight tools:
- PostHog for analytics
- Ahrefs for keyword research
- A ghostwriter on Upwork for content
- Meta / LinkedIn / Reddit for paid
- Clay + Smartlead for outreach
- Spreadsheets for competitor tracking
- Spreadsheets for experiments
- Friday afternoons reviewing session replays
Eight tools and about thirty hours a week. If any one slips (the ghostwriter misses a deadline, the outreach sequence breaks, the experiment queue empties), growth stalls. You won't know for weeks.
It works at small scale and falls apart the moment you try to grow.
What the category is missing
Physical commerce got Shopify. Digital distribution got the App Store. Payment infrastructure got Stripe.
Growth for small software businesses is still the Wild West.
The closest options:
Agencies. $4,000-$20,000/month, slow, briefing-heavy, incentive-misaligned.
Freelancers. Cheaper, same pattern. You brief, they execute, you review, you wait.
SaaS tools. Each does one slice (ranking, ads, A/B). You're still the general contractor.
None of them are the software version of "the business runs itself."
What an AI growth team looks like
A category this big wasn't buildable until models got good enough to run on their own. Now they are.
An AI growth team:
-
Reads your product. Your repo, landing page, pricing, positioning. Forms an opinion about who you serve and how you talk.
-
Acts. Writes the article today. Ships the experiment now. Kills the ad this hour.
-
Learns. Watches results. Updates its strategy. Doesn't ask permission for the obvious stuff.
-
Reports. You get a daily digest of what it did and why. You can override anything. You won't need to often.
This is the layer nobody had built. It's the layer that decides whether your software becomes a real business or stays a side project with a Stripe button.
The opportunity
AI builds millions of new software products a year. Most never get found because nobody taught AI to do growth.
Now somebody has.
You're not alone at the wheel
The founders who win from here ship the product, turn on the agents, and go back to building.
You run the product. AI runs the business.