Most developers hate marketing. Cold emails, sales funnels, content plans — all of it feels like someone else's job. Banu from India found a way out: he does what he does best — builds products. And this has become his main marketing channel.
Banu is a former engineer of a large startup who realized eight months into the job that employment was not his path. He returned to his parents' home to avoid worrying about expenses and started building his own products. The first one, Feather, grew to $6,000 a month, after which Banu sold it for $250,000. His second product, Site GPT, now brings him $13,000 a month, and the total revenue since its launch in March 2023 has exceeded $500,000.
What’s remarkable is that he hasn’t spent a dollar on marketing. All traffic is organic. Over a million people have visited his site. About 60–70% of the traffic comes from Google, and almost 90% of this search is through free tools he built himself.
What is Site GPT and How the Business Model Works
Site GPT is a SaaS service that allows any business to create its own AI chatbot based on the content of its website. Essentially, it’s a 24/7 virtual assistant that knows everything about your business and answers customer questions without human intervention.
The business model is classic for SaaS: a monthly subscription, the cost of which depends on the number of chatbots, the volume of content, and the number of messages. The average check is about $100 per month. Banu currently has about 130 paying clients. Of the 50,000 monthly site visitors, 200 become leads, 60 become trial users, and 25–40% of them end up paying. The lifetime value of a customer (LTV) is about $1,700–1,800 — an impressive figure for a product of this scale.
Engineering Marketing: Build Instead of Sell
Banu’s strategy is called engineering as marketing. The idea is simple: instead of buying ads or hiring a marketer, the developer creates free tools that solve specific problems for the target audience. These tools rank in search, drive traffic, and some of this traffic converts into paying customers of the main product.
Banu has created about 50 such tools. Among them are a PDF to Markdown converter, a chatbot name generator, and various AI text generators. Each tool solves one specific task and contains a call to action that leads to Site GPT.
The logic is straightforward: a person who needs an AI-based response generator is a potential client for the AI chatbot service. They find the free tool through Google, use it, see a subtle call to try Site GPT, and transition. No ads, no cold contact — just organic interest.
Creating a new tool now takes Banu less than five minutes: he simply points Cursor AI to existing tools and asks it to create a similar one for a new keyword.
Seven Steps: How to Find Ideas for Free Tools
Banu shared the exact algorithm he uses. Here’s his step-by-step process:
Step 1. Open Ahrefs — Keywords Explorer and leave the search field empty. This allows you to get a mass of all keyword queries without a pre-set direction — especially useful when you don’t know what exact phrases users are using.
Step 2. Add a filter for keywords. For example, if you’re building an AI product, add the words “AI” and “generator” — and get all queries like “AI generator X.” This gives a list of potential tools that real people are searching for.
Step 3. Filter by keyword difficulty (KD). Banu sets a maximum of 10 out of 100. This means that any site with minimal authority can rank in the top for such a query. He intentionally avoids highly competitive niches like “AI image generator.”
Step 4. Add a filter for search volume. The minimum threshold is 1,000 queries per month. The tool should attract real traffic; otherwise, it’s pointless. The intersection of low competition and sufficient volume is the gold mine.
Step 5. Collect the list in Notion. All suitable keyword queries with search volume and difficulty indicated. This is the working base from which priorities are later chosen.
Step 6. Come up with a call to action for each tool. This is a critically important step that many skip. The free tool should lead to the main product. For example: “You tried interacting with one PDF. What if you create a chatbot based on all your business content? Try Site GPT.” Without this bridge, the tool just attracts traffic to nowhere.
Step 7. Prioritize the table. Banu builds a table with columns: keyword query, search volume, difficulty, labor costs for creating the tool, relevance to the main product. Then he sorts by overall priority: high volume, low difficulty, ease of creation, high relevance. Work begins with these tools.
Why This is Perfect for Developers
Most marketing advice assumes skills that engineers don’t have and aren’t interested in developing: copywriting, working with audiences on social media, setting up ad accounts. Engineering marketing requires none of this.
The developer does what they know: writes code. Only now this code works not only as a product but also as a client acquisition channel. Each tool is an asset that continues to bring traffic months and years after creation. Unlike ads, which stop as soon as the budget runs out.
Another advantage is scalability. The first tools require time to create templates and infrastructure. But with the advent of AI tools like Cursor, the process takes minutes. Banu directly states: it’s enough to show AI an existing tool and say “make the same for this keyword.”
The Main Lesson: Do What You’re Good At
Banu’s path is not a story about the perfect marketing strategy. It’s a story about the best marketing for a specific person being the one that aligns with their strengths.
If you’re a developer and hate sales — don’t try to become a good salesperson. Instead, build things that sell themselves. Create tools that solve real problems for real people, and make sure those people know about your main product. Google will do the rest.
The advice Banu gives to his past self is universal for anyone launching a SaaS: don’t spend months perfecting the product before launch. Launch with a minimal set of features — and let feedback from real users determine which direction to go next. The market always knows best.