What was the specific moment or frustration that made you decide to build oChats?

I was using the WhatsApp Business app for my own business 8 years ago. Then, I realized that I am limited by WhatsApp Business limitations. Later on, WhatsApp Business launched WhatsApp Business API for the first time. Then I encountered respond.io. I subscribed to it. The onboarding was steep, to a point where I needed to use Postman to get onboard with WhatsApp Business API. Even with that, I could not migrate to another BSP or etc. I am stuck with respond.io. There was no offboarding. Then, it slowly became better as Meta kept improving the platform. That was the time when respond.io started to increase the subscription fee to $79/month for the lowest package. That is when I decided to build my own app. Little did I know building an app was costly, especially for a non-developer (AI is not here yet). I hired an agency to build this app for me and ended up spending a lot of money on it.

What's one growth tactic that worked and one that completely flopped?

Worked, I use SEO to acquire leads and the ROI is much better than running ads. Of course, I also try my best to list as many marketplaces as possible (hopefully this works). I am still constantly trying to discover what other channels I can use without spending money. Flopped, I spent $3,500 on Google Ads targeting highly competitive keywords and got exactly one signup who churned a week later. This is definitely not something I had forecast. Waste of money. I tried social media marketing by posting regularly on Facebook, LinkedIn, and IG but it doesn't work. I got little to no engagement. Pending to try, I am going to try Reddit and LinkedIn again. Perhaps webinar and YouTube too. For sure, I am going to try Google Ads again after I get a good amount of paying clients using ochats first. I am going to use bootstrap style to grind this through.

What do your customers complain about most, and how have you handled it?

Early on, users complained about a lot of things, - message delays - no notifications - messages don't pop out until they refresh the browser; only then do they see the message - the app is buggy and slow in loading - the workflow module sends multiple messages in a loop - the server gets jammed because too many workflows are on the queue - no proper error handling when creating contacts and WhatsApp template messages At the end, I put everything on the shelf. Lately, I opened the shelf again because now there is a code to fix them all. So I fixed all of the above by changing the core engine to Next.js, Redis, and more for a faster engine (suitable for instant messaging software). Then, I improved the error handling and fixed all those bugs. Now, I am revamping other modules and building new functions at the same time.

What's one thing about your market that most people get wrong?

In this industry, there is a high demand for bulk WhatsApp blasting. But it turns people off when they see the pricing of the Meta template messages charged by Meta. It is so high. This is really unfortunate. Most of the people who are looking for this product are generally in a grey industry (for example, gambling industry, etc). Meta doesn't accept this kind of industry. That's why it is hard for the lead to come by. Don't even need to talk about securing a paying client. Also, most people want many messaging apps they want in one platform. Certain messaging apps are not available because the API is not readily available in the public. Because of one missing app, they decide not to subscribe to this type of software. Most people think there are no charges for using the WhatsApp Business API. They are wrong because Meta charges users for using WhatsApp API template messages.

What's next for oChats in the next 6-12 months?

Over the next 6 to 12 months, we are launching an intelligent AI Agent powered by custom Retrieval-Augmented Generation (RAG) functionalities directly within oChats. This will enable small businesses to upload their own documentation, FAQs, and files so that the AI can automatically handle complex customer inquiries on WhatsApp and live chat using accurate, business-specific context. Also, we are building full Model Context Protocol (MCP) compatibility into oChats, transforming the platform into an open, MCP-ready ecosystem. This architecture will allow developers to securely connect external AI engineering tools like Claude Code and Cursor, empowering them to interact with oChats data, manage live tickets, and trigger automated messaging workflows programmatically through standard AI-driven interfaces.

Before you started building, what was your relevant experience? Had you worked in this industry, built other products, or were you starting from scratch?

I came from a marketing background as a digital marketing manager (no coding experience at all), so I knew firsthand how painful and chaotic customer communication can be because I needed to communicate with clients consistently through multiple messaging apps on a daily basis. I taught myself how to use Claude code to fix and continue building this product. The barrier of entry to continue developing this product has lowered. I am glad. After I learned how to use Claude code, I also built other products such as Yaeris MCP (yaeris.com/mcp-server), developed WordPress plugins and Shopify apps to list OChats in many marketplaces.

What's the biggest mistake you've made so far building this product, and what did you learn?

Spent too many development dollars to build this software through a web development agency. The main reason I had spent that amount was because I listened to 1-2 prospects who gave advice and didn't pay any money, assuming that all clients wanted the same functions. So, I continued to build the software until a point where the development cost became bloated. This was the time when there was no AI yet, and I didn't know how to code at that point in time, so I had to depend on other people. What a steep price to pay to enter the software industry without prior experience and knowledge.

If you could go back to day one, what would you do differently?

Listen to the client who subscribes to my software instead of a person who never subscribes but just advises. Only execute the recommendation after getting a good amount of feedback from a good number of paying clients.