INSIGHTSROOM

Guys, got my first 47 users after one week, not a big number – I admit, but still want to share with you guys the progress and what I learned.

Here’s the breakdown of where they came from and - more importantly - what I realized I got wrong during the build.

1. High-Trust Referrals > Broad Traffic

• LinkedIn (541 Impressions => 12 Users): Targeting HR professionals with a specific "Employee Engagement" angle worked. It was high-intent and niche. But the activeness of this segment is still in question. I haven't had any follow-up post since then.

• The "Champion" Effect: A friend shared the tool on her company’s Slack. She was very impressed with Prompt-to-Survey feature - not just by the AI "gimmick," but by the logic, wording, and flow of the questionnaire that was generated by AI. That one Slack post brought 10 users.

• The Reddit Void: I got traffic, but zero conversions.

2. "Minor" Features are actually Dealbreakers

I thought "batch pasting" options for questions was a "nice-to-have" polish. My users disagreed. They told me immediately that manually typing out options for a question was a dealbreaker.

A very interesting case is that an user reached out to me and shared that they really need the “batch options input” since they had a question with around 300 options (a list of schools they wanted respondents to choose from).

Acknowledged this need, I have updated the platform. Now you guys can copy options text from elsewhere and paste to the survey instead of manually typing them out.

3. The UX blind spot

This is quite silly. I was so focused on other complex logic that I completely ignored the "door" to the house.

After signing up, I had users redirected to their Profile Page which cause confusion and stuck. . They didn't sign up to look at their profile; they signed up to build a survey. I’ve since fixed this so they land straight in the Workspace so users can build the survey straight away without navigating to the Workspace themselves.

What’s next

That’s it. The high from 1st week launching is wearing off. Now come the phase of testing distribution channels to find a scalable one. I am setting a goal of 100 users by the end of March. If you guys have any suggestions, feel free to leave that in the comment. Also very much welcome you guys to my platform and test it out.

Thanks guys.

Hello guys,

I have been in the market research industry for around 15 years. That’s 15 years of surveying and interviewing customers, and the truth is, I am not happy with any survey tools out there.

Most of the form builders out there - from the costly ones like Qualtrics, SurveyMonkey, and Typeform to the free ones like Google Forms, MS Forms, and Tally - are really good at helping people build surveys. But the bottleneck happens after the user gets the data.

Crunching even a small set of hundreds of survey responses is not a breeze. At Nielsen, we have a specialized team just for that task, and things go well until you request a data analysis ticket with a tight deadline. In so many cases, I’ve had to apologize to clients and promise answers later simply because we didn't have the processed data on demand.

• Not to mention the nightmare of open-ended answers.

• Not to mention the unreasonable subscription plans you have to pay just for basic features.

• Not to mention the learning curve (and the cost) you suffer through if you decide to use a separate tool just for data processing.

Most tools on the market are designed for the "Collector." They make it easy to drag and drop a question, but they leave the "Analysis" entirely up to you. If you aren't a data guy, you end up with a mountain of raw data and no clear direction. Even if you are, handling such data takes time and energy that could be productively spent elsewhere.

I decided to build the tool I wish I’d had back then - one that focuses on the output as much as the input.

Deciding to build was not an easy task, especially since there are already so many tools out there. I spent one year just thinking about it. In the end, the pain points were still there with no one really solving them, so I said, “Fuck it! Let’s build the tool!” And that’s how InsightsRoom was born.

What I did

Deciding to build was the hard part, but only mentally. Actually building a production-ready app as a solo, non-traditional dev was the hardest thing I’ve ever done.

I’ve always been a hobbyist coder, but I never felt "ready" to build a full-scale web application. For a long time, I stayed in my lane as a researcher, thinking that "real" products were only built by "real" engineers.

This wasn't a weekend project. It took months. As I dove deeper into the build, the codebase expanded rapidly because I refused to cut corners on the architecture. I needed a system that helps users build surveys with complex logic as quickly as possible, as securely as possible, and as conveniently as possible for data handling. Moreover, I wanted to provide a new survey-answering experience.

I am quite proud of what I have done - not only because of the features or the UX, but because I never thought I could build something like this in the past.

What InsightsRoom does

It has features that I believe will change the way you conduct surveys (for the better, of course):

1. The Survey Builder is intuitive. I tried to make the survey editor as easy to use as possible with a really lean UI. You can build a survey in seconds using the AI builder tool. Just tell the AI what you want, and the survey - complete with screening logic, branching, and options - will be generated. To be honest, the output is quite impressive the very 1st time I tested, and based on various feedbacks from users.

2. A chat-based interface. For the respondent’s side, I used a chat interface. This is friendlier and likely increases data quality compared to traditional interfaces, which can feel more like an interrogation.

3. AI integration in responses. Imagine you ask an open-ended question like “How do you feel about our product?” and a respondent answers “Nothing special.” For other tools, this is a dead end. But with InsightsRoom, if you enable the AI follow-up function, it can generate a question like: “What do you mean by ‘nothing special’? Can you please share more about what didn't meet your expectations?”

4. Prompt-to-analysis. You can ask questions like “What are the differences between Gen Z and Gen Y in dealing with office stress?” and get instant charts and answers. Say goodbye to hours of data processing. But if you feel nostalgic for those hours, you can always download the raw data for manual processing (though that will hurt my feelings!).

I’m currently in the market validation phase and would love to hear from you. All non-AI features are free - not only for now, but forever. For AI-related functions, I am giving you 200 credits to experiment with, which is equivalent to a small survey project. There is absolutely no obligation to use the AI features. Every advanced feature found on other survey platforms is here, for free.

Give it a try and let me know what you think.