I am helping thousands of dev bloggers attain creative freedom with Hashnode. I am Sandeep Panda, ask me anything!

View other answers to this thread
Bhanu Teja Pachipulusu's photo

Hey Sandeep, Thanks a lot for doing this AMA.

Almost all the questions (that I have), have already been asked.

Here are a few more questions that I have.

  1. How did you know when you hit a product/market fit?
  2. Why did you move away from the initial model of Hashnode? how did you know that it was time to move away from it and focus just on Developer Blogs? Are there some metrics that you used, which helped to make this decision?
  3. What other metrics do you guys usually use to measure the growth?
  4. How do you prioritize one feature over another? Is it just intuition and consensus among the team?
  5. How are you guys able to make so many changes that fast to the platform?
  6. What is the high level architecture of Hashnode like?
  7. Do you have single frontend and single backend? Or do you follow the approach of micro frontends/micro services?
  8. Are you guys using Serverless architecture? If not, have you considered using it even in the initial stages of Hashnode?
Sandeep Panda's photo

Hi Bhanu! Thanks for the questions!

How did you know when you hit a product/market fit?

You know you have hit Product Market Fit when you no longer need to ask if you have PMF 😀.

To be honest when PMF happens you automatically feel it. When we launched Hashnode as a blogging platform 2 months back, we instantly onboarded a lot of dev bloggers onto the platform. Every day we monitor whether or not those developers come back to the platform.

Our weekly retention curve looks healthy and indicates that our users love to come back to the platform on a daily and weekly basis. This indicates that our offering definitely has a strong value proposition.

Why did you move away from the initial model of Hashnode? how did you know that it was time to move away from it and focus just on Developer Blogs? Are there some metrics that you used, which helped to make this decision?

To be honest, the old model of Hashnode (i.e Q&A) was more of a vitamin than a pain killer. We wanted to build a conversational community for developers and we succeeded to an extent. However, we were not growing fast enough.

Eventually, we felt that blogging was a better problem to solve. Since none of the existing platforms focused on content ownership, we thought we could make a difference by helping developers start a blog on their own domain easily. Developers relied on platforms like Ghost/Wordpress (for publishing) and Medium/devto etc (for community) -- we thought of combining the best of the both worlds and offer developers both content ownership and an active dev community. This seemed like a bigger and more interesting problem to solve. Hence we pivoted.

What other metrics do you guys usually use to measure the growth?

We have several growth metrics such as DAU, MAU, Weekly Active Writers, # of blog posts per week and so on. Apart from that we also measure daily, weekly and monthly retention.

How do you prioritize one feature over another? Is it just intuition and consensus among the team?

We have a feature requests page and our roadmap is driven by it. If we think of building something into the platform, we log a new FR and take opinion from our users on Discord.

How are you guys able to make so many changes that fast to the platform?

Despite being a small team, we believe in shipping and iterating fast. That way we can experiment faster and move on to a different idea if an experiment fails.

What is the high level architecture of Hashnode like?

It's Next.js and MongoDB. A comprehensive article is on the way. :)

Do you have single frontend and single backend? Or do you follow the approach of micro frontends/micro services?

We have a set of geographically distributed nodes so that we can serve content faster. These nodes are on DigitalOcean and AWS both. And we have several micro-services running. :)

Are you guys using Serverless architecture? If not, have you considered using it even in the initial stages of Hashnode?

We are using Cloudflare Workers to implement SSO which is the only serverless component of Hashnode. :)