Start building your own chatbot now

Last week, we launched on ProductHunt! It was a huge boost for SAP Conversational AI, but it didn’t happen quite as we expected. Here’s our story.

How we prepared ProductHunt

A month ago, we got an invite from another cool company we know well: We started planning our ProductHunt launch: what to do, how to do it, when, where, and why. We were planning to launch on Tuesday, 19th.

We registered our entire team on PH and joined the community. We upvoted cool projects, asked questions and there was generally a lot of excitement. What started as a need to bond with the community ended up being a real pleasure, and our team members are still active today on the platform.

We had joined Slack communities to get insights from companies who had already launched, we’d read tons of articles on “Being in the top 5 of Product Hunt in 5 minutes”. We combined actions to create more reach.

One question came up (among millions of others, rest assured): should we keep our waiting list or should we open up our platform entirely? By keeping the waiting list up, you risk loosing valuable users, but letting users test the platform right away, you risk a server overload. We chose the latter with a backup instance for the infrastructure, in case of emergency.

We set up our Intercom and our Slack Community to handle users’ requests and allow them to  reach us easily. We finally wrote down our introduction text, prepared visuals, tweets, people to reach out to, emails and lots and lots of candies.

What we expected

We knew it’d be hard and exhausting. We knew, as a French company in a different timezone, that we’d spend the night in the office. But that was quite exciting.

On ProductHunt, when you submit a product, it goes to the “newest” tab, and when you have enough upvotes, you’re pushed onto  the frontpage. At least that’s what we assumed. With our friends and families ready, and faith in our product’s quality, we were quite positive. We expected a true intellectual debate in the comment section, real insights about the future of bots and artificial intelligence, and so on.

We saw ourselves, tired but happy, the next morning, trending on PH’s front page, and masters of the world. Well…..

How it actually was

It didn’t happen quite that way. Last Thursday 14th, we launched a new feature and posted SAP Conversational AI on HackerNews. That brought us a lot of traffic. Out of the blue, one of our team spotted a “?ref=producthunt” on GoogleAnalytics.

ProductHunt - Google Analytics

“Wow. That’s odd. Maybe they came from PH but by clicking on one of our comments.”

Then another user came from PH. Then three, then four. Then we had to face the exciting reality: somebody had posted us on Product Hunt!

Everybody went into hardcore mode. I checked the status of our infrastructure, our communication manager finalised all our texts, tweets and emails, our designer updated our homepage, and others monitored PH and Google Analytics.

It boomed so quickly we were sure our servers wouldn’t hold it together. They did. We quickly reached 50 people on the website simultaneously, with thousands of visitors per hour from all over the world.

SAP Conversational AI Analytics ProductHunt

We spent the night doing shifts, eating candy, and answering questions.


It wasn’t quite the intellectual debate we expected, but we got great feedback from both PH and via e-mail. By the end of the night, we’d reached 300 upvotes on PH, were #1 in Artificial Intelligence, on the front page as the 8th project, and got a growth of 250%.

Not too shabby. 😉

What’s next?

ProductHunt was a great experience and team building opportunity for everyone. It was a terrific boost for us, now we have a real community engagement, lot of bots in development on the platform and great feedback and feature requests, so we work hard to deliver as fast as possible and keep this powerful dynamic going!

Want to build your own conversational bot? Get started with SAP Conversational AI !

Follow us on
There are currently no comments.

This site uses Akismet to reduce spam. Learn how your comment data is processed.