At the end of 2021 I quitted my job to create a company with an ex-colleague. It did not work out. One year later, both the first and the second product were a failure, my business partner had quit to create a VC-funded startup, and I was left to pick up the pieces. Not fun.
Since then I have been trying to find a new product to build while looking for contract work. This is a summary of my thought process around my last idea.
It all started with Hacker News
Recently one of my articles was posted on Hacker News. It is a big deal: lots of tech people read HN, and having your content featured on the first page is really nice when you are trying to get feedback, connect with like-minded people or simply raise awareness about your work. I do not post my own content on HN (though I have been told by someone from a YC company that “everyone does it all the time”), and I spotted the submission almost by accident.
Last week I was thinking about it again; missing the submission would not have been the end of the world, but it would have been useful to receive some kind of real time notification. A simple Telegram notification would work, how hard could it be? And it was not just Hacker News. I would love to be notified if my content was being posted on other websites such as Lobste.rs. My imagination started running wild. There are a lot of things on the internet I watch from time to time. But I would really prefer a real time mechanism notification. Some services support it; most use email which is useless because I do not check my emails every 5 minutes.
I suddenly remembered one of the discussions we had during a user research interview. A consultant told us very clearly: “automation is nice and all, but sometimes all I want is an SMS”. Could it be a viable product?
PingMeWhen: the concept
The core idea is a small SaaS where users can watch for various events and be “pinged” in real time when these events occur. No code, no complex rules, just an email, SMS, Slack message or Telegram notification with a link and maybe a custom message. It would start with watching Hacker News of course: when a link to your website is being posted, or when specific keywords appear in a title or in comments. Then I would extend it to other platforms. I could then consider more complex events: security vulnerability alerts, financial data, anything happening anywhere really. My mind was buzzing with all the possibilities! I always liked the idea of letting the computer do the work and tell me when something interesting happens. This could work.
I exposed the idea to a few friends; they tought it was interesting. I found a name, PingMeWhen, and immediately bought the pingmewhen.net domain. One thing done. The technical aspect was not too hard: regularly poll various data sources on the Internet depending on what users want to watch, emit events when specific things happen, and send notifications. I was confident I could build it quickly.
Seemed solid. But it was not.
Who is going to pay for that?
Something I learned from my last two failures is that there is no point in building a product unless you have established that someone (ideally multiple someones) is going to pay for it. For real, not just “yeah probably” but with the well known “take my money!” reaction. So this weekend I started to think about the potential market. It was not great.
Your product is either B2B, where you sell to other companies, or B2C where you target simple consumers. In this case, consumers are a really tough market: there is nothing really attractive about real time notifications, and people really do not want to pay for anything. Even if I could manage to somehow market the product to them (and I have no idea how), I could not hope to have anyone pay more than a couple dollars a month. Hardly a good business model: I am alone and have no way to gather thousands of consumers.
So why would a company use this product? As some Google search showed me, companies which really benefit from reacting quickly to what happens on the Internet already have specialized tools for that. All of them have their limitations of course, but they all target specific use cases (e.g. content marketing) and are therefore much easier to sell. Specific beats generic all the time. For more technical use cases, the situation is not really different. IFTTT already lets users react to a wide range of events with way more flexibility than a simple notification. And it is so cheap I cannot start to think about how to compete with them. Furthermore IFTTT is just one of them, there are dozen of similar powerful tools, including open source ones such as Huggin.
Something is really wrong: if I have trouble figuring out who is going to buy the product, does it really make sense to build it? Probably not.
It took me three days from the first idea to the end of this article. I usually do not mind discarding ideas, but this one frustrates me: IFTTT does not have a recipe to let me know when one of my articles is being posted, and I really do not want to manage a Huggin deployment, so I’m back to square one. I feel like if have this kind of problem, others may have too. I am just not convinced enough people would pay for it.
At least I learned a good lesson: do not buy the domain name before being sure there is something worth pursuing. 20€ for a .net domain is way more expensive that it used to be.
I imagine I will keep the idea around just in case. And I might get some interesting feedback from readers, who knows.