I just launched v2 of my side project Newsletter Spy

Hey,

I finally pulled the trigger. Newsletter Spy 2.0 is live on Product Hunt right now. Here's the link.

Even though I've launched 10+ projects by now, I'm still not immune to the fear of launching.

The project was ready three weeks ago.

But I kept postponing the launch for no good reason, really. I added all kinds of features only to remove them again because they were confusing and added no value. I was waiting for early user feedback that never came. I invented all kinds of excuses why it might make sense to wait a bit longer.

Luckily, two days ago I was on a call with Ty Frankel and when I told him about Newsletter Spy he bought it on the spot. Even though I offered him free access, he insisted on paying for it and told me that this is incredibly useful for him as he's about to start a paid newsletter himself.

This reminded me that Newsletter Spy can be genuinely useful to some people, and it's really stupid to not share it with others.

It's definitely not perfect yet. I fixed some very embarrassing bugs just this morning. But it's usable.

I'm already using it myself to find newsletters I can book ad slots in.

Pete Davies is using it to find Substack users with a podcast that he can invite to his new platform Jam.ai.

Ty Frankel is using it to find partners for his new newsletter.

So the product is definitely already providing value and I can always improve the product later based on feedback by paying users, as I always do.

Anyway, enough of my self-doubts. Let's talk about Newsletter Spy itself.

I launched the first version of Newsletter Spy 9 months ago. At the time, I was inspired by a tweet by Mike Rubini who handed the problem over to me. I finished the project over the course of a weekend. So it was literally just a weekend project.

But it got far more attention than I expected and was used by some of my favorite writers and even researchers from Harvard.

Unfortunately, quickly after the launch, Substack changed their website in a way that made it impossible to fetch subscriber counts. As a result, I was no longer able to provide updates and the project fell into hibernation.

To my surprise, people kept buying access to Newsletter Spy even though I added a big warning that the dataset was outdated and no longer updated. Also at least once a week someone would reach out to me and ask for an update.

Thanks to these conversations, I quickly learned that Newsletter Spy could be useful to many people even without exact subscriber counts. So I decided to revive the project.

Here's what's new:

• The Substack database now contains 50,000+ newsletters. Instead of subscriber counts, it now contains metrics like the category rank and the average number of reactions, which usually give a good idea how big a given newsletter's audience is. Also, in contrast to v1, the database is now updated daily.
• In addition to the Substack database, there's now also a Revue database. This database contains exact subscriber counts for all newsletters that share this number publicly.
• Finally, a third database contains information on newsletters from other major email service providers like ConvertKit, MailChimp, etc. I started experimenting with a machine learning model that predicts the subscriber count, but the results are still meh and should be taken with a grain of salt. However, the database also contains columns like "Fame" (how well known is a given newsletter), "Buzz" (how much are people talking about the newsletter online), "Traffic", Traffic Change" (growing, stagnating, declining) which are more accurate and make it possible to estimate a newsletter's audience size.

If you want to check it out, you can do so here.

And of course, if you have any comments, questions or feedback, I’d love to hear it!

Best,

Jakob