James He

Reach by Artificial Societies - Test content in a simulation of your own LinkedIn audience

by

Reach simulates your personal LinkedIn audience. Use it to write, test, and improve your content before posting. Simulations used to be sci-fi, now they’re for everyone.

Add a comment

Replies

Best
James He
Maker
📌

Hello Product Hunters, I’m James, co-founder of Artificial Societies! 👋

I couldn’t be more excited to bring you Reach, an AI simulation of your actual LinkedIn audience - yes, your very own LinkedIn!

What’s the story?

At Artificial Societies, we found ourselves posting a lot on LinkedIn, trying to hack the formula for successful content. So we built Reach to help us, and both my co-founder Patrick and I doubled our LinkedIn followers in 6 weeks.

That’s when our friends and YC batch mates started asking for it, so we’ve decided to make it available to everyone.

What makes Reach unique?

Reach isn’t a writing assistant. It lets you take your authentic thoughts and opinions and test them before you post. Not only does this help you write content that your followers engage with, it takes away the risk of your posts falling totally flat and getting ignored - if you know, you know. 🤦‍♂️

Because Reach simulates exactly what happens on LinkedIn, it understands exactly how your LinkedIn audience will respond to your posts. This is how we are able to predict how your post will perform 2x as accurate as predictions made by ChatGPT.

The really crazy part is that, when you test a post, we spin up 100 variants using your tone of voice and everything else we know about you, and automatically test these alongside your original. You can take the best ideas, incorporate them, and keep iterating until you get it right.

How can I try Reach?

Free and paid plans are available now at https://societies.io/

Would love to hear your thoughts, feedback, and questions! 🙏

Masum Parvej

@james_k_he How does it handle niche audiences or evolving LinkedIn algorithm changes?

Patrick Sharpe

@james_k_he @masump  every audience is made bespoke to the user based on who interacts with their posts. We’ve baked in a whole bunch of the LinkedIn algorithm’s features into the design of the simulation and we’ll continue to evolve it over time

Manu Hortet

@james_k_he Woah this is so cool! I was recently looking for A/B testing tools for posts on Linkedin and X, this can totally deliver the value I wanted. Will try it out.

Congrats on the launch and good luck! 🍀

John Tan
Launching soon!

@james_k_he have been looking for something like this. very cool launch and will test this out! 🔥

James He

@john_tans thank you!! let us know how we can make it better!

Konrad S.

That sounds quite fascinating!

Could you explain how it knows how to simulate the audience? Does it analyze the activity of my followers?

Patrick Sharpe

@konrad_sx thanks Konrad, we analyse your own LinkedIn profile as well as your activity (who you engage with and who engages with you in return). This allows us to build both the individual personas and the network. So, you're exactly right, we look at your activity and the activity of those in your audience

Andrew White

never thought of doing a/b testing on linkedin this way! will definitely try. are you planning to do this for ads?

Patrick Sharpe

@andrew_white_13 absolutely, ads / comms / products are all in scope for the future. Let us know if you have something specific in mind

Maciej Balasiński (magic)

Interesting! That opens a bridge for completly new predictions!

James He

@mbalasinski thanks! we want to simulate everything everywhere all at once haha

Simona O'Neill
This is very cool 😎. Will give this a test today. Congrats on the launch guys! Definitely a unique product I haven’t seen before .
James He

@simona_o_neill3 Thank you so much!

Ohans Emmanuel (remention.ai)
Interesting! How accurate are the simulations if hypothetically you don’t have access to the underlying LinkedIn algorithm?
Patrick Sharpe

@ohansemmanuel honestly, scarily accurate. best way to prove it is just by writing a post and seeing how close the prediction is to reality

Tomas Vasso
Congrats on the launch! Looks interesting and will give it a try. Wonder how accurate the simulations are?
Patrick Sharpe

@tomas_vasso Our predictions are getting an 83% correlation with actual performance on LinkedIn, compared to about 40% for the typical reasoning LLMs, GPT-4o is actually only around 17%!

Patrick Sharpe

So hyped to be sharing this with the Product Hunt community!

Manish Choudhary

Wow... If this product delivers what it says. I am sold😁

Patrick Sharpe

@manish_choudhary19 Thanks! Let us know if it lives up to the expectation!

Manish Choudhary

@patrick_sharpe Tried, I love the promise. Will spend a day or two to see the impact. Will reachout to you on linkedin

Jona from ProductRadar

Would be nice to know what works before putting it out there.

James He

@zihui_zhan Thank you! That's why we built it!

Slavisa Dujkovic

Sounds cool! I'll definitely give it a try!

Patrick Sharpe

@slavisa_dujkovic 🙏 let us know what you think

Shushant Lakhyani

How do you ensure that the reach prediction is always accurate?

Patrick Sharpe

@shushantlakhyani It's rarely going to be perfect, that's why we give a range with a lower and upper estimate. But we calibrate our predictions to your profile and audience to make sure it's aligned

Andy Li
This is so cool! Congrats on the launch guys
Patrick Sharpe

@andydy25 Thanks Andy, loving Keywords!

Andrew White

btw the link in your text is a 404

Patrick Sharpe

@andrew_white_13 legend for spotting this! 🙌 We've updated

Tania J

@james_k_he @masump Hi Guys, the product looks promising! However, when i went to have a go at the LinkedIn community, a message pop up "We can only build your audience if you have posted to LinkedIn at least 3 times. If you believe this is in error, please contact us at support@societies.io" - which isn't the case.

James He

@masump @tania_j Thank you for flagging this to us Tania! Let us look into it

Dan Schlung
Launching soon!

just a few years ago I worked on a product that did a/b test automation for landing pages, and was in awe of the ML our engineering team was working on. Never expected I'd see something so sophisticated available to consumers this quickly. This looks great y'all - I'm excited to try it out.

Patrick Sharpe

@dan_difr thank you - we're excited to make simulations possible for everyone - even for something as simple as a LinkedIn post. Would love to hear what you think once you've given Reach a try

Taro Fukuyama

I love using Artificial Societies. It's incredibly helpful to simulate my audience's reaction before I post, ensuring that I'm always choosing the best language for them. The Variants feature is a game changer—it consistently provides better angles and wording for my social posts. I can't imagine working without Artificial Societies now!

Patrick Sharpe

@taro_f Thanks Taro! The variants feature is our favourite too

Owen Oktay

This is awesome. I just signed up. Cannot wait to give this a try. Are you considering X as well?

Patrick Sharpe

@owenoktay absolutely - started with LinkedIn but other socials are on the roadmap

Tom Whittle

Had a viral post last week with this super cool solution!

Patrick Sharpe

@tom_whittle first of many brother

Shivam Singh

Your journey with Reach sounds both innovative and practical, particularly in addressing the challenge of generating engaging LinkedIn content. It's fascinating how you've leveraged AI to simulate real audience interactions, effectively iterating content creation to boost engagement. The ability to test varied content versions using personal tone is a powerful tool for personalization and strategy.


Congrats on the launch! Best wishes and sending lots of wins :) @james_k_he

Patrick Sharpe

@james_k_he @whatshivamdo thanks Shivam - appreciate it