Let's Talk Distribution
Now that we've cleaned some things up: we want to get it into the hands of users
I added in a couple more features to buildinpublic.coach over the weekend that I felt were important to ship before focusing too much on getting more users.
I fixed an issue where too many documents were found when looking things up, which would blow up the request by using too many tokens in the API request to OpenAI. I fixed this by changing from the “stuff” method of retrieval, where it would simply grab all of the relevant documents it found and stuff them into the request (surprising, eh?). Instead, I’m now using map-reduce, which gets all of them and iterates over them, building up a composite context from a little bit of each. It takes more OpenAI requests, but it will be more robust and hopefully end up with better results.
I also added in the ability to view previous chats and automatically name the chats based on the first message sent. That makes it a lot easier to go back to conversations you had previously.
Next, I’m going to adjust the pricing I set. I didn’t have anyone take me up on private coaching sessions for the annual plan, so I’m going to scrap those. And now that I’ve got a more full-featured product, I’m not going to discount it so heavily. Over this week, I want to get two things accomplished:
Figure out how many tokens each chat message costs. This will allow me to do some basic modeling on how much a typical user will likely cost, depending on how many messages they send daily/weekly.
Set up the free trial or freemium. I’m thinking about offering around 10 messages for free, but I’m not sure if I want to do it per month, or just initially. I kinda like the freemium model though, so I’ll probably do that. Just give them 10 messages per month for free and if they want more, they can upgrade for unlimited use.
I’m going to bump the monthly cost quite a bit, but I’ll probably still discount the annual plan more than I normally would. I want it to be highly incentivized to sign up for an annual plan so that I have a bit more stability going forward as I develop it out.
Distribution Plan
Now, I’ve got to think about my distribution plan. I’ve been starting to expand my content talking about the product, which aligns with week of of Buildspace Nights & Weekends season 4. In addition to all of the content I’ve been sharing here, on Twitter (sorry, X. Still can’t get used to that…), and on YouTube, I’ve started posting on Instagram Reels and TikTok as well. I want to see if I can leverage the shorts I’ve been creating for YouTube in multiple places.
And I noticed something for the first couple of videos I posted over there. I got a fair number of views on TikTok, but no real interactions. On Instagram, I didn’t get many views, but I had multiple people comment and seem interested. Really made me think about how I think about distribution in terms of getting to the right audience vs the biggest.
I’m also planning on reaching out to a couple of people running Build In Public courses and offering the product to their students for free. This will help me hopefully increase usage and possibly gain extra visibility. There’s a good sized audience between the two people I’ll extend the offer to, so maybe that will help things grow a fair bit.
Finally, I’m planning a Product Hunt launch next week. Still not sure exactly which day, but I’m leaning towards Wednesday. That should drive some guaranteed traffic, although we’ll see how much it converts. Having a freemium offering should help get people signing up and trying it out at least.
Content Corner
It’s been a really great week for content. I’ve listened to a few podcasts that really kinda blew my mind this week.
First up, My First Million with Connor and Breanna Price.
I think there are a few really important ideas to take away from this: they have a great strategy to get people off of TikTok and on to their Spotify. They also own their catalog, which means they don’t have to pay a label for anything. Finally, they keep his music that he’s done solo (can’t do it for his collabs that touch a label) in a playlist of DRM-free music. This means streamers/YouTubers/etc can include his music without fear of copyright strikes. That’s going to really pay off in the future. I love that approach a ton. I don’t think most people have that vision.
Next up, Arvid Kahl talked with Danny Postma.
I really enjoyed this conversation as well. I think it touched on a lot of really important ideas, but something that really resonated with me: Danny tends to block his competitors on Twitter/X, because if he doesn’t, he focuses too much on what they are doing instead of what he needs to do. There’s a ton of other wisdom in there too. I recommend listening to it.
Finally, KP talked with Calum Johnson on the Build In Public podcast about starting your own podcast.
I loved how they talked about digging into your own authenticity and throwing out the scripts that people love to use. It’s not about doing things the way everyone else does, it’s about finding ways to tap into your authentic voice and sticking with it long enough to be successful.
My Content
I’ll start by sharing two of my videos from the past week that underperformed, but I think are important in two different ways. First up, a video on the ethics of using AI
I think it’s an important thing to think about. YouTube seems to disagree.
Next, I’ve got a video that I spent quite a bit of time on. Not as important a topic, but I thought it was fun. Meet the super team behind your favorite Indie Hackers:
To end, I’ll share my video from last week that overperformed my averages from quite a bit (I think it’s all about the word “sucks” in the title…)
What happens when you launch a product and it sucks? I go into what my approach was when I launched buildinpublic.coach initially and got very little usage from the people who signed up.
That’s it for this week! See you next week!
~Leo