
Hey there,

I hope you had a great week. I [spoke at a conference](https://twitter.com/fintechdevcon/status/1562477596343816193) called Fintech Devcon last week in Denver, Colorado, about how we use Friction Logs to make our products better at Stripe.

It was the first time I had traveled for work in a while, and I found it extremely refreshing. The recording for my talk should be available soon - I'll be sure to share it once it is live.

I was also able to explore Denver a bit while I was there. The River North Arts District (or Rino) was great fun - I can see why people are moving there in droves.

## Learn with me

📝 Finding balance as a remote worker - I wrote an article on the Third Space and why it is important for remote workers finding balance. It was a super fun project to work on and I'm excited to share it with you. It's called "[The third place - the secret to a happy life for remote workers](https://mikebifulco.com/posts/remote-work-and-the-third-place)".

📹 Supercharge your video call setup - This [twitter thread by @theKevinShen](https://twitter.com/theKevinShen/status/1565791729906745348) shares some helpful tips for improving your video setup for calls. This one is really useful if you're doing a lot of calls with other people for work. Your presentation on video can have a real psychological effect on the people you're meeting with!

🧘‍♂️ Write more, write better - I've been working on developing a daily writing habit. This has helped me become more aware of how I use language which is making me more effective at writing great content for my job, and for Tiny Improvements!

Thanks for reading Tiny Improvements. If you found this helpful, I'd love it if you shared it with a friend. It helps me out a great deal!

Until next time - be excellent to each other!

-Mike


Balance for remote workers


## A Quick Win with text-wrap: balance

Here's a ludicrously simple tip to make your website more readable: use the `text-wrap: balance` CSS property. It's one line of CSS, but it makes a noticeable difference in how text appears, especially for headings.

### Why Use text-wrap: balance?

When text breaks across multiple lines, it often creates uneven line lengths, which can feel awkward to read. The `text-wrap: balance` property ensures that text lines are balanced, creating a more visually pleasing and readable layout.

Take a look at these two versions of a heading:

**Before:**


### What is text-wrap: balance?

`text-wrap: balance` is a CSS property designed to improve the visual balance of text that spans multiple lines. By default, when a block of text wraps onto a new line, the lengths of those lines can vary significantly, creating uneven and sometimes awkward breaks.

With `text-wrap: balance`, the browser automatically adjusts the wrapping of text so that the lines are as close to equal length as possible. This makes the text easier to read, and looks great - particularly for headings, short paragraphs, and blockquotes.

### Key Points:
- **Purpose**: Ensures evenly distributed line lengths in multi-line text.
- **Use Case**: Best suited for headings, short paragraphs, or any text that benefits from visual balance.
- **Compatibility**: Works on most modern browsers and gracefully falls back to default text wrapping behavior where unsupported.

By enhancing the balance of text, this property helps to create a polished and professional design with minimal effort.

## See text-wrap: balance in action

Here's a quick side by side example of a heading before and after using `text-wrap: balance`:

<heading className="text-3xl font-bold text-purple-700">This is a heading that doesn't have the rule applied; text will break onto a new line unevenly</heading>

<br />

And by contrast -

<br />

<heading className="text-balance text-3xl font-bold text-purple-700">This is a heading that will feel nicely balanced if it spills onto a new line</heading>



(Note: you may need to resize your browser window to see the difference)




## How to use it: CSS

One little line of CSS does the trick:

```css
h1, h2, h3, h4, h5, h6 {
  text-wrap: balance;
}
```

## How to use it: Tailwind

Recent releases of Tailwind have built-in support for `text-wrap: balance`. You can start using it with the `text-wrap-balance` class.

```
<h1 class="text-wrap-balance">
  This is a heading taht will feel nicely balanced if it spills onto a new line
</h1>
```

### Browser Support
According to Can I Use, `text-wrap: balance` is supported by approximately 90% of browsers. While there are exceptions (e.g., IE11, Opera Mini, and some niche mobile browsers), the fallback is harmless: the text will wrap as it normally would without the property.

### Why It Matters
Legibility and readability are critical for user experience. There's even a psychological principle—the [Aesthetic Usability Effect](https://mikebifulco.com/newsletter/aesthetic-usability-effect) - that suggests people perceive more attractive designs as easier to use. Balancing your text lines is a subtle but effective way to make your site look polished and professional.


## Try It Today
Improving the look and feel of your site doesn't have to involve a complete overhaul. Small changes like text-wrap: balance can make a big difference. Add it to your site, and give your visitors a better reading experience—one balanced line at a time.



<details>
  <summary className="text-lg font-bold inline ml-4 cursor-pointer text-pink-600">
      Video Transcript
  </summary>

  [00:00:00] **Mike Bifulco:** Hey, what's up friends. This is Mike from tiny improvements. I have a ludicrously simple tip for you.

  [00:00:03] That's going to make your sites look better with one line of CSS. And I cannot believe how easy this is, which is why I'm making a video about it.


  Side by side Demo: Single post

  [00:00:10] **Mike Bifulco:** What we're looking at here is my website. I've made a change. That is exactly one line of CSS between this and the version you're about to see.

  [00:00:16] And I think the change is tremendous. Take a look at this.

  [00:00:21] Again, version one. Version two.

  [00:00:23] Version one. Version two. Version one. How to set up self healing, URLs, and next JS for better. SEO. Version two, how to set up self healing, URLs, and next JS for better SEO. I think that's a little more readable. Don't you?

  [00:00:38] Maybe he doesn't pronounce enough here. Let's look at a more detailed example.


  Side by Side example: Lots of posts

  [00:00:41] **Mike Bifulco:** this is my newsletter page, where I list off many newsletters that I've written in the past.


  Subscribe to 💌 Tiny Improvements

  [00:00:45] **Mike Bifulco:** If you're watching this and you don't subscribe already I would love it if you did. I write a weekly newsletter called tiny improvements with product tips for startup founders, product builders, and indie hackers, based on my experience, building startups in the past and currently building my startup Craftwork, which is a Y [00:01:00] Combinator backed startup.


  Looking at the change

  [00:01:00] **Mike Bifulco:** now let's take a look at this change here.

  [00:01:02] It's much more pronounced when you see it multiple times.

  [00:01:04] So, what we're looking at is CSS change that brings balance to text when it breaks across lines. And that is almost exactly what the change is. It is CSS, text balance.

  [00:01:14] If I pull this up on the Chrome for developers site, you can see from my pal, Adam Argyle, a super straight forward tutorial on what this thing looks like.

  [00:01:24] And the change is pretty easy to see when it's animated like this. Although it's a little hard to describe what's happening is the balanced version is trying to make the text break into lines that are roughly evenly sized.

  [00:01:33] So it has a lot more visual balance. It's a little easier to parse because your eyes and jumping from one line to the next isn't quite as far of a distance on the balanced version versus the unbalanced version. It also works with Arabic text, right to left left to right. It doesn't matter.

  [00:01:46] There are some limitations here.

  [00:01:48] This balance rule works on up to six lines, which means that if you apply it to everything, it won't work for long paragraphs, for example. But the nature of this rule really applies to headings. This is more or less the way that this should be applied [00:02:00] and I'll show you how I did it on my site.

  [00:02:01] Next.


  Implementation details in Next.js with Tailwind

  [00:02:01] **Mike Bifulco:** This is what you have to do to make this thing happen on your site, where you're implementing headings. You need to add a text balance rule to it. This has done with tailwind. So it's applying the text wrap, colon balance CSS property to my headings. And this component happens to apply to all the headings on my site. So it was super easy to do, but if you had to do it somewhere else, it would look something like this. Const balanced, heading. Equals. Return H one style equals there you go.

  [00:02:31] Tax rate balance. This would do it. If you need to do it, one-off for some reason, but you could also apply this to block quotes. Like the Google article explained. So const balanced block quote equals tax wrap balance, or in tailwind land, you could do. Class name equals text balance, not work just as well.

  [00:02:51] Matter of fact, let's go do this for the block quote on my site too.

  [00:02:53] I have this. Lovely block quote, component that I use to render block quotes on my site. And I think it's a reasonable thing [00:03:00] to add text balance here as well. So we're going to do that and that will improve the aesthetics of my site. That's super nice. It's very subtle. People may never notice that this happens, but it should make it so that when your title's going to a second line, the look that much nicer, pretty cool, right? I thought so.


  ## Can I use it yet though?

  [00:03:13] **Mike Bifulco:** Last thing I know what you're thinking. Is this something that I can actually use yet? Well, the good news is pretty much, yes. According to can I use this is used by almost 90% of browsers on the internet, which is fantastic. My apologies to people who use IE 11 opera mini UC browser for Android QQ, browser, Baidu browser, or Caio S browser.

  [00:03:32] You're going to have to wait a little bit longer to see nice titles on my site, but in the meanwhile, the fallbacks. All back scenario for this is that your text will rep as it used to, if you're using this rule. So I consider this relatively safe to use.

  [00:03:44] It's also already supported by tailwind, which is fantastic.

  [00:03:47] I think you should go add this to your site and give it a shot. Legibility and readability are really important for user experience. And there's this thing called the aesthetic usability effect that the better looking your site is the better looking your product is.

  [00:03:58] The more people will [00:04:00] inherently like it. It's a scientifically proven thing that, the better looking something is the more people put value in it.

  [00:04:05] So why not? I don't want line change like this.

  [00:04:07] That gets you some real wins for free.

  [00:04:09] All right.

  [00:04:09] That's a quick tip. I hope it's helpful for you until next time.

  [00:04:12] My name is Mike Bifulco. This is a tiny improvements. Give them hell out there.

  [00:04:15] ​



</details>


CSS Tips - text-wrap:balance will make your sites feel better


In this video from my Tiny Improvements series, I'm diving into product analytics to improve newsletter signups on my personal site. Join me as I share my engineering process, from discovering a bug with my old provider to setting up an A/B test with PostHog to find the best-performing newsletter signup form. Despite fewer-than-hoped signups, I've gathered key insights to refine my strategy. Plus, a sneak peek into my next steps, like redesigning the signup popup and enhancing site SEO. If you're a startup founder or indie hacker, you’ll find plenty of actionable tips here!

## See the code changes I made in this video

Code changes for this video are available in this pull request are available at [mbifulco/blog#856](https://github.com/mbifulco/blog/pull/856)

## Timestamps for this video

00:00 Product Analytics for Engineers
01:18 The Results are in!
04:09 Code review in Next.js
04:49 I switched from VS Code to Cursor
05:50 Looking at the variant that won
06:31 Ship a variant directly from PostHog
07:30 A surprise I didn't see coming
08:05 What we did and what's next




The results of my PostHog AB Test are in!

See the code changes I made in this video



In this video, I discuss my recent efforts to improve the product analytics experience on my website. Over the past two weeks, I addressed a sign-up issue by switching to a new email service and implementing AB testing on my newsletter forms. I analyzed the results, but discovered a bug that prevented complete data collection. I walk through identifying the problem, show the fixes I made, and reset the experiment to gather accurate data. I also encourage viewers to check out the open-source code for my site and invite feedback and questions.


## See the code changes I made in this video
Code changes for this video are available in this pull request are available at [mbifulco/blog#840](https://github.com/mbifulco/blog/pull/840)


## Video Timestamps
00:00 Introduction and Recap
00:44 Analyzing AB Testing Results
02:32 Sponsorship Shoutout
03:01 Deep Dive into Data Analytics
05:41 Fixing Implementation Bugs
08:33 Resetting and Restarting the Experiment
10:28 Conclusion and Next Steps


How I found a missing change on my next.js site with PostHog


## Test, trust the data, and optimize your site for better results

In this video, I dive into A/B testing with PostHog to boost newsletter signups on my site. I walk you through my process—from identifying issues with signups, switching email platforms, to experimenting with different UX elements. I'll show you how to set up and run A/B tests, and explain why trusting the data is crucial for making smart, data-driven decisions. If you're a developer or UX enthusiast, this video will help you refine your own website's user experience.

The code for my site is available on github - you can see the Pull Request for the changes associated with this video here: https://github.com/mbifulco/blog/pull/832

## Watch the first two episodes in this series:

1. [Product Analytics for engineers: Debugging a conversion problem on my next.js site](https://mikebifulco.com/posts/debugging-a-conversion-problem-on-my-nextjs-site)
2. [Fixing the bug behind my Conversion Problem](https://mikebifulco.com/posts/fixing-my-conversion-problem)

## Video Timestamps

- 00:00 Introduction and Series Overview
- 00:35 Identifying the Signup Problem
- 00:59 Switching to Resend
- 01:11 Exploring AB Testing
- 01:19 Sponsorship by PostHog
- 02:11 Website Elements and Hypotheses
- 04:07 Implementing AB Tests
- 05:03 Setting Up PostHog Experiments
- 07:20 PostHog Implementation Details
- 08:31 Testing and Analyzing Results
- 10:52 Tell me what you think!

<details>

  <summary>

    <b>Transcript</b>

  </summary>

[00:00:00] **Mike Bifulco:** Hello, and welcome back to tiny improvements. My name is Mike And since we last talked, I've been working on a problem on my website. This is the third video in a series that I'm doing on trying to get people, to sign up for my newsletter.

[00:00:11] If you're new here, I read a weekly newsletter called tiny improvements.

[00:00:13] That is meant for product developers, startup founders, indie hackers, people who are interested in building products. Where I share the things that I've learned, building products that companies like Google and Stripe and Microsoft, and more recently building a venture capital backed startup I'm co-founder of a company called Craftwork. We went through Y Combinator last summer. And so this is my way of giving back. I'm trying to give away the things that I know for free, because I've been really lucky in my career.

[00:00:34] And I want you to be as lucky as I have been. So since you're here, let's talk about where we've been in the first video. I discovered that I had a problem with signups and people weren't subscribing. I was using convert kit for my newsletter at the time and was having some sort of mystery problem with signups.

[00:00:47] So I added a bunch of telemetry to show me when signups were happening and to try and detect failures from that experience, I found that. I was getting some mystery failures with convert kit in my implementation of my newsletter signup form. And that's [00:01:00] what led me to what I did in the second video, which is where I replaced convert kit with Resend.

[00:01:03] So now I use Resend email platform for both sending my newsletters and getting people to sign up for my mailing list. That's been really nice so far. In this video I'm exploring the next thing I wanted to get into, which is AB testing stuff on my site.

[00:01:16] So let's take a look at that and see what that actually means in practice.

## Sponsored by Posthog

[00:01:19] **Mike Bifulco:** Hey, all this video series is sponsored by PostHog. After seeing my first video, they reached out about a sponsorship, which is really cool.

[00:01:25] And frankly, I'm flattered to see it. I was planning on making these videos either way and that's because I really enjoy PostHog. I also really enjoy building products and seeing things improve over time. . I wouldn't take a sponsorship from something that I don't believe in and don't use, I hope you're enjoying this series and I hope you're learning something.

[00:01:39] If you're really into it, it might also be worth giving. PostHog a shot. One of my favorite things about the platform is that they're free tiers, unbelievably generous. You can get a whole lot out of PostHog for free before paying anything to decide if it's worth it for you.

[00:01:50] There's also a self host option. If you want to go that route and PostHog also has a ton of open source software, which means you can have influence over the product itself.

[00:01:58] I think that's really amazing. And the team has been [00:02:00] super responsive so far. So if you want to check it out, go to PostHog dot com slash Mike Bifulco. There'll be a link in the description. It will let them know I sent you and it means a lot to me.

[00:02:09] Thanks so much. I hope you enjoy the rest of the video.

[00:02:11] Okay, this is my site and this is one of the things that I'm interested in testing to see if I can make a difference in the amount of people signing up.

[00:02:17] When you visit my site, read one of my newsletters or any of the articles on my site. There's a bunch of things that are made available at the top of the pages, a cover image and a title things you're probably used to seeing, I have this nice dynamic. Table of contents, which is really helpful I also record audio versions of some posts, which is why we see this podcast embedded here. if we scroll through the post, when you get towards the end of it, a couple of things happen.

[00:02:36] This pop-up, that just flew in from the left is a thing called polite pop. This is where the first call to action typically comes for someone to subscribe to my newsletter. I've got a couple of hypotheses on how we can do better here. What I want to do is test the copy.

[00:02:48] That's rendered here to see if there's maybe a different treatment, that we can show. To people to get them more compelled to sign up. I might also choose to remove this. It kind of is just an ag and I think people are really well-trained to just dismiss pop-ups and [00:03:00] try and get them out of their face.

[00:03:01] And it's a little bit of a rude experience while you're on a site. I see loads of people coming to my site and clicking this. No thanks button. I've got tons of telemetry on that. That may not be long for this world.

[00:03:09] The other thing that I see, and I think this is worth testing straight away is at the bottom of each post.

[00:03:14] I have this giant banner, which is this big striking thing that says ship products that matter. And I really like this. This is something I workshopped for awhile, trying to come up with a slogan or phrase, something that was sort of attention grabbing. That's supposed to give people a reason to sign up.

[00:03:28] However, I think what I've done here is I built something that is so different from everything else. That's that seen on my site that. By the time you get down here, as you scroll down and you get this great big dark banner that appears, it's pretty obvious that the article is over and then I'm not giving you any more value here. And so what I want to test is if I can replace this with something that's a little more subtle.

[00:03:47] So maybe not a giant banner like this, but something that says, Hey, I write this cool newsletter. I think you might like it. Please subscribe. And maybe Tiki into people's psychology a little bit to keep them reading on the page. watching back some replays of people visiting the site. [00:04:00] It's pretty common to see someone get to about here and go away.

[00:04:02] And if they leave before they get down here, I've lost. So. we're going to try a couple of things here and that's. Where AB testing comes in.

## A-B Testing in PostHog

[00:04:08] an AB test is exactly what it sounds like. I have version a version B and sometimes C D E F a number of different versions of one bit of user experience that I want to test against each other to see how they perform. The idea here is to see that if I do this versus that is one going to get me more people signing up for my newsletter or not.

[00:04:24] **Mike Bifulco:** And we can kind of test some hypothesis from there. Now there's some really important statistics involved in how this works. What I really like is that when you collect data, the math will tell a story. And if you're willing to trust the math and listen to what the statistics say, you can probably make a good data driven decision to move yourself from doing well, to doing better and incremental change over time will do really well.

[00:04:46] But you have to be comfortable with trusting what the statistics say. And let me be the first to tell you that there's a lot to learn about that, and it can be really counter-intuitive. I am far from a statistics expert. I have learned a lot of mathematics in my life.

[00:04:57] You have to learn to trust the statistics. [00:05:00] And so that's what we're going to try to do here. Let me get off my high horse. Let's talk about this. So this is my first experiment is called newsletter banner treatments. So you give it a name, which is a human readable thing, and then our feature flag key.

[00:05:10] So this is the key that we'll use programmatically to make decisions based on this experiment. So I gave it newsletter banner treatments as the key. The description here again is a human readable thing. It's meant for me when I go review these later on to see what this experiment was all about.

[00:05:23] So I may fill this in with more detail, but the idea is I want to test different UX presentations of the newsletter banner that appears at the bottom of articles. To see which one performs best.

[00:05:32] By default PostHog will give you a control, which is the fallback value. So the control is meant to be in scientific method speak.

[00:05:39] That means the thing that we know how it performs and the baseline for the experience right now, that's that big dark banner that I show on my site that I was just showing you that people sort of. Seem to scatter away from. I'm running two tests against that in this case. So I'm going to test a simple implementation and a hero implementation. Those are bits of UI that I've built alternatives for.

[00:05:58] And I'll show you what they look like. [00:06:00] So if I come through and I hit, continue on this experiment, we'll see that PostHog does a little bit of thinking. And it goes to do some give me some numbers on roughly what this would look like. And what I want to check is my newsletter subscribed event.

[00:06:11] So I'm hoping to see more of these newsletter subscribed things happening. With one variant of this versus the other, they'll probably all perform slightly differently. We're going to pick the one that wins out of the three. So. I've picked the event that I want to use. If I go to the next page here, what I get is the goal for the experiment.

[00:06:28] So this now shows me everything in one view. What it's saying at the very top is this is a draft. I haven't actually started it yet. By results are not significant, meaning I haven't collected enough data to make a decision. And the feature flags name is here.

[00:06:39] Down here, we get implementation details from PostHog, which is literally code snippets that you can use in your application. So you can switch through these and see what they look like.

[00:06:47] There's react native there's no, there's Ruby, all kinds of stuff in here. What's really cool is they show you exactly how to do this. So you load the feature flag using the feature flags name. If you have a user ID, like an email address, if someone's logged in, you can pass that [00:07:00] into, they always get a consistent experience per user. And then you make a decision based on what comes back. So the feature flag will say for this person render a simple one. It might come back and say for this person render a hero one and you do different things based on that. So what I'll do is I'll hit launch on this thing, It'll sort of thing for a second and go live and I'll, show you what the implementation looks like, next.

[00:07:20] Okay, implementation time. So I'm using post hogs library called PostHog JS.

[00:07:26] Thanks sports, some really nice react components, which I'm using in this implementation here. So this experiment name, newsletter, banner treatments is what we just saw on the site. These are the three different experiment variants.

[00:07:35] So these are the things that the experiment will tell me to render for different users. And we'll collect data on that. And I'm using a thing called PostHog feature, which is a react component exported by post hog, which takes in the name of the flag. As well as the part of the experiment to match.

[00:07:50] So what this is saying is we're going to do this with the newsletter experiment. So if it comes back and says control, we're going to render this thing, the fancy component, which is the newsletter banner that's currently shown on [00:08:00] the site with the dark background.

[00:08:01] I've created two other variants. The simple one here is based on a fairly standard tailwind component library. And this one called hero is one that I was using on other pages on my site, but wasn't consistently showing on the bottom of articles.

[00:08:13] now I'm taking all three of these things. And for about a third of users, I'll show each one of these and we'll track. How many of them convert to newsletter subscriptions over the next. Week or so, and we'll see what comes back.

[00:08:23] Next thing to see is what this looks like in the browser, because PostHog also provides some tools for you to be able to poke around with these and make sure it works well before you roll it out to your users.

## PostHog's A-B Testing tools in the browser

[00:08:31] **Mike Bifulco:** All right back here on post-talk and here's what I want to show you.

[00:08:35] from any page on PostHog, you can get to this handy thing called toolbar. If I click into this, this gives me the ability to launch a PostHog specific toolbar within the context of my site. So it doesn't really fancy injection of content. And it renders PostHog using what I believe is a Chrome extension.

[00:08:50] And so if I click on this, it's going to load my local host, which I've told is where I'm running my site. This is what the tool bar looks like.

[00:08:55] Typically you'll see it. Like this, when it first comes up and we're really interested in this [00:09:00] feature flags. Functionality here. So there's a bunch of things here. You can do actions and heat maps and all kinds of stuff like that. But let's take a look at feature flags first. I'm going to navigate to one of my newsletters, the most recent one, we were just checking this out and I'm going to scroll to the bottom and I'm going to go to this feature flags thing here and already something looks different, This is not the banner that was showing before. And this is a sign that my experiment is working. you can see here, I can manually override the experiments that I'm seeing personally in development.

[00:09:25] And as I toggle through these, I get three different presentations. this one might look familiar. This is the hero one I had before. This is the simple implementation that I wanted to take a look at. And then here's the control implementation. the important thing to test here is we're trying to test things that could produce the same output against each other. When users come to this page, I want to make sure they can submit this form and get the output that I want from each of them, which is that newsletter subscribed event.

[00:09:46] So from this control one, This is a newsletter subscription form. First name and email address is exactly what we want to see the simple form, same thing. First name, email address, and then the hero form. Once again, first name and email address. So. [00:10:00] All the primitives are here. I have three different variants of this thing.

[00:10:03] The next thing to do is to set this live and just record some traffic for awhile. there are some statistical minimums. We have to get a certain amount of users interacting with each one of these, to be able to say whether it worked or not. And the next part is just waiting. So we're going to spin this up on PostHog.

[00:10:17] We're going to collect some data over the next week or two, however long it takes, and then we'll start to make some decisions based on that and go into the next step from there.

[00:10:23] So that's all there is to it. It's a little bit of implementation. It's a little bit of work to set up the three different variants, but now I need to trust the statistics. I need to listen to what comes back from post-hoc as the experiment runs. And tweak my site design based on the data that comes back. My guess is it'll probably take about two weeks for this to come back with meaningful results. But lucky for me, I can run a couple of experiments in parallel, as long as the two different experiments don't interfere with each other. I can probably still draw a valid conclusion from it. So over the next week or so I'll watch this and I'll see what happens.

## What do you think?

[00:10:52] **Mike Bifulco:** If you've been hanging on for the whole series so far, I'm really curious what you think of this.

[00:10:55] Is this something you would find useful for your site? Do you have an idea of experiments you'd like to run? Do you have [00:11:00] questions about this stuff?

[00:11:01] I spent a bunch of my career working in design and user experience. So this sort of thing is a process that I really enjoy, but I can understand that it might be frustrating too, especially if you just want to get to results.

[00:11:10] If you have questions for me, feel free to drop them in the comments. . Your ideas are always welcome here. There's no criticism too big or too small. I'd love to hear what you're thinking. Thanks so much. My name is Mike bi-fold co this is tiny improvements and we'll catch up for the next one.

</details>


Unlocking A/B Testing with PostHog: Improving Newsletter Signups

Test, trust the data, and optimize your site for better results

Watch the first two episodes in this series:

PostHog's A-B Testing tools in the browser


Watch [Product Analytics for engineers: Debugging a conversion problem on my next.js site using @PostHog](https://youtu.be/4nWcB1a8M1s) on YouTube

Thanks so much to PostHog for sponsoring this video and supporting my work! Give it a whirl over at https://posthog.com/mikebifulco and tell 'em I sent you!

## Summary

In part two of my adventure with product analytics, I tackle a significant drop in my newsletter subscribers. Join me as I dive into the data using PostHog, identify and fix critical errors, and ultimately switch from ConvertKit to Resend – a more developer-friendly email platform.

Along the way, I'll show you how I integrated tRPC into my Next.js site for a type-safe API, created hooks for better data management, and set use posthog to chase down the root of the bug.

Whether you're a startup founder or an indie hacker, you'll find valuable insights to improve your product. Plus, transparency is key – this video is sponsored by PostHog, but all opinions are genuinely mine. Let's build something great together!

## Video Timestamps

00:42 ConvertKit Stats

01:06 Subscription Form

02:45 Sponsored by Posthog

03:16 Resend broadcasts

03:36 tRPC

03:54 tRPC Demo

04:43 Subscription form code walkthrough

06:27 useNewsletterStats

07:33 What I'm seeing in posthog

08:37 What's next

09:45 Next PostHog tests

10:11 What do you think I should do next?

10:31 Wrap up

## Links

- Watch part 1 of this series: [Product Analytics for engineers: Debugging a conversion problem on my next.js site using @PostHog](/posts/debugging-a-conversion-problem-on-my-nextjs-site)


Fixing the sign-up problem on my site using PostHog Product analytics


Watch [Product Analytics for engineers: Debugging a conversion problem on my next.js site using @PostHog](https://youtu.be/4nWcB1a8M1s) on YouTube

## Summary

I've been experiencing a problem with my personal website, mikebifulco.com, where my newsletter sign-ups have mysteriously plummeted. In this video, I take you through the process of diagnosing and debugging this issue using Posthog, a powerful product analytics tool.

We'll start by taking a look at the site, identifying potential problem areas, and then diving into the code. I'll show you how I implement Posthog to track user interactions and gather valuable insights. We'll explore how to handle form submissions, track events, and use Posthog’s features to pinpoint what's going wrong.

Throughout the video, I'll share my thought process, hypotheses, and the steps I take to experiment and test different solutions. If you've ever had a non-bug issue in production, this video is for you. Let's get to the bottom of this conversion rate mystery together!

You might be interested in this video if you want to learn about debugging and optimizing website performance, especially using tools like Posthog. Relevant technologies and subject areas include React, Next.js, form handling, product analytics, serverless functions, and conversion rate optimization.

## Video Timestamps

0:00 Product Analytics for engineers: Debugging a form conversion rate problem on my website with Posthog

01:05 Site walkthrough

01:30 Are people annoyed by the popup?

02:53 Long form article pages

04:02 Looking at the code

05:20 Dive into the code

07:13 Newsletter form refactor

10:45 Serverless function for newsletter subscribe

12:26 The results in Posthog

13:29 What we did, and what's next

14:43 Why Posthog and product analytics are helpful

15:41 Epilogue: We learned something! Next steps...

## Links

- Watch [Product Analytics for engineers: Debugging a conversion problem on my next.js site using @PostHog](https://youtu.be/4nWcB1a8M1s) on YouTube
- Part 2 of this series: [Fixing the sign-up problem on my site using PostHog Product analytics](/posts/fixing-my-conversion-problem)


Product Analytics for engineers: Debugging a conversion problem on my next.js site


Watch [The SECRET you need to hear about the difference between Taste and Skill](https://youtu.be/Ym6HQopqe-4) on YouTube

## Summary

Mike discusses how important it is for creatives, especially those building startups and indie projects, to understand this gap and use it as motivation.

If you're on this journey, it is to your benefit to draw from a broad array of influences, and relate it to your creative output.

In addition to the critical step of building your chops by practicing your craft, gathering inspiration from others who excel in the field is a game changer. This is borrowed from my extensive experience in building software. and a career spent learning about the critical role of good design.

If you're aiming to produce better and more user-friendly products, this video offers valuable guidance and inspiration to help you along your creative journey.

## Video Timestamps

0:00 Ira Glass on Beginners
05:35 Mobbin
06:03 Wrap up
06:31 💌 Tiny Improvements, my newsletter

## Links

- Watch [The SECRET you need to hear about the difference between Taste and Skill](https://youtu.be/Ym6HQopqe-4) on YouTube


Your tastes will always outpace your skill, and what to do about it


OpenApi.tools is an Open Source and community-driven list of high-quality, modern tools for OpenAPI that we've been maintaining over at APIs You Won't Hate for the past few years, and it's overdue for a rebuild.

🔧 In this session, I'll be using Astro.build to recreate the openapi.tools site, with a focus on building a highly performant, maintainable, and SEO-driven experience for the next version. If i'm lucky, we'll even get into building some fun new features to help our community find the right tools for their use case with OpenAPI

I'll be building with:

- Astro: https://astro.build
- Hosting through @VercelHQ
- Visual Studio Code
- TypeScript
- React.js
- Markdown
- Fathom analytics @FathomAnalytics
- PostHog @PostHog
- Tailwind CSS @TailwindLabs

📅 Mark your calendars and join me at 10am EST for an insightful journey through the intersection of design and development. If you dig SEO and markdown, or you're a React developer, JavaScript enthusiast or TypeScript pro, or you've been curious about what Atsro brings to the table, you'll enjoy this one!

Don't miss out on this opportunity to learn, create, and connect. Be sure to subscribe and click the notification bell to stay updated on this and more exciting content from Craftwork.

For additional resources like my newsletter for founders and product builders, 💌Tiny Improvements, visit my website at https://mikebifulco.com/newsletter


Rebuilding an open source content-rich site with Astro, TypeScript, and React


Join me for an exciting live stream where we dive into the world of design and development at Craftwork! 🎨💻

🔧 In this session, we'll explore how we use a variety of image templates to elevate our software's appearance. From stylized portraits of our talented team members to vibrant paint swatches and eye-catching Open Graph (OG) preview images for our website, we'll unveil the secrets behind our visual magic.

🖌️ We'll kick things off by using the power of Figma, our design tool of choice, to craft stunning templates. Then, we'll dive deep into the development side, where we'll transform these designs into dynamically generated images. How? With the incredible Satori library and the magic of Edge Functions on Next.js, brought to you by the team at @VercelHQ .

I'll be building with:

- Satori & Next.js from @VercelHQ
- Figma
- Visual Studio Code
- TypeScript
- Tailwind CSS @TailwindLabs

📅 Mark your calendars and join us on Jan 27th at noon EST for an insightful journey through the intersection of design and development. Whether you're a React developer, JavaScript enthusiast, TypeScript pro, designer, or simply curious about UX and tech, this livestream is for you!

Don't miss out on this opportunity to learn, create, and connect. Be sure to subscribe and click the notification bell to stay updated on this and more exciting content from Craftwork.

See you there! 🚀🎉


Using Satori and Next.js to generate dynamic images


## Notes about today's Livecoding stream
This is a YouTube Livestream, where I'll be working on integrating Resend Broadcasts into my next.js site to use for sending newsletters. Resend released a new set of features this week that makes it easier to send mail to large audiences.

I'll be building with:

- Resend
- Next.js
- TypeScript
- Vercel
- Tailwind CSS

With Substack recently catching headlines for their boneheaded content moderation policies, lots of creators and developers have been looking at other tools for managing their mailing lists. Between Beehiiv, ConvertKit, Ghost, Mailchimp, and some of the other usual suspects, there are many options to choose from.

Let's take a look at Resend's new API to see how it works, and if it's a good fit for my workflow!

## Don't forget to subscribe

If you're interested in more content like this, please subscribe to my **[YouTube channel - @mikebifulco](https://www.youtube.com/@mikebifulco?sub_confirmation=1)!**


Configuring Resend Broadcasts for newsletters on my Next.js website


Building a website using the Remix.run web app framework? This tutorial will teach you how to add privacy-first analytics to your Remix site with Fathom.

## How to set up Fathom on your Remix site

1. Sign into your account with [Fathom](https://usefathom.com/login), and create a new site.
2. Open the settings for that site, and look for _Script settings_. Hit the clipboard icon to copy the Script tag for your site.

   <Image
     publicId="posts/add-fathom-analytics-to-remix/script-settings"
     alt="Script settings for one of my sites on Fathom's dashboard"
     caption="You site's embed code will be on the script settings page. Hit the clipboard button to copy the code"
   />

3. Open up your Remix site in your IDE of choice. open `root.tsx` (or `root.jsx` if you're using JavaScript instead of TypeScript), and paste your embed code right before the `<Script />` tag:

```jsx
return (
  <html lang="en">
    <head>
      <Meta />
      <Links />
    </head>
    <body>
      <Layout>
        <Outlet />
      </Layout>
      <ScrollRestoration />

      <script
        src="https://cdn.usefathom.com/script.js"
        data-site="TRSSCIOR"
        defer
      />
      <Scripts />
      <LiveReload />
    </body>
  </html>
);
```

<small>
  
  _Note: ☝🏽 I've edited the `<script>` tag here to be self closing, rather than having `<script></script>`. It should work either way._

</small>

4. Start your site with `npm run dev` or `yarn dev` and visit any page (you may have to hit reload if you had the page open already). You should see this reflected in your fathom dashboard in real time!
   <Image
     publicId="posts/add-fathom-analytics-to-remix/tracking-results"
     alt="You'll know it's working when you see a live visitor on your fathom dashboard"
     caption="You'll know it's working when you see a live visitor on your fathom dashboard"
   />

## Try out Fathom - it's great!

I've been using Fathom for years, and am a happy paying customer. If you'd like to give it a shot, you can get $10 off of Fathom when you **use my referral link [https://usefathom.com/ref/DPSSYB](https://usefathom.com/ref/DPSSYB)** to get started.

## If you enjoyed this video

- I wrote a post when I initially adopted fathom, called [Why I switched to Fathom for analytics](https://mikebifulco.com/posts/why-fathom-analytics)
- Make sure to [subscribe to my YouTube channel](https://www.youtube.com/channel/mikebifulco)


#video: 12 posts tagged