My 2023 Annual Review

2023 is the most significant year of my life so far.

I cannot think of another year with as many milestones as this.

On the personal front, my wife Ying Yan and I moved into a new house—our house—seven years after we “ordered” it, and we welcomed our son into our lives. We are now a little family of three, living in our own home. Thinking about this always gives me a warm, fuzzy feeling.

Professionally, SK and I launched Pebblely, which changed the trajectory of our startup, becoming profitable almost immediately and reaching a million signups in seven months. We were invited to speak at several events, including one in Dubai, and were featured by well-known companies such as HubSpot and Aspire.

Of course, in between the glamor are missed goals, mistakes, and struggles, which I’ll share below.

Reflecting on 2023 goals

Ramen profitability

After killing Dashibase because it wasn’t growing as fast as we would like, we set out to find fast-moving water. Our plan for the year, inspired by Pieter Levels, was to launch 12 ideas, kill the unpromising ones, and invest more in the rest. While it wasn’t an explicit company goal, I wanted us to reach a level of profitability to pay ourselves a salary. This was not because we desperately needed an income but because I believed it would help us psychologically in this long, tough startup adventure. More startups die from implosion than competition.

The first idea was Pebblely, a product photography AI. Our goal was a meager $200 in revenue by the end of January. Or we will kill it. We exceeded the target within two days. Every week, we thought we had reached the peak. Every week, we made more money than the previous week. This went on for months. By March, we decided to pay ourselves a salary. We agreed that, in the worst case, we can still put the money back into the business.

Goal achieved.

With Pebblely going well, we focused on improving it, instead of launching new ideas every month. But we did build a few more generative AI apps because we felt they could have a much bigger potential: Vispunk Editor (AI-first Photoshop), Vispunk Motion (AI video generator), and Vispunk Visions (AI image generator). Admittedly, we made perennial startup mistakes, such as not having a clear user and over-scoping (I suspect the former caused the latter).

With those lessons freshly imprinted on us, we launched Pebblely Fashion last month. It has a clear user (fashion ecommerce owners and marketers) and fewer things than we would like (works only for female tops). If it has everything we wanted, we likely over-scoped.

Overall, I’m delighted we hit this goal out of the park. But after our year-end retrospective, we felt we should have aimed higher. I’ll share more in the next section (Goals for 2024).

Tweet daily, publish weekly

These were two separate goals but I put them together here because I failed terribly at both. I wanted to tweet daily to encourage myself to share my thoughts more and to publish a weekly newsletter to improve my writing. But both goals went out of the window once Pebblely started taking off.

My justification was that the time is better spent growing Pebblely. For example, I went from not having a TikTok account to creating more than 100 videos. But I now think I should still have kept working on these two goals. One, it’s important to always allocate time for self-development. Two, not 100% of the time “saved” was spent on growing Pebblely.

I’m going to write more regularly in 2024 but with a different motivation.

Exercise daily, sleep well, eat healthily

Being a triathlete for 14 years, I value staying healthy. I work out every day, sleep 7-8 hours every night, and eat vegetables and fruits. This keeps my mind sharp, motivation high, and stress low. In my last review, I wrote about wanting to stay physically and mentally fit.

I kept my streak for many months until, of course, the arrival of my son. While we could have hired helpers, we opted to look after him ourselves. The first few months were all about survival. We were (and still are) constantly sleep-deprived, exercising became an afterthought, and ice cream and brownies became a regular treat.

But this is a trade-off I was prepared for and willingly made. Children grow up so quickly, and they will never be 1-week, 1-month, or 2-month-old again. Days (and nights) are long, years are short.

Goals for 2024

Build a company

As Steve Blank wrote, “a startup is an organization formed to search for a repeatable and scalable business model.” While Pebblely has been successful by certain metrics, we don’t have a repeatable and scalable business model yet. An advisor nicely told us we have built a feature but not a product or a company yet. It stings because we know it’s true.

Before we even launched Pebblely, several well-intentioned friends told us our idea was undifferentiated and would be copied quickly. While that turned out to be true, we would still have launched Pebblely, given what we know now. What we would have done differently, though, is to figure out how to turn it into a company with the temporary early-mover advantage we had. Both SK and I felt we could have executed better and moved faster.

In our first year, we had an idea that went nowhere (Dashibase). In our second, we improved and built a cash-generating software (Pebblely). Next year, we aim to have a repeatable and scalable business model. A company.

Our big vision is to empower businesses to use generative AI media to grow. Pebblely Fashion will be our focus for a start. If you are an innovative fashion brand looking to adopt generative AI to create more content at a tenth of the cost, I’d love to chat.

The plan to achieve our goal is beyond the scope of this essay. But one key skill I want to pick up is B2B sales. I'd appreciate any advice or resources. (Shoutout to Cedric Chin for sharing his B2B sales reading program with me.)

Frame challenges as opportunities

The biggest struggle I had this year was dealing with challenges. Like most people, I would lose motivation after being stuck for a long time. And this year, I let that spiral downwards longer than I should have. I remember forcing myself several times to run outdoors to break myself out of the negativity. Or it took my wife asking me how am I really doing (I highly recommend asking your loved ones this question regularly). I used to be much more creative and resourceful around challenges.

The turning point for me, which was very recent, is re-reading Ryan Holiday’s The Obstacle Is the Way. Challenges should not be avoided but embraced. They are opportunities in disguise. This reframing doesn’t make the challenges any easier. It makes dealing with the challenges easier. Once we view them as opportunities to get better, instead of obstacles to stop us, they are like enemies in a game whom we have to defeat to level up.

In the movie Evan Almighty, Joan prayed for their family to be closer but her seemingly insane husband Evan chose to build an ark, apparently under God’s instructions. Frustrated, she left Evan, only to meet God (of course, it had to be Morgan Freeman) disguised as a waiter at a restaurant:

God: Let me ask you something.

If someone prays for patience, you think God gives them patience? Or does he give them the opportunity to be patient?

If he prayed for courage, does God give him courage, or does he give him opportunities to be courageous?

If someone prayed for the family to be closer, do you think God zaps them with warm fuzzy feelings, or does he give them opportunities to love each other?

If I ask to become better, I get opportunities to get better.

Get better at using AI

For someone selling an AI tool, I don’t use AI as much as I should. Besides Pebblely, the other AI tool I use often is the free ChatGPT (for getting copy variations, solving coding issues, and explaining legal terms). I also used A1111 for a bit when we were building Vispunk Editor.

I’m excited to get good at using tools like ChatGPT Plus, Midjourney, and Pika, and maybe even venture deeper and explore tools like ComfyUI. To ensure I achieve this goal, I plan to incorporate these tools into my work at Pebblely and on this blog (which is often my playground for testing new ideas and tools). For example, I used Midjourney to generate the images for this blog post.

If you have a similar goal of getting better at using AI in 2024, let’s connect!

Write daily, publish monthly

I will again attempt to write more regularly this year. The goal isn’t to become more creative or better at writing, though I might in the process. My main motivation is to improve a little every day, and I remember things best when I write them down.

I have begun to journal privately before I start my workday and before bed. I use the Day One app, which I have been using for a while. A private journal allows me to comfortably write about my life. It should also reduce the unnecessary stress of wanting to publish something perfect.

The downside, though, is I have less external pressure to polish my writing. So I plan to publish once a month to consolidate my learnings. But ultimately, I have to hold myself accountable for improving my writing.

Eat healthily, exercise weekly, sleep enough

Since I have a baby, I have to be realistic about my health goals. I don’t plan to win triathlon races but to be, for lack of a better word, a highly functioning person and to set a good example for my son.

While I think I have good self-discipline, eating healthily is a matter of setting up the right environment. I know, for certain, I don’t eat as many snacks if I don’t have any at home. So the game plan is to replace the unhealthy food at home with healthier options like fruits. Even if my self-control fails when I’m sleep-deprived, there is no junk to eat.

I’m trying to be creative with my exercise routine. I know it will be hard to do my usual long swims, rides, and runs. I’m replacing those with shorter but more intense static exercises that I can do in my living room before my son wakes up. (I have even used him as weights for squats.)

I used to subscribe to the idea that I must have eight hours of sleep to function well. I now think part of that was self-fulfilling. Over the past few months, I learned I can function with less sleep—to a certain extent. On days when I only get two hours of sleep, I’m an absolute train wreck. That I’ll admit. But I can still function well enough with five to six hours of sleep (and lots of coffee). And well enough is good enough. As my son grows bigger, this should get better (but again, I don’t want to rush to the future and miss the present).

Anti-goals for 2024

For once, I’m setting anti-goals. Things I will not do.

The common advice for doing things well is to do fewer things. But most people simply say no to things they would have not done anyway. Saying no should be painful. With that in mind, here are the things I’ll say no to in 2024:

1. Start side projects
2. Use Twitter and LinkedIn on my phone (I have deleted them and plan to replace this habit with reading with the Kindle app)
3. Scroll through social media without posting
4. Watch TV/Netflix/etc. aimlessly (instead pick shows and schedule time)
5. Go for friend gatherings (instead invite them over)
6. Attend events not related to Pebblely or generative AI

One thing I’m not sure about, though, is networking. Oftentimes, who we know can be more important than what we know. But spending too much time attending events and networking doesn’t feel as productive as creating something. And while I can be selective about the events I attend, the right encounters seem serendipitous. If anybody has any advice, let me know.

—

I hope you are ready for 2024 too.

Let’s go!

You can also reach my past annual reviews here: 2022, 2021, 2020, 2019, 2018, 2017.