Andrew Chen Archives

Hi readers,

Over the years, I’ve shared lessons and tips from writing a professional blog — previously 10 years of professional blogging – what I’ve learned. Recently, I’ve had some major changes in my workflow, and wanted to share more.

Pre-AI blogging workflow
My pre-AI workflow for blogging — what I’ve been doing for over years — looks something like this:

• Have interesting conversations at work or with friends/colleagues
• Randomly grab my phone, and email myself the title of a blog post
  • These have punchy opinions/titles that expand easily to blog posts – some examples:
    • “what needs to happen for web3 gaming to work”
    • “lessons learned from launching games that apply to web and mobile apps”
    • “why startups should ignore social media haterade”
    • etc
  • Recently, I’ve been using this app, Email Me, which is opens to a text box and it’s a one-button send (also nice for articles to read later, todo list items, etc)
• Later, when I’m doing email, I do one of two things:
  • If I’m motivated, start a blog post right away with the title
  • Or put it into my Notes app where I collect a long list of 100+ titles that are on the backlog
• I generally write on the weekends and try to do a full essay in one sit – usually take 2-3 hours
• Very very rarely do I go back and proof read or rewrite them (don’t blame me! I’ve been writing for hours!)
• Usually the essays are scheduled to go out Monday/Tuesday morning at 9am
• The day the post goes out, I do a light proof-read and then once it goes out, I write a tweetstorm directing traffic

This whole process has served me well for years. In the end, there’s no real magic to it — 99% of the battle is carving out a few hours on Sundays to write when there’s a many dozens of emails that need to be returned, Netflix shows to be watched, and games I should be “researching” for my job as a games industry investor.

But in the past year, I’ve experimented with changing my workflow to incorporate a number of new generative AI tools — particularly ChatGPT. I wanted to share some experiences with what it seems to be good at, and what it’s not.

Post-AI blogging workflow
After ChatGPT was released, I began to experiment with it in various ways — here are a few ways where it’s been useful.

First, writing from a blank page is hard. Sometimes it’s easier to just generate a first draft of something, even if it’s not great, just to get your creative juices going. I might use a prompt like this:

I’m writing a blog post about how Gen Z consuming more video, playing games, and other forms of rich and interactive content will mean that they will read less long-form text like newspapers and books. Write me a first draft of a 5 paragraph essay that explores this topic, but only use 2-3 sentences for the opening and closing so that it’s terse and punchy. Be opinionated.

And ChatGPT returns something that sort of makes sense? But reading the results more closely, it lacks examples and story-telling, and statistics or anything else. I don’t like Paragraph 3 here, since it misses an opportunity to talk about how much of YouTube is education purposes, for example. The writing is a bit too stiff and formal, and doesn’t match my own personal style.

In general, it’s just not that good.

But after staring at a page like this and finding all the flaws, I can then iterate pretty easily. I can put in a followup prompt to change out Paragraph 3 and connect it to a different point. Or I can just start editing it and rewriting it completely, and even if at the end 90% of it is different, it’s still better than starting with a blank page.

Brainstorming outlines, lists of topics and questions, and more
If ChatGPT is weak at actual writing, where I find it excels is as a non-judgmental brainstorming partner that doesn’t care if you have stupid ideas or if your writing is boring. It’s supremely useful for generating outlines of blog posts, making lists of topics, and creating a bunch of questions you can answer as inspiration, etc. In this use case, the AI doesn’t have to have a 100% hit rate on its content. If it creates lists and lists of content, but 20% inspires you in your actual work, then that’s a success.

For example, let’s say that you know you want to write a blog post about opportunities to build new apps in VR. You can ask something like:

I’m writing a blog post about new app ideas in VR. What are 15 questions that someone might ask me about this topic? Make them spicy, and if helpful, cite statistics to make the questions more interesting

It comes back with a mishmash of questions, some which are dumb and uninteresting — like the first, one, which is just asking what VR apps will be useful. But it also has a bunch of interesting ones on esports, content creation, corporate training, etc.

You could also imagine using this as a pretty helpful tool if you are making a new podcast and are interviewing guests. Or if you want to write a new essay that critiques some existing paradigm. (“Give me 10 questions from someone who is skeptical of of the value of AI safety”). And once you have an interesting question, it becomes easy to ask it to outline an entire blog post about it.

For example, I often find it useful to write prompts for outlines specifying specific topics and ideas.

Take the concept of K-12 education and VR – can you write an outline expanding on the promise and also skepticism around this idea? In particular touch on the concept of blending entertaining and gaming and learning, as one of the points. Create 6 sections, and put 3 sub-bullets under each. Use strong opinions and ask tough questions.

I’ll leave it to an exercise for the reader, but the result is not bad. It’s a decent starting point.

ChatGPT works great for outlines for long form writing also. And as many of you know, I wrote my book 3 years ago without the benefit of AI. The book is on network effects, and I ended up making an outline that acts in various stages, and uses examples from iconic tech companies. I can prompt ChatGPT with something like this:

Create an outline for a book on network effects for mobile apps and websites. Stage out the outline so that it starts on launching a product from zero, and the issues there, versus the stages of scaling the product, then hitting market saturation. Make sure you are able to weave in the issues facing network effects-driven products like social networks, marketplaces, dating apps, and collaboration tools. Write a 10 section book, with 3-5 sub-bullets. The first sub-bullet in each section should always be an example of an iconic tech company, like Uber, Airbnb, Dropbox, Tinder, Slack, or others. Please make the comprehensive outline for a book focused on network effects targeted at professionals working in the tech industry

The result actually isn’t bad. It’s not the way that I wrote it, since it starts to dive into each category of product rather than expanding on the broad concept. But again, it’s not a bad starting point. I could easily have imagined using this for brainstorming and prototyping purposes, iterating through different versions in describing the idea. It might have helped me work at more of a conceptual level in the early months, without getting dragged into the minutiae of writing all the various sentences.

Cleanup — at the beginning, and at the end
Finally, one of the tips I have for folks who are struggling with writing is to simply talk out their ideas aloud. Some of us are often more verbally wired, and we will connect disparate ideas and make interesting connections when asked a question, but flop when we’re just staring at a blank text. The voice feature in ChatGPT’s mobile app — and also products like Oasis AI — are interesting in that you can go stream of consciousness on a topic, and these products can clean them up significantly. The output isn’t that usable, IMHO, but you can then go and add/edit the result and develop it further.

It’s powerful when you combine this with AI-generated outlines. Just give ChatGPT a topic or better yet, an opinion (“open source products have bad consumer UX!”) and have it come up with an outline of the argument. Then start talking through the argument, and record it. Process it in one of the new AI apps, and then get it down to text. Then edit from there.

The other place where AI-driven cleanup can help is at the end, if you want to make sure you sound professional, or like an expert, and not an idiot. You can have ChatGPT rewrite content and incorporate a different tone. Or feed it some examples to cite and build off of. The ability to clean up ideas and textual content during and then after the creation flow is all pretty interesting.

There’s a bunch of things ChatGPT can’t yet do
I’m excited for all the ways that generative AI will help the writing process. I find edges to its capabilities that will soon be addressed, I’m sure. But here are some obvious ones that I notice all the time:

• I want it to have all the data, up to today (or in real-time), because I want it to use examples and cite numbers that are as recent as possible
• Some genAI tools can generate images, which is useful, since I want to create charts and figures, sketches of concepts, and even the little lead image that sits on top of each essay
• It’d be ideal if it was connected to Twitter and other social media platforms, so that you could tweet out content easily — writing tweetstorms via AI would be great!
• I want to train it on a corpus of my own writing, of course, and also writers that I respect. Right now it takes some work to customize the tone of voice to what I want
• There’s data, documents, and other info that I want to feed into ChatGPT, and then have it utilize the content as part of its arguments and suggestions. Imagine a routine task like, “hey, take this 10,000 word chapter and summarize it into a 100 entry tweetstorm” would be hugely helpful. Right now it can’t do that

There’s a lot more to go here, but it’s incredibly promising.

When I’m blogging these days, no wonder I find myself with WordPress open in one window, and then ChatGPT in the other. And I find myself spinning up new chats, copying and pasting back and forth. It’s the first time in 10+ years that I’ve found my workflow significantly change, and I’m excited to see what happens next.

 

 

 

There’s been a big wave of Creator Economy startups over the past few years, as the rise of social media platforms has empowered content creators to become a focal point for consumer engagement. This wave of startups promised creators that they could help them better monetize their audience on social media if they only promoted their products. We’ve all seen these — creators promote a startup’s new offering via a link in bio, or mentions in video or via links — and drive their followers to a landing page that enables some new interaction or functionality involving the creator. Initially, these started almost as “tip jars” but over the years, many, many creative products have been tried, spanning e-commerce to newsletters to Q&A, and more. These products all promised a win/win with creators so that when their fans spent money, the company would only take a % of earnings, usually something like 10% plus or minus.

There have been big successes, with some of these Creator Economy companies hitting billions of earnings paid to creators, while others have struggled. The successful creator startups are much more defensible than previously thought, and new entrants (often with splashy celebrity backing) have struggled to launch. Now that a few years have passed, what have we learned about the dynamics of this sector? Why have some Creator Economy startups worked and why have others lagged?

I have a few theories of the dynamics at play:

• The creator power law: A small, concentrated number of creators have all the audience, which makes Creator Economy startups potentially fragile and dependent
• Battle for the bio link: Creator economy companies acquire their audience from larger social media platforms that often just have one spot — the link in bio — to promote a single company. It’s a zero-sum game to overpower other companies
• The graduation problem: Startups often charge a take rate — % of bookings — and if the creator is acquiring own their customers and also doing the underlying work, they want to pressure you towards reducing costs. The biggest creators often “graduate” from a platform, building their own, and taking their revenue with them
• Algorithmic feast and famine: Creator traffic is driven by social feed algos, which lends itself to big spikes in traffic that appear and then go away — the opposite of the steady, durable growth that startups seek

These are all concepts that I’ve learned from meeting dozens of creator companies over the past few years. And as the next generation of Creator Economy startups emerges, these are some of the dynamics they’ll have to figure out how to navigate. Let’s jump in.

The creator power law
So you want to start a Creator Economy company? The biggest dynamic you have to master is the power law of audience and earnings within the creator class itself.

Here’s a graph that shows % that the top creator earns on a platform like Patron, versus the 2nd and 3rd and 4th creators, all the way down (credit: Power Laws in Culture). You can see there’s quite a dropoff:

Imagine if you graph this all the way out, to the many millions of creators on these platforms on the x-axis. You’d see that it eventually flattens just one tick above 0%. There’s a lot of reasons why this is the case, starting with the idea that these creator platforms build themselves on top of social media which themselves have well-documented power law distributions for followers and content engagement. In turn, social media platforms have power law curves because of algorithmic discovery, but a small number of social butterflies just know a lot more people than that.

Thus any creator economy product that builds on a social platform inherits these power law curves. OnlyFans creators offer free content on many social platforms that then drive traffic to their private landing pages. Below is a graph of creator earnings, which show a similar curve – via API scraping in the essay The Economics of OnlyFans – showing that while some creators earn up to $100,000/month, the median is closer to $180/month. A familiar curve emerges:

While power laws naturally emerge in social media platforms, that can’t be the only other explanation. The reason is that creative work — including TV, films, music, and more — generally follows a power law pattern. Here’s an example from TV, from the essay Power Laws in Culture (worth reading in its entirety):

A few hit shows get all the viewers. And if you look at video games, movies, fiction, directors, authors, and more:

There are a lot of things going on here that might explain the universality of this phenomenon, but one core issue is the uneven distribution of creative skills in the world. A top writer or film director is really just that much better than the 100th. You can look at research output, the distribution of so-called “10x engineers,” and patent filings, for a parallel universe of power laws as well.

So what does this mean for Creator Economy companies? Well, it means a few things:

• When Creator Economy companies first launch, the long-tail creators they initially attract are too small to be meaningful
• To hit scale, they need to attract the largest creators — the ones who are most likely to be distracted with many other projects and products
• And even once you have large creators on your platform, revenue is often quite concentrated into a small group — so that if they churn, the financial impact can be big and negative

These dynamics all mean that the initial phase of a startup’s launch can be perilous. The best companies can aggregate so many small creators that the numbers start to matter, or organically attract large/mid-sized players. If a startup finds itself manually DMing/acquiring/handholding many creators (read: high cost of acquisition and ongoing service), then that’s a sign that the product might not solve a big enough problem for things to happen on their own.

The battle for the bio link

Social media platforms like Instagram and TikTok have advertising business models, and as a result, they don’t want to give people *too much* organic traffic. Better they make you pay to sponsor posts, creators, and ads. One way they’ve done this has been to offer a single link for driving organic traffic — the infamous “link in bio” — that appears at the top of a profile.

This is insanely valuable real estate for Creator Economy startups. If you can convince a creator to place your startup into this link, then organic traffic will appear in your product. With some monetization mechanics in place, the startup takes its cut. And initially, it worked. Early in the Creator Economy cycle, startups were competing with non-monetizing links — either links to other social media profiles or personal websites. But as time went on, people began to fill their bio links with highly monetizing links to Patreon, Substack, Twitch, and otherwise — this is much fiercer competition.

It’s now a zero-sum battle to displace another startup’s link in bio. The only way to gain organic traffic from creator profiles is by monetizing better than other older, more proven competition. If you simply match what an incumbent might make you, then that’s not enough – it has to be significantly more. Or you have to find a different piece of real estate, whether that’s inside the creator content itself – whether that’s video, text, or otherwise. Either way, new entrants will find this a major barrier, and while they might be tempted to subsidize earnings with investor money initially, that may not be enough to reach a meaningful scale.

The graduation problem
The graduation problem is what happens when your best customers get big, and eventually “graduate” — taking themselves and their customers off of your platform. Why does this happen? Creators provide obvious value to startups — driving traffic, creating content, and monetizing their users — and that makes the Creator Economy model attractive. But work with creators long enough, and they often think to start to think it’s *too* attractive. They start to think, they’re doing all this work, what gives you the right to charge XX%? Why isn’t this a $99/month WordPress subscription, why do I have to pay a %? This is particularly problematic because of power law curves, where a small number of whales often dominate top-line revenue. If a whale starts to ask, couldn’t they replicate your product by hiring an agency and paying them to build a custom website, then there’s a huge temptation to drop take rates to accommodate them. They eventually are tempted to “graduate” from the platform, reaching sufficient scale to build their own platform.

Contrast this to marketplaces startups and the on-demand wave to which the Creator Economy is often compared. In that sector, a company like Airbnb or Uber aggregates both the supply and demand sides of the network independently. These 2-sided marketplaces work best when each side is highly fragmented, which is why the biggest outcomes have been consumer-to-consumer or consumer-to-SMB marketplaces, versus B2B. (More on this from an essay of mine from a few years back, What’s Next in Marketplaces). In their initial formation, Creator Economy startups look more like B2B networks or maybe even SaaS platforms — their customer bases (the creators) are highly concentrated, and the creators bring their consumers. No wonder the frustration.

To overcome the graduation problem, Creator Economy startups have to provide a significant amount more value than the utility of payments and other commoditized tech. They need to have a moat, not just for external companies but also for their own customers who are tempted to graduate over time. The best version of this is to create network effects on their own — by acquiring and cross-pollinating customers and bringing them to each creator, a 2-sided network forms, with all of its usual advantages. (I describe all these dynamics more in my recent book, The Cold Start Problem). The additional functionality that the startup creates should ideally be proprietary on its own. If an AI-enabled creator economy company develops a very good foundational model that allows creators to monetize 10x more than before, it’s unlikely the creator will ever leave.

Algorithmic feast and famine
Creator economy startups often find themselves highly dependent on the whims of social media platforms and on the hits-driven nature of viral content. If a video goes viral on TikTok, a big spike in user acquisition might ensue. But startups are always trying to grow steadily month by month, and unlike SEO or referral programs, or paid marketing, it’s hard to create a consistent march of 20% MoM growth. Compare this to marketplace startups, which add value by doing the work to aggregate each side of the market — often spending billions of dollars to build buyers and sellers. When I was at Uber, during the hypergrowth years, the annual performance marketing budget to acquire Uber riders was a billion, and the driver side was close to $2B, and that was diversified across SEO, brand marketing, paid, referral programs, partnerships, and otherwise. This added a ton of value since the two sides couldn’t connect otherwise.

Creator economy startups are different in that they use creators to find their customers, but in doing so, they are highly dependent on a single channel. A dependency on a single marketing channel is always dangerous, as we’ve seen in prior years where changes to SEO algorithms obliterated multiple generations of SEO-dependent content sites. A dependency on social media is even more fragile since the content is naturally more ephemeral and delicate. I think this is also one of the reasons why subscription (with upgrades) has become the dominant business model for successful Creator Economy companies — allowing creators to build a long-term, durable revenue stream from each follower is just much more stable than a transactional model. It’s just much easier to stack revenue over time this way.

Algorithmic feeds also play into a competitive factor. In recent years, we’ve also seen YouTube, Twitch, Twitter, and other underlying platforms try to go after directly paying creators themselves and playing a more vertically integrated role in the Creator Economy. As this happens, you could imagine a multitude of platform shenanigans where they try to hoard creator relationships at the expense of new startups.

The best solution here, of course, is to layer on additional marketing channels to drive predictability. Combine a spiky social media channel with steady retention, an inflow of traffic from referral, SEO, mobile installs, and otherwise, and the growth curve becomes much more durable. But in the early days of a Creator Economy startup, they’re often going all-in on social, and it’s only with success that they can choose to invest in the other channels.

The upside and the future
Creator Economy companies are going through their second and third generations of startups. The bar has gotten higher. Instead of providing functionality akin to fancy tip jars, startups are building full-blown products — supporting multiple platforms, new forms of interaction, and providing new functionality for creators to interact with their followers. These products will have network effects of their own, sometimes becoming destinations of their own. And instead of launching a product anchored by one celebrity and expecting it to succeed, instead, startups are building real technology — often involving AI — combined with a broad go-to-market strategy.

The upside of this sector is that mobile use, and thus social media platforms, continue to grow incredibly fast, taking time away from the hours that people used to spend watching TV:

 

A lot of this movement is of course driven by younger generations:

(btw, can you believe that most 18+ people still watch 4-5 hours of TV a day?)

The point is, social media continues to play a huge role, and creators are ultimately a new class of participants in the economy that continue to gain power in both cultural and economy terms. And the products and tooling they use to fulfill their goals will continue being attractive. This is especially true because in the end, creators don’t want to be dependent on one social platform themselves — if they are strong in video, they want to go to podcasting, and to have a huge Instagram. And startups can always seek to be friendlier to the creators than the mega-social platforms.

Thus, I argue that the future of the Creator Economy continues to be promising, but the approach has significantly evolved and the bar has been raised. Startups will need to provide new functionality, create new forms of monetization, and adopt new technologies that make them more defensible to competition and in-house efforts by creators to replace them. Personally, I’m much more interested in Creator Economy startups that are AI- or video-first, and act more like marketplaces in providing a highly managed solution to both sides. I’m more bullish about startups that know how to collect $1000 from a smaller niche of users — thus creating more value — rather than a tip jar model that collects $2 from everyone. In coming years we will see many more variations that will work, and given the underlying consumer trends, I’m bullish this will remain a source of highly valuable startups.

A simple question
The last few years have been crazy, and no wonder there’s a ton of folks thinking about making job changes right now. I know this since I’ve been getting the calls. Often the conversations open with a laundry list of different companies, roles, and compensation packages. Every opportunity is completely different and hard to compare. There’s got to be a better way to organize your thinking about these opportunities.

Here’s my favorite question to ask:

“What do you want to be your next next job? And why can’t you get it right now?”

And then, of course, you work backward from that. This is the “Next Next Job” framework for thinking about career moves, particularly in the highly chaotic situations that we find ourselves in today where there are many many opportunities across different industries and company stages. This reflects the very natural flow of the recruiting process, where recruiters and colleagues often make referrals across a wide swath of companies that are making. It’s always fun to talk through the various roles, but also it feels chaotic.

I know how it feels because, of course, I’ve faced this exact situation before.

The Next Next Job is an evaluation framework that I used myself many years ago, to make an important decision: As an early 30-something-year-old, at the tail end of a startup adventure that had gone awry, I had a big decision to make. A few months after putting my startup team/myself on the market, I was choosing between several very strong acquisition offers at pre-IPO startups. Each had its idiosyncratic benefits — some of the team cultures were a better fit for me than others and in others, I had a stronger connection to the founder. The packages were also very different. It was an emotional rollercoaster to meet dozens of companies over several months, and then need to choose amongst them.

It was tempting to pick based on a gut reaction, but I felt like there must be a better way. I sought a more analytical approach to augment the rollercoaster. I have tremendous gratitude to my close friend Bubba Murarka who coached me through all the conversations. Once the offers came in, he challenged me to stack rank the opportunities based on my “next next job” — almost a throwaway comment — but something that’s stuck.

How to answer a simple question
Let’s go back to the question – “what’s your next next job, and why can’t you get it today?” — it’s straightforward to ask, of course, but surprisingly hard to answer. Often we don’t know what we don’t know.

The first is that we often don’t know what our next next job might be — after all, if it’s unclear what the next job is, speculating about the next next job seems even more nebulous. Yet there’s an advantage here because you can make a few pretty big buckets of next next jobs. Or you can at least start to, based on what you know today.

It might look something like this:

• 50% – Become a startup investor
• 30% – Start another company
• 10% – Join a high-growth startup as a C-level exec
• 10% – Random stuff? (Switch into a new cutting-edge industry, become a blogger/writer, etc)

For someone who’s earlier in their career and the product management function, the goals might be more focused on becoming a first-time manager of PMs, becoming “employee 1” of a high-potential startup, or getting accepted into YCombinator, or something like that. Others might be thinking about transitioning from a non-tech role into a tech job or maybe going from a non-product role into becoming a PM/designer/eng.

Of course, sometimes it’s not obvious what other roles might be interesting or appealing — this in itself can be a useful thing to focus on when meeting with mentors and colleagues in the industry. But assuming you have some pretty big buckets to think about, the next step I’d encourage you to do is to pick the top 2-3 of these and do your research. Meet as many people as you can who have your next next job. What were their career paths? What did they need to accomplish before they could get the job? And you can ask them straight out — “what are the gaps in my skills that I need to fill, to get your type of role?” Keep asking questions and meeting people until the answers start to sound pretty similar, and the delta of new information decreases substantially.

Sometimes there’s a shortcut (and sometimes there’s not)
A funny thing sometimes emerges, particularly for people who rank “start a new company” as their next next job — it turns out they’re already qualified. Some of these jobs have high degrees of emotional baggage, because of Imposter Syndrome and not feeling ready. But the reality is, sometimes people over-prepare for a future job out of a deep sense of risk aversion. These are folks who are getting as credentialed and qualified as possible, rather than jumping in. These are the “wantrepreneurs” who are wasting their time getting multiple advanced degrees, working at all the top companies, and who are often very smart — but just can’t bring themselves to actually do something on their own. Usually, when this is one of my friends/colleagues, I try to talk them into taking the largest degree of risk possible :)

On the other hand, often the next next job isn’t attainable and it’s for good reasons. Maybe you’ve only worked at a series of failed startups, and you need a “shiny” role or two that helps add some credentialing. Or perhaps you’re in marketing and interested in becoming a PM but aren’t yet close enough to the engineers and the technical details. Perhaps you’ve never managed anyone, and want a role to demonstrate strong managerial ability before jumping into a team lead role. Identifying these gaps can help form the basis for evaluating potential job opportunities — which ones help fill them better and faster.

Gaps might encompass a number of things — skills, but also network, experiences, mentors, and ideas:

• What new skills do I need for my next next job?
• Is there a new network of people that would help me?
• Are there experiences that I need to demonstrate to land the next next job?
• Which mentors do I need, and how would I meet them?
• How do I get exposed to the ideas that might inspire me in the future?

Understanding these gaps are great, but that’s just playing offense. A “superpower” is often important, and there are superpowers that are so important that they overcome an imperfect set of gaps. I sometimes talk to folks who are interested to get into investing, and in the end, you can check off every skill on the list, but unless you have a specific superpower I care about — getting in the flow of new startups we’d be interested to meet and invest in — it doesn’t matter how good your analytical skills are, or that you attended fancy schools. If you have an incredible network of founders who seek you out, you can learn some of the other skills. For your industry and specialization, figure out the superpower that might trump everything else. Think offense (building up a superpower), not just defense (filling in gaps).

I want to give an example. For someone interested in investing as their next next job, and have substantial internal-facing roles at successful startups, I often find the list looks something like this:

• The next next job: Become a professional investor
• Gap: Need to develop a personal brand for other external-facing networks
• Gap: Haven’t done any angel investing
• Gap: Need to develop opinions on cutting-edge spaces
• Potential superpower: Get in the dealflow of recent spinouts/alumni of my prev companies

Again, this is just a hypothetical example – you can run your analysis and figure out what you need to do to close some gaps and develop a superpower. Of course, if you are tracking 2-3 options for next next jobs, you might find that a few gaps appear and re-appear. That’s great! This means the next role that helps you develop against that should be weighted more heavily.

Evaluating the grab bag of jobs
Now we can go back to the original question — “what do you think of XYZ as a company, and should I take this ABC role there?” The approach then becomes more obvious. First, you need more than one option. Go run a process, meet enough people, and look broadly at enough companies that you have different options to compare. Building up these options — while the opposite of instant gratification — will mean that you can truly make a good decision.

Then work backward on your options. Which ones are best at filling your gaps, and what will help you develop a superpower? Or if none of the options do much, then be patient. Develop more options. I think a lot of the reason why we often see a long series of 12-month stints on resumes is that we live in a world of instant gratification — whether that’s short videos for instant entertainment, on-demand food, groceries, cars, or online dating. But when it comes to making a big career decision that might commit you to work somewhere for (ideally) multiple years, the quality of the decision is important. Taking your time is key.

Of course, in the end, this big decision is very emotional. I really believe that. There are a ton of little things that go into getting excited about a new role: You’ll need to vibe with your manager, and you’ll want to like the work. But at the same time, having a bit of an analytical framework behind your decision will help. It might ultimately be 80% emotional and 20%, but I think that’s still better than 100% emotional rollercoaster.

To wrap up my story, almost ten years ago, I decided to go to Uber rather than the other very good options on the table. The background was that I knew it was likely I’d want to become an investor one day, and that pointed toward Uber as the right choice. The theory was that Uber would be a great place to meet future founders, would have problems at scale, and would be a very interesting place. And boy, was I right about that latter point :) It did OK on the other things I cared about — the scope of my role, the compensation package, etc — but I chose not to optimize for that. I knew it wasn’t my forever job. It was a stepping stone to the forever job I’d try to gain in startup investing, many years later.

The crisis arrives slowly, then all at once
At first, everything seems rosy. The growth rate of a new product is spiking, and growing quickly, maybe even hundreds of percentage points a year. But weirdly, a year or two in, there’s some softness in the latest numbers. Maybe it’s seasonality, or maybe something else. But worryingly, it keeps slowing. First to 300% a year, then 200%. Then 100% – a mere doubling annually in a startup ecosystem that demands a much faster target. More features are planned, and some are even shipped. Eventually, there’s a back-to-back where things are completely flat. What starts as a slow boil – where the team has a well-planned roadmap and a big vision – becomes a sudden crisis. There are late evening phone calls and emergency sessions. Analytics dashboards are pulled and re-pulled, to figure out what’s going on. The team needs a new plan.

There’s a saying that no military plan survives first contact with the enemy, and similarly — no product roadmap survives first contact with stalled growth. Instead, a crisis ensues, and the entire roadmap has to be rewritten. Particularly for startups, where continual growth is life and death.

When this crisis hits, the question is, what to do about it? How do you come up with a plan?

For better or worse, I’ve had this conversation with product managers and entrepreneurs many times over the years. The easy answer that people generally want to hear either falls into the camp of:

• This next magic feature will fix all our growth problems –The PM
• We need to spend more money on marketing –The Marketer
• Have you considered adding more AI? –The Investor

Don’t listen to these people :)

Instead, I offer the idea that you can analyze growth stalls systematically. You can ask questions, gather data, and assess the stall to zero in on the problems that are driving the metrics downwards.

Assessing the stall – starting with retention
First off, let me explain what’s happening during a growth stall. Yes, of course, it’s when a top-line number (like revenue or active users, or otherwise) stops growing. But what’s happening under the covers? At its core, a product stalls when its churn catches up with its customer acquisition.

I encourage y’all reading the entire thing, but I’ve written about this in the past in the deck The Red Flags and Magic Numbers That Investors Look For, which shows this growth of the underlying dynamics:

That is, a stall occurs when a product is churning enough users that it overpowers the counterforce – the product attracting new users and reactivating users (though this latter term is less important for startups). This happens because typically churn happens to a % of the user base, as anyone who’s seen cohort retention curves knows. But unfortunately, new customer growth channels tend to be fairly linear — most marketing channels don’t scale up as the user base scales up, and even the channels that do, like viral marketing, eventually saturate and slow down. All while churn continues to creep up over time as a percentage.

Because of these dynamics, I start by asking questions about retention to establish a baseline.

These questions are just a starting point, because once you ask them, the question is — what do you do with the answers?

• What is the D1/D7/D30 of the product? (if consumer?) How does it compare to other products in its category?
• If it’s a workplace product, how many days per week does the typical user engage? (This is the Power User Curve)
• Are people as active and engaged as you expect them to be? If it’s a daily-use product, does your DAU/MAU ratio reflect that?

There are many benchmarks out there for all the product categories, but as a very rough guideline, you need a D1/D7/D30 of 60/30/15% to be at respectable numbers for a social app. You need DAU/MAU over 20%, and if subscription based, you want churn <5% if SMB (and free acquisition). There are equivalent numbers for net revenue retention, session lengths, and lots of other metrics too.

A marketplace company might look at a different set of metrics. Often the demand side can have heavy churn, but the supply side should retain well (>50% YoY). An enterprise SaaS product would have its own set of metrics. It’s important to benchmark, to see if there are successful products with similar metrics that have gotten to scale. If you have similar numbers, then probably these underlying retention metrics are not the problem.

Let’s look there first, but understand that you might find a devastating truth.

Admit it when people don’t want your product
There’s an ugly truth that when most products are put under a microscope, most of them simply don’t have the retention to sustain growth over time — this is “pouring water into a leaky bucket.” A slow growth rate is inevitable because products start at a mega disadvantage of needing to replace all their existing users who churn, in addition to building new marketing channels that grow the overall number significantly.

But “my product is not retaining” is also sometimes a fancy phrase for “people don’t want to use my product.” I say this because it’s a blunt way of stating what’s often true – that a new product is too experimental, too unpolished, or so poorly positioned, or underdeveloped, that no one wants to use it. I think this was especially a problem in the Web 2.0 days when folks would combine their favorite random set of product mechanics — disappearing text messages sent to strangers near you, but you can only reply with a video — and launch them as the latest app (Disappr! – gotta love those 2010 app names). When people don’t want your product, no amount of new customer acquisition is going to solve that. Yes, you can sometimes generate very fast growth rates for a few weeks or months, but eventually, it catches up to you. And then the product stalls, per the graph above.

Instead, when initial product/market fit is low (yes, another fancy way to say people don’t get it), I usually recommend the exercise of positioning more closely to existing product categories. As I argue in Zero to Product/Market Fit, any founder can instantly get to product/market fit by simply going after an existing category — of course, we all know how to build and design a coffee cup such that there’s product/market fit. You incur other problems, of course, such as competitive differentiation, but if you combine a well-known product category with innovation, and picking at the right time and place in the innovation cycle, it can work.

There are major questions to ask here:

• Does my product have a clear, successful competitor? Is there a there there? (and do I have strong differentiation?)
• When I ask people to describe my product back to me — without the jargon — what do they say?
• When I ask people during user tests what kind of people might use the product, and what they’d use instead, do the answers make sense?
• Do people actually like my product, or are they just being nice to me? And a famous question- is it a painkiller or a vitamin?
• Are there any well-known product categories I could position against? Is there a way for me to test that positioning in user testing or otherwise?
• Is my growth the fault of shitty retention? Or do I need better user acquisition?

When retention sucks, but you haven’t growth hacked yet
What if retention sucks, but you haven’t added email notifications yet? What if you can just do a big marketing push, and that might spike the numbers? I can tell you as someone who has seen many underlying metrics for a wide variety of products, moving the retention number is the very hardest thing to move. Usually, the initial numbers are a ceiling, and it only goes down from there. So if your numbers are bad, don’t think that adding notification emails will solve it.

There is a very very narrow set of situations where I will take this back:

First, long-term retention is often most improved by better initial user activation. A few years ago, in Losing 80% of mobile users is normal, and why the best apps do better, I show that the biggest difference in the retention curves of the best apps and mildly good apps wasn’t as much in their long-term retention curves, as much as their ability to get the numbers in the first 7 days up higher than others. So I often will ask the question to product leaders- what differentiates someone who’s activated versus not, in your product? What % of users become activated? And how do you make that 100%?

Second, there’s a narrow class of products that have network effects — social apps, workplace collaboration tools, dating apps, marketplaces, etc — and they will often have a “smile curve” when retention goes up as time passes, and the network fills in. I wrote a whole book about this so I won’t belabor the point, but the main point is, if a product is more useful when more of your friends (or colleagues) are using it, then retention will naturally float up as the product grows. Thus, a product that has poor retention in the early days might just need more network density. For these situations, I might suggest the team do a completely manual, hands-on build of a network — launching at a high school or a single office — and measure retention there. Sometimes it’s much higher, which means there’s a there there, and the product just needs to be launched in a network-by-network manner as some of the great companies have done via college campuses, cities, workplaces, or otherwise.

Whatever you do, don’t fall for the idea that you can fix your retention by simply adding features:

The Next Feature Fallacy: the fallacy that the next feature you add will suddenly make people want to use the entire product. -@bokardo

There’s a longer explanation of the idea here, but the TLDR is that when you add features that engage hardcore users, that’s going to be such a small % when in reality you need to stem the bleed in D1/D2/…D7. That is, in the activation step of the product. If you get 10% of your hardcore users to engage more deeply, the reality is that it won’t move the needle enough mathematically to lift your entire retention curve. This means that you need to listen to the “silent majority” of users who churn, rather than the core users who stay and are highly vocal.

Thus, I’d ask myself the following questions:

• How is my retention? Am I counting on the ability to move metrics far beyond what’s reasonable? (You can increase 20%, but probably not 100%)
• Am I betting the farm on some product magic that hardcore users want? Or am I working on things that cause more newbies to love the product more quickly?
• Is my product in the category where network effects might substantially grow retention? Is that reasonable to think?

Top of funnel
It makes me happy when I see strong retention numbers with a flat growth curve. Funny enough, I consider this a very good thing. The history of fixing these situations is much better, and the approach is usually quite simple: Find more marketing channels, and scale existing ones. And if you can, find a self-repeating growth loop where users sign up for your product, use it, and then help generate more signups over time.

Just avoid the random lightning strikes. This could be from tech news coverage, a viral TikTok video, or a one-time email blast. You feel good for a moment, and when the excitement (and growth curve) dies down, then the crisis begins. It might be a fine way to solve a cold start problem or to get your first few hundred users. But it’s not a real growth strategy and leads to a product that’s lurching from crisis to crisis. Instead, the focus needs to be on repeatability, particularly once retention is established.

The easiest way to find a repeatable strategy is by simply fast-following other companies in your space. Finding and scaling marketing channels is typically pretty easy. If they are doing paid marketing, then go into those channels and test for CAC and measure payback periods. If they are marketing via Twitch creators or Instagram influencers, try that too. This method of simply experimenting and copying the competition goes a long way and often leads to success.

Testing marketing channels, alongside ad creatives and call-to-actions, requires an entrepreneurial spirit. There’s a huge advantage to testing a lot of different ideas, creatives, and landing pages and experimenting with messaging.

Growth loops scale and scale

Figuring out a growth loop is even more powerful. The idea here is that the loop helps attract users, who take actions that attract even more users, and so on. Thus a product with 10,000 users will grow quickly, but when it hits 1M actives, it can go even faster. This means user acquisition is a function of the size of the user base, and thus, it will keep up with the churn curve that’s stalking just behind it.

I have a few examples in my Magic Numbers deck, where I illustrate these as some of the classic and ideal growth loops:

Above: Viral loops are important because they are extremely scalable, free, and don’t require a formal partnership. This is based on users directly or indirectly sharing a product with their friends/colleagues, and having that loop repeat itself.

 

Above: A product like Yelp or Houzz fundamentally is a UGC SEO driven loop. New users find content through Google, a small % of them generate more content, which then gets indexed by Google, and then the loop repeats. Reddit is also like this. So is Glassdoor. And so on.

The process of figuring out these growth loops is not an easy task- it’s a form of product-led growth that requires an understanding of marketing, product, and sometimes growth hacking the underlying platforms/APIs to get a leg up (as Zynga did on Facebook, and Paypal did on eBay). But it’s very powerful when done well.

Polish your the UX flows that matter to growth — signup, inviting, payment — and ignore your hardcore user features
For teams that are focused on growth, it’s uncomfortable but necessary to ignore your best users and instead focus on UX targeted at users who may not be vocal at all. If you can polish your new user flow, then you can often make 20-50% gains to conversion, which then fall straight into the bottom line (whether that’s revenue or an active users count). When you polish your friend invite flows or referral flows, then you might get 20% of users to invite 100% more of their friends. And then that larger group of invitees will invite each more friends, and so on, with a larger viral factor. This is why when I assess product UX, I tend to focus on the less sexy stuff: Signup flows, invitations/referrals, and payment. And even surface areas like the lost password flow, which is for larger products, often block engaged users from getting back into their accounts.

Unfortunately, this is a product surface area that isn’t considered particularly sexy. If you’re at a large company, you may not get promoted to the next level of PM for delivering this type of project. In these settings, PMs are often rewarded more often for coordinating massive cross-functional projects than to move the needle on growth, by simply testing dozens of variations of signup flows.

And yet, this is often what matters!

There are a couple of key things I’ll often assess when looking at these growth-critical user flows:

• Are the value props clear, the headlines crisp, and generating urgency for the user?
• Are all critical elements — buttons, form fields, etc — above the fold?
• Are extraneous links removed, to not divert the user, or otherwise moved to below the fold?
• Instead of asking users to scroll, can content turn into a video, animated GIF, or slideshow?
• How does it look on desktop versus mobile?
• If the signup process is multi-step, can some steps be skipped for now, and done later?
• Is the order of the signup right? Can you bring forward the magic moment, rather than asking people to fill out form after form?
• Are there critical asks — getting a credit card, asking people to invite friends — that should be baked into the first few steps of the signup flow?
• Does the signup flow activate people correctly? Should the user be “forced” to activate in any way, by adding required signup steps?
• … and on and on

For new user flows, I try to get more users that hit the landing page to ultimately become activated users. I use tons of A/B testing and experiments in messaging to make this happen. For invite flows, I often try to stick them to the end of sessions so that users repeatedly see them as they engage the product. Maybe they create content, and you ask them if they want to share their newly created content with friends/coworkers. Do that every time, and you’ll be generating viral factor as you go, rather than just at the beginning. Payment is similarly important for products that focus on paid marketing to grow. The earlier you harvest purchase intent — often in the signup flow — the more you can plow that money into growth programs.

There are these flows and more, and they are the unsexy product features that drive growth.

Some final thoughts
Even great products stall growth. Famously, Facebook grew in its early years to take over colleges, but then saw a stall as saturation effects took over, and the product needed to be expanded past universities. Then there was another period of flatness, just before they expanded internationally. And another, before mobile. The same was true for Dropbox in its early years, as it saw a spike on Digg and Hacker News, but it needed a referral system and shared folders to push it to the next level. And in recent years, TikTok stalled as a platform for dance videos before it was acquired, and a very large paid marketing effort helped push it over the top based on building out a massive library of content.

These stories are common because successful products inevitably saturate a market, or need to jump from one acquisition channel to another, or any number of problems. When this crisis happens, it’s easy and reflexive to simply try to spend more on marketing. Or to try to develop more features. Or some other simplistic rule like that, sometimes based on the natural ability and interests of the product team.

Keep yourself from doing that.

Instead, consider that every stalled growth curve has its idiosyncratic issues. Sometimes it’s poor activation. Sometimes the novelty has worn off. Or perhaps the product is seasonal, or a marketing channel has been saturated. For better or worse, finding the levers to correct the stall requires patience, analytical abilities, and deep customer empathy. It’s hard, and every stalled product has its own story. But to identify the problem, fix it, and see the graph return to its previous glory — well, that’s just an amazing thing.

Hi readers,

I wanted to share a fundraising deck for a16z’s GAMES FUND ONE that we put together a little over a year ago. Of course a lot of fundraising decks have been published for our favorite companies — in fact, I have a folder full of the v1 pitch decks for Dropbox, Coinbase, Uber, etc — but rarely do you see decks for venture capital funds. I wanted to share this deck with some voiceover so you have a sense for what’s involved. Obviously there are lots of redacted slides, but you’ll at least get a flavor.

Thanks, Andrew. (Venice, CA)

 

A16Z GAMES FUND ONE PITCH DECK

We’ve hit the 1 year mark!!!!

A year ago, we raised GAMES FUND ONE, our first $660M fund focused on games. 25 investments and 10 new hires later, we’ve learned so much.

To celebrate, I’m going to share some slides we used to raise the fund. Yes, even VCs have to pitch :)

 

We structured our pitch deck into a discussion and overview about the games industry – many folks outside the industry needed some context to catch up. And then our investment areas, and then how we approached the team. (I’ve redacted notes about returns/companies/etc, of course)

 

For the overview – Games are driving the GenAI Revolution and this games are having their “Marvel Moment” showcasing their cultural power as some of the top shows – The Last of Us, Movies – Super Mario Movie, and Games are expanding entertainment IP like Hogwarts Legacy.

And by the way, the industry has changed substantially from when people were buying cartridges one at a time to decade-plus long-running games where friends come together as a next gen social network.

Games are often the killer app for new technologies. 3D, GPUs, and virtual items all came from games, though we might re-label them as metaverse, or NFTs. Before freemium, there was shareware. Games helped bring computing to the home, with Atari and Nintendo.

 

Here’s the “why now” slide – and today we’d update this slide with AI AI AI AI AI AI :) – but there are some amazing tailwinds that the industy has seen over the past decade

 

Last year we were very focused on studios, web3, and infrastructure. This year we’ve continued to maintain that focus, but have really started to lean into AI. In Q1 alone, we met over 100 AI x Games companies. And in 2023, 80% of the Games Fund’s investments have had a major AI component – either reinventing core gameplay or creating tools

We need to balance our approach to include all key areas of games, but also adapt to new trends.

As we invest across the ecosystem – both game studios but also infra, next gen consumer, and other areas – we need to also help the companies succeed. We do that by building a team that’s “games native”

A major part of a16z’s differentiation has been our Operating Team that helps companies. In fact the vast majority a16z’s staff focuses on that, and it’s a small minority that focuses on investing Gaming startups want help here: – hiring – creators – launching – publishers

 

Last year, we hired operators like Doug McCracken from Supercell to lead Marketing, Lester Chen from Youtube to lead Creators, and Jordan Mazer from Riot to lead Talent. Everyone has a startup mindset we all act like owners and work without clear directives. (Many more folks to come!)

Given the economic downturn, we’ve spent innumerable hours working with our portfolio to maximize their probability of long term success. This is when the operating team is so powerful. But the games market is resilient!

As we round out the year, we’re continuing to build and innovate. Builders are needed to help builders. We’ve been building the team in year one and hired from Blizzard, Supercell, Youtube, Riot, Twitch, and Unity. Join us!

Thanks to everyone who’s supported us in Year One!

Kudos to the new team, our founders/co-investors, and our friends across the ecosystem

And of course, my close colleagues Jon Lai, Marc, Ben, and the entire team who have been building this initiative from the start!

PS. Finally, our lawyers made me add this at the end :)