Discover more from How They Grow
🌱 5-Bit Fridays: Product-led sales, advanced Opportunity-Solution Trees, how to build a growth model, automatic creativity, and writing great PRDs (with templates)
👋 Welcome to this week’s edition of 5-Bit Fridays. Your weekly roundup of 5 snackable—and actionable—insights from the best-in-tech, bringing you concrete advice on how to build and grow a product.
📬 To learn something new each week, drop your email here and join 9K+ others:
Happy Friday, friends 🍻
We’ll start with Pappa Elon. Musk has moved through the electric car industry, social media, payments, and space.
Now it looks like he's tapping into another unexpected market —cat products.
A bizarre listing on Tesla's China website shows a "Cybertruck Multifunctional Corrugated Cat Litter"—essentially a cat litter box made out of cardboard. It needs to be self-assembled. And as a cat owner, it looks…like a nightmare
Imagine being an engineer at Tesla and having Elon drop this idea on your desk. I think he misinterpreted some Cybertruck feedback on Twitter. They said it looks like shit, not that it needs to hold shit.
Not much else to say here.🤦 Let’s just move on.
Here’s what we’ve got this week:
PLG → Product-led sales
The advanced way to think about Opportunity-Solution Trees, and rapid assumption testing > A/B testing
How to build a growth model
Great PRDs, with templates from world-class companies
Small ask: 👉 If you enjoy reading this post, feel free to share it with friends! Or feel free to click the ❤️ button on this post so more people can discover it on Substack 🙏
(#1) PLG → Product-led sales
As you no doubt know, there’s been a paradigm shift in the software world in recent years. SaaS businesses have switched from the sales-led GTM to a motion of product-led growth.
In large part, that’s because PLG helps products to scale a lot quicker. This happens for two reasons:
PLG creates a wider top-of-funnel. Allowing people to try before they buy brings prospects into your product much earlier than if they waited for sales outreach. This is powerful because it gets the evaluation stage going way sooner.
PLG enables rapid global scale. Instead of scaling your sales team headcount (timely, costly), you improve onboarding (which drives better activation, engagement, and retention) to serve your much larger inflow of users…all in a quicker and more iterative way.
Speed aside, and more interestingly for longevity, PLG creates a nice moat. Leah Tharin—the PLG expert—tells us why:
Faster sales cycles: By having your prospects onboard themselves, you can significantly reduce your prospect’s time-to-value and sales cycle. Once people experience the value in your product, the next logical thing to do is upgrade. The quicker your users can accomplish a key outcome in your product, the quicker you can convert your free users into paying customers.
High revenue-per-employee (RPE): Software was always built to scale well, but with a product-led approach, you’re able to do more with fewer people on your team. Less hand-holding means higher profit margins per customer. Just take a look at Ahrefs in 2019. They have a $40 million ARR business with 40 employees.
Better user experience: Since your product is built for people to onboard themselves, people can experience meaningful value in your product without any hand-holding
And product-led, does indeed, mean product-led: PLG shops lead with the product across every department, and all teams leverage the product to hit their goals.
A product-led marketing team asks, “How can we use our product as the #1 lead magnet?”
A product-led sales team asks, “How can we use the product to qualify our prospects for us? That way, we have conversations with people that already understand our value.”
The product-led customer success team asks, “How can we create a product that helps customers become successful without our help?”
The product-led engineering team asks, “How can we create a product with a quick time-to-value?”
With PLG nicely established, and more and more founders making this the core driver of growth, product-led sales is the logical continuation of product-led growth.
In a great deep dive/guide into PLG, Leah Tharin unpacks this in more detail. As usual, we’ll just scan the bottom line.
Simply, product-led sales introduces “touch” into the process. Instead of the product doing everything, self-serve owns the entire lower half of the market, and for more upmarket prospects, it helps to qualify the right people for a sales team to then step in for the final mile.
This creates product-qualified leads.
The goal, as Leah points out, is to “touch” the customer as late as possible. Where now, for the high-ticket customers, sales can close PQL at a much higher rate than with marketing-qualified ones (MQL) because (1) they have real insights into how and why the company is using their product, and (2) the prospect has a familiarly with the product already.
But, how does PLS affect an individual's buying decision?
Contrary to what you may believe, product-led sales hands you more control over an individual buyer's decision than sales-led, not less.
You might have fewer leads (initially).
Buying decisions may take longer the more transparent you are.
However, the transparency entailed by no-touch shows low-quality leads out all by themselves. “Convincing” those leads by optimizing sales pitches alone is just an illusion.
PLS, done right, empowers your sales team to leverage user data. There’s a shift from basic questions to targeted conversations. This leads to better product development as product usage data is a table stake for great experimentation.
Better product-market fit.
High-level takeaway: You can have your cake and eat it too. 🍰 Mix both product and sales-led together, especially when you start looking for customers in the mid to upper market.
And while we’re talking about the power of product-led growth, and how buyers expect a “Show, don't tell” experience, here’s a question for you…have you ever tried building a product demo for your product before? If yes, you know that doing it natively can be a pain in the arse to maintain, and time-consuming to build. If you haven’t and you operate a SaaS, they are a great onboarding tool to think about. That’s where Storylane comes in, as they help you build killer product demos in <10 minutes.
With Storylane, you can leverage the power of interactive and personalized demos to shorten your sales cycles, increase prospect interactions, and convert more visitors. It also gives you actionable data on demo engagement, helping you figure out which demos are driving qualified leads.
To join other GTM teams at companies like Gong, Clari, and Twilio, hit this link to check out Storylane.
…yes, that was an ad. Just FYI, I’m testing sponsorships out only for 5-bits. And I will always be very selective about the products I suggest, only doing sponsorships with products I believe can either help you be more productive personally, and/or help your product grow.
Join 9K+ people learning about product, growth, and company building. Free for now, not forever.
(#2) The advanced way to think about Opportunity-Solution Trees, and rapid assumption testing > A/B testing
Earlier this week,shared his advanced techniques around Teresa Torres’ Continuous Discovery Habits. An epic post with tons of takeaways, as expected. But one thing that really stood out was Aakash’s take on Opportunity Solution Trees (OSTs).
Wondering what that is? 🌲
Let’s start with how Teresa described it—the classic way.
The original OST
The Opportunity Solution Tree has four parts: a clear outcome, opportunities uncovered from research (i.e. unmet customer needs), solutions to target opportunities, and then experiments to evaluate those solutions and the riskiest assumptions behind them.
Visualized, it looks like this:
A really great tool for seeing how what you plan on building aligns with your business goals.
But it gets better. Here’s Aakash’s spin on it. 👇
The advanced way:
In short, Aakash combines the OST with the OKR framework and adds an important element to the mix: the problem
I’ll just let him explain:
Traditional OSTs can leave teams stuck in a nebulous world of opportunities, while OKRS (Objective - Key Results) alone may lack a clear path to implementation. By incorporating problems and solutions into the equation, you align every opportunity with a tangible path forward.
If you’ve read my work before, you know that I’m a fan of OKPS (Objective - Key Result - Problem - Solution) trees. These are a slight variation on Teresa’s OSTs.
I find that they solve three major problems:
We’re all working on OKRs anyways - OKPS clearly ties elements of the OST to the OKR framework.
It skips the confusing Opportunity v Solution distinction - theoretically, they seem clearly different. But in practice, it’s not so easy to tell the difference between the two. So I’ve taken to skipping it.
They make it really easy to organize and present your roadmap - the challenge with moving towards outcomes over outputs is meeting waterfall requirements from execs. OKPS give you a healthy middle ground.
OKPS are a small change on OST with big quality of life impacts.
This technique is essential for product managers, team leads, and stakeholders who need to see both the forest and the trees - understanding the big picture while making concrete progress.
The issue Aakash points out that resonates the most with me there is #2: opportunities and solutions often get intertwined. But, it's hard to talk about a problem as a solution. 💡
Now let’s break down how to make your tree. In short—start from the top and make your way down:
Start with a few clear objectives that align with your product strategy and vision.
Define the key results that will tell you if you reach those objectives.
Now that you know where you want to go, speak to customers and do research to figure out what the key problems are to getting there.
Again, rifting off your customer discovery, brainstorm ways to overcome those obstacles. Think big, think small, and be sure to balance the short and long term. This will help you prioritize solutions with things like user impact, feasibility, and objective alignment.
Make a nice pretty tree so you, your team, and your leaders can see how it all fits together.
Test your assumptions. And importantly, that does not mean just A/B tests. While definitely useful, be careful not to over-index on A/B testing everything. It slows you down. Rather, run rapid assumption tests too
Like a living tree, your OKPS is not a one-and-done artifact. Always review and iterate based on your tests and new inputs. Aakash advises on a bi-weekly basis.
Go deeper: 🧠
Aakash’s full deep dive (paid): Advanced Techniques: Continuous Discovery
Teresa Torres’ book (paid): Continuous Discovery Habits: Discover Products that Create Customer Value and Business Value
(#3) How to build a growth model
A growth model shows you…the model of how a product grows. 👀
But it is that simple (in theory).
Think of it like an equation that tells you how changes in different variables in your business lead to different growth outputs.
Of course, there is a spectrum for how deep you go if you make one. But that really is the crux of it: a way to conceptualize and summarize your business in a simple equation, which allows you to think about growth in a holistic and structured way.
If acquisition goes up, but engagement and retention go down, we predict this growth change will happen.
A handy mathematical tool. 🔮
So, how do you make one?
As with any mathematical model, like Volume = l × w × h, yours will need:
Inputs to put into your function
And a dynamic output based on how the above changes
In Andy Johns’ interview with First Round Capital, he shares a basic growth equation that he learned from his former boss, Chamath Palihapitiya, who ran growth at Facebook—one he suggests everybody internalizes (and later builds upon).
The first step though, as Hila Qu (Head of Growth at Gitlab) points out in this post, is to focus on the fundamentals.
Leave the numbers and equations aside and start with a high level conceptual model that captures the core levers of growth. This alone will provide a valuable framework to help you make good decisions. Bottom line is this model should explain how your business will grow, in a way that is different from that of other companies. Andy’s framework below is a pretty good one, because it covers the key steps in delivering value to your customers: acquisition, activation, and long term engagement.
After the first step, if you want to go deeper and you have the analytical power to get all the data, you can begin to construct a more data-rich model similar to Hubspot’s, with all variables linked by mathematical functions to produce the output: revenue or user growth, which can then allow you to perform all sorts of forecasting and sensitivity analyses.
This is how I read that advice:
Identify your qualitative growth model. AKA, do you understand and can you verbally articulate the drives behind how your business grows? Tip: think holistically about each cohort of your users and how they can acquire the next cohort of users.
If yes, then go create a more mathematically savvy (quantitative) model in a spreadsheet with your assumptions. Start with an MVP though, because you want to test some stuff. 👇
You probably don’t have the data chops to do this step yourself, so partner with an analyst on your team to find your key metrics by conducting a sensitivity analysis on your spreadsheet.
Expand the MVP of your model
And like with most things in the world of using data to drive decisions—it’s about directional accuracy, not perfect precision. As British statistician George Box says, "All models are wrong, but some are useful”.
If that sounds like work, it’s because it is. So you may be wondering, do you even need one? 🤷
Well, it takes time. And depending on where your product is at, this could well not be the best use of yours. If you’re pre-product-market fit or have a small user base (and small data set), rather invest your time building and iterating on your product. A much better ROI.
Yes, a growth model can help you do that, but it can be a distraction in the early days when you still have big leaps in metrics to make. If you understand the dynamics of your business well enough and focus on the right initiatives (using the OKPS framework ✌️), you should be able to drive growth well enough.
Okay, so why even talk about a growth model then?
There are unique benefits of going over the exercise of creating a growth model and making it part of your day-to-day decision making, which is also why so many successful growth leaders and teams advocate for it:
The process of uncovering all variables and then synthesizing them into one single equation will deepen your understanding of the business
Growth model also helps you assemble individual metrics into a big picture, and force you to think about their relationships and the relative importance
When entire team look at a single equation with a couple key variables, it helps prioritize and focus
Once you have growth model version 1.0, you will naturally begin to find questions to ask. As you ask more questions, you change how you
— Hila Qu
Just think of it as an advanced tool for driving growth, one that should be employed as long as it does not distract from finding and solving user problems.
Go deeper: 🧠
(#4) Automating creativity
The core irony of generative AIs is that AIs were supposed to be all logic and no imagination. Instead we get AIs that make up information, engage in (seemingly) emotional discussions, and which are intensely creative. And that last fact is one that makes many people deeply uncomfortable.
— Ethan Mollick
Creativity is a hard thing to define, let alone measure. Nonetheless, researchers have tried various flawed tests to see how good we are at coming up with unique ideas.
This was not a problem, as Ethan Mollick points out, until AI came along as passed them all.
Now we live in a world where GPT-4 beats 91% of humans on a variation of the Alternative Uses Test for creativity and exceeds 99% of people on the Torrance Tests of Creative Thinking.
Well, to get up to speed, researchers have set out to create some new experiments (did we rank their creativity?) to see how we compare to AI at idea generation in settings with real-world implications. Shoutout to Ethan for finding and reporting on these papers.
The first recent example of this came from a paper out of Wharton (where Ethan is a professor).
They staged an idea generation contest: pitting ChatGPT-4 against the students in a popular innovation class that has historically led to many startups. The researchers — Karan Girotra, Lennart Meincke, Christian Terwiesch, and Karl Ulrich — used human judges to assess idea quality, and found that ChatGPT-4 generated more, cheaper and better ideas than the students. Even more impressive, from a business perspective, was that the purchase intent from outside judges was higher for the AI-generated ideas as well! Of the 40 best ideas rated by the judges, 35 came from ChatGPT.
Then, AI also beat us in an experiment to come up with business ideas based on reusing, recycling, or sharing products as part of the circular economy.
A third paper focused on creative writing ideas, rather than business ideas. The study by Anil R. Doshi and Oliver P. Hauser compared humans working alone to write short stories to humans who used AI to suggest 3-5 possible topics. Once again, the AI proved helpful: humans with AI help created stories that were judged as significantly more novel and more interesting than those written by humans alone.
What do these three papers tell us?
AI can generate creative ideas in real-life, practical situations. It can also help people generate better ideas.
The ideas AI generates are better than what most people can come up with alone.
In other words, you’ll be more creative when you mix AI into your creative process. 👇
Putting it to use 🔨
The notion that you need to be super specific in your prompts to get value isn’t always true. The above papers confirmed that at least for idea generation, you don’t need to get fancy with your words.
For instance, in the research paper around business ideas, the style of the prompt was simple:
You are a creative entrepreneur looking to generate new product ideas. The product will target college students in the United States. It should be a physical good, not a service or software. I'd like a product that could be sold at a retail price of less than USD 50. The ideas are just ideas. The product need not yet exist, nor may it necessarily be clearly feasible. Number all ideas and give them a name. The name and idea are separated by a colon.
However, if you’re looking for the best ideas possible, the papers found that a helpful addition to your prompt is applying few-shot learning. This is where you provide the AI with examples of the kind of results you’d like to see (the “few shots” rather than “zero-shot” learning, where you provide no examples). For example:
Zero-shot: Come up with 10 startup ideas that would be a good fit for Y Combinator
Few-shot: Here are 400 of the latest startup ideas from Y Combinator. Come up with trends, then generate 15 original ideas combining these concepts
And Ethan describes an awesome addition to this.
You can also use other techniques that take advantage of the ways that AI can hallucinate plausible, but interesting, material, and use that as a seed of creativity. Consider asking it for interview transcripts for fake interviews: Create an interview transcript between a product designer and a dentist about the problems the dentist has, for example. Or ask it to describe non-existent products: walk me through the interface for a fictional new water pump that has exciting new features. There is an art to this that you can learn from experimentation.
That’s a very nifty idea. With AI connected to the internet, you can see how a prompt like that (run an interview for me) might actually uncover real problems people have been speaking about online. A nice addition to boost your Continuous Discovery Habits, perhaps. 🤷♂️
🚨 This post is about to get cut short if you’re reading this in your email. Continue reading in your browser. 🚨
(#5) Great PRDs, with templates from world-class companies
The Spec, PRD, or Project Brief is a big part of what PMs do. It’s one of the few deliverables we’re personally responsible for.
So, what does a good one look like?
A “good enough” PRD usually contains the following:
Problem & Opportunity Statement
Goals & Success Metrics
Roadmap & Go-to-market plan
We don’t like good though, do we? We like great, and great PRDs add more thought and depth. As Shyvee Shi says, great PRDs touch on:
Where should we play in this problem space?
Why are we uniquely positioned to “win” this?
Why now? What drives the urgency?
What user insights give us the conviction?
What are some product principles we follow?
What are the key trade-offs & decisions?
What are the risks & their mitigation plan?
What are the riskiest assumptions? What must be true for this idea to work?
And two notes about the PRD worth calling out:
Writing is an excellent opportunity to clarify your own thinking. Use the PRD to do the same.
Use the doc as the focal point for feedback. Writing the PRD isn’t about handing your perfect plan over to the team, it’s about making the plan better.
All that being said, here are a few templates to keep on the decks. ✍️
Kevin Yien’s PRD template (PM at Square)
🌱 And now, byte on this if you have time 🧠
Want to have your mind blown? I just read this on the Morning Brew, it’s wild.
Earlier this week, celebrity astrophysicist Neil deGrasse Tyson made this TikTok video. It’s well worth tapping play and watching (<1m).
In short: He revealed that the possible number of ways you can shuffle a 52-card deck is so unfathomably large (52 factorial, or 52!) that you could gather 1 trillion people, hand each of them a deck of cards, tell them to shuffle it 1 trillion times per second for 1 trillion years, let that happen across 1 trillion civilizations in the universe, let that happen across 1 trillion universes…and a deck of cards you shuffled right now would only have a 40% chance of matching the sequence of any single one of the decks in this experiment.
And that’s a wrap, folks.
If you learned something new, or just enjoyed the read, the best way to support this newsletter is to give this post a like or share. It helps other folks on Substack discover my writing.
Or, if you’re a writer on Substack, enjoy my work, and think your own audience would find value in my various series (How They Grow, Why They Died, 5-Bit Fridays, The HTG Show), I’d love it if you would consider adding HTG to your recommendations.
Until next time.