I decided to dedicate this week’s essay to answering this question, so let’s get started. As always with these kinds of questions, it is important to take a couple of steps back and reason about it from a more holistic perspective.

1. UI is a form of communication

A UI is just a language between a human and a machine. It is how you tell a computer what you want and how it shows you what it did. The job is simple. Minimize ambiguity. Minimize effort. Make outcomes obvious.

This matters because different “modes of communication” (reading, speaking, pointing, seeing) have very different bandwidths and error profiles. If you choose the wrong channel for the job, you may be compromising efficiency or speed.

2. Human bandwidth: reading, writing, speaking… and seeing

Here’s the quick reality check on our I/O limits:

• Reading (silent): Average adult rates cluster around ~238 words/min for nonfiction (faster for fiction). (source)

• Listening/Speaking: Comfortable speech/listening is around ~140–160 wpm in everyday use. Presentations and audiobooks target that range. Conversational can go higher but comprehension can drop. (source)

• Typing: Keyboard input for typical users is often ~20–40 wpm (power users go much higher). On phones, a massive 37k-person study found ~36 wpm average. The conclusion: voice can be ~3× faster than typing when conditions are ideal. (source)

• Seeing (vision): This is where humans are exceptional. The retina pushes ~10 million bits/second to the brain, and we can extract the gist of an image in ~13 ms. Vision is parallel and insanely high-throughput. (source)

• Language information limit: Across languages, human speech carries ~39 bits/second of information on average. That’s a universal bottleneck for verbal channels. (source)

Translation to UX: If you want to show lots of options, comparisons, and state at once, a screen is far superior to voice. If you want to express a short intent or capture a thought hands-free, voice can beat a keyboard.

3. Two levers to make communication efficient

You can improve the “dialog” in two ways:

A) Increase channel bandwidth.
Replace typing with voice, complement voice with visuals, add structure to chat (chips, forms), or use richer layouts for scanning, etc.

B) Reduce the information that needs to be said at all.
Let the system do more by default:

• Built-in business logic & automation: Move the repetitive and predictable work behind the scenes and surface only decisions that truly need the human.

• Great defaults: Users rarely change them, so pick them carefully to remove decisions and clicks.

• One-click actions: One click can carry identity, context, preferences, and history, so the system can execute the entire workflow with no extra fields, steps, or questions.

4. Point-and-click isn’t just due to lack of alternatives

I find it strange when people say that conversational interfaces (voice/chat) will make traditional UIs obsolete, as if the only reason we needed point-and-click interfaces was the absence of a good voice/chat alternative. This is simply not true. There are many reasons why we have needed, and will continue to need point-and-click interfaces:

• Faster for single actions: Clicking Mute is quicker than saying a sentence and waiting for recognition.

• Precision: point-and-click interfaces allow a level of precision that is much harder to achieve with voice or natural language.

• Language is non-deterministic: Natural language is ambiguous. Buttons, menus, and forms constrain inputs and make outcomes predictable. This is why we invented programming languages and protocols to to remove ambiguity from language and achieve universal agreement and reproducibility.

5. Humans are visual creatures

We choose clothes, furniture, and food by looking at them. We don’t book a hotel room without seeing photos. We browse destinations on Instagram and travel sites before we commit. Even with a perfect voice interface, people will ask to see options before they buy or book.

6. What this means for UIs in the world of AI

Human to Human

Collaboration is visual. When people plan a trip, review a funnel, or choose a design, they point at maps, charts, and screens. These types of interfaces are here to stay, and while AI will be able to produce many of these UIs on the fly, there is still an argument to be made for the importance of familiarity and pattern recognition in building great user experiences. I won’t even get into the dimension of human taste, because the topic of good taste in the world of AI probably needs an essay of its own.

Human to Agent

State intent in plain language. I believe that a combination of “rich” conversational interfaces and really easy-to-use UIs are going to win. First and foremost, your focus should be on “making the machine do more”. This means defaults, automation, one-click actions, etc. Second, find the parts of the experience where switching to voice or chat increases the information bandwidth and reduces ambiguity instead of increasing it.

Agent to Agent

A few years ago at Google IO, Google Assistant made a phone call to book an appointment. One person on Twitter, who I can’t find anymore, said soon enough we’ll have an AI assistant on the other end answering the phone and taking the appointment. But if that became true, and we ended up with two computers on the two ends trying to communicate with each other, is human language the most efficient way for these computers to communicate? At a speed of ~39 bits/second? Of course not. Agent to Agent communication will look more like API calls or digital signal transmissions than natural language, and I believe something like the MCP Protocol was the first step along that path.

Closing thought

Traditional UIs are not going away. They will sit alongside language interfaces and automation. Use language to capture intent quickly. Use screens to compare, verify, and fine tune. Let agents do the heavy lifting behind a clear, visual surface.

Build for humans first. Then let AI remove steps the human should not have to take.

I’d love to hear from you, so if you have any thoughts you would like to share, please do so in the comments, or find me on Twitter.

In this essay I will try to answer one fundamental question:

How do we build the most effective autonomous agents in a world built for humans?

“Agents” here is used loosely to mean both software agents and hardware agents (robots). Let me start with the immediate thing that comes to people’s minds:

1. Robots

When people think about autonomous agents in a human world, they immediately think of humanoid robots. I’m a software products guy, so I promise I’ll eventually get to software agents in this essay, but trust me: thinking about robots is very helpful for reasoning about software agents.

Many have argued that humanoid robots are the superior form of robotics for the future, because we live in a world built for humans and a humanoid robot would allow the robot to be as “general purpose” as possible, thus justifying the cost. Instead of a thousand specialized robots, you just have one humanoid that can do a thousand tasks.

So are humanoid robots really the most effective form? Well, for most things, they are not. Do we really need general-purpose robots? Also probably not, at least not in the traditional sense. Let me explain both points.

1.1 The human form is inefficient

Look at the photo above. In agriculture, automated machinery looks nothing like human farmers and packs 1,000x the power and productivity. These machines not only replaced thousands of farmers, but they have also enabled innovation and possibilities that were simply unimaginable if we relied solely on human beings, because they broke free from the limitations of the human form.

We don’t see this only in agriculture. You see this across the board in manufacturing, supply chain, etc. So with this in mind, you have to imagine how dumb I think this promo video from Figure Robotics looks:

I mean, compare the idea of a humanoid package-sorting robot to this 3-year-old Amazon sorting system sorting packages in ways a human couldn’t.

So the conclusion I want to draw at this point is that specialized robots are more effective than general-purpose robots when the following three conditions are met:

1. We gain efficiency by going beyond the limitations of the human form.

2. The efficiency gains outweigh the cost of specialized robots.

3. The environment can be reshaped in a way that is less optimum for humans but more optimum for the specialized robot.

While #1 probably holds true for most tasks, one cannot say the same about points 2 and 3. Very good examples of this are tasks around the house like cooking, cleaning, doing the laundry, etc. So let’s dig into these to understand why they are different and what that means for robots.

1.2. General-purpose home robots are still specialized robots

And their speciality is operating within your personal environment. Let’s quickly take a look at how conditions 2 and 3 from the previous section apply here:

“The efficiency gains outweigh the cost of specialized robots.”

The efficiency of any home robot is capped at the needs of that particular home. A family of five would only have as much laundry as a family of five would. Having a highly efficient robot will not meaningfully affect the consumption habits of any household. The difference in convenience between a robot that finishes the task at human or near-human speed versus one that does it at superhuman speed is non-existent.

“The environment can be reshaped in a way that is less optimum for humans but more optimum for the specialized robot.”

The limited square footage in your home is, and should always be, optimized for the comfort of the human beings living in it. It doesn’t make much sense to redesign your home in a way that is less convenient for you but better for the robots, with no significant gain. Even if the gains were there, it is highly unlikely that they would justify such drastic measures.

It’s likely this applies to other areas of life, and perhaps even in some domains of service labor, but the main arguments and conditions probably hold across most of them.

2. Software Agents

2.1 Specialized & general-purpose software agents

Software you will find is a very peculiar thing and software tasks are wildly different from real-life physical tasks. Let’s look at the 3 conditions for specialized agents:

1. We gain efficiency by going beyond the limitations of the human form: This is inherent in software, because with theoretically unlimited compute, you can run a theoretically infinite number of tasks. In addition to this, computers are many orders of magnitude faster than human beings at most things a computer can do.

2. The efficiency gains outweigh the cost of specialized robots: Software is really cheap to run, that’s why SaaS was such an incredible business model with ridiculously high gross-margins. Even the GPU-heavy cost of intelligence, or in more practical terms the cost of AI inference, while expensive compared to traditional CPU computation, it is still cheap and getting cheaper every day. I am aware that while the cost of each model drops by 10X every year, state-of-the-art (SOTA) models remain relatively expensive, but even the costs of SOTA models is going down. Given the low cost of software, the efficiency gains are there even at a small or even personal scale.

3. The environment can be reshaped in a way that is less optimum for humans but more optimum for the specialized robot:

   Software exists in infinite virtual space. The environment where software runs is malleable, adaptive, and can not only be reshaped in every way we want, but it can evolve and change in real time.

But with LLMs, we’re not just talking about automation. We are talking about intelligence on tap. And intelligence I believe is one of the areas where Jevon’s paradox applies: unlike the case made earlier about home tasks, when intelligence is the product, people will consume more of it when it gets cheaper.

2.2 General purpose LLMs, and specialized Agents

To understand better the ability of any piece of “AI” software (i.e. software build on top of LLMs), to execute or reason about a variety of general purpose tasks, I need to explain a few concepts.

The first one is the general level of intelligence of a model, which is directly correlated with the pre-training and post-training of that model. As models get bigger and smarter, they become smarter and better across more and more domains. For example, at the time of GPT-3 a lot of companies found that they need to fine-tune the model on domain-specific data-sets to get the level of intelligence they need, but the vast majority of these fine-tuned models became obsolete once GPT-4 came around. I think we can expect this trend to continue.

The second concept is function or tool calling. Most models now have the ability to use tools provided to them, either directly or via a standardized protocol like MCP. This pushes the specialized to the level of the tool vs. the level of the model. You can build tools to get the LLM to do pretty much anything you want it to, or retrieve any information the LLM might need from virtually any source. A good example of tool calling is web search.

The third concept is Reinforcement Learning (RL). RL is a machine learning technique where an "agent" learns optimal behaviors by interacting with an environment through trial and error. This type of technique is compute heavy, and requires a very large number of environments for the agent to train on for each different type of task or system. For example an environment to learn how to play chess will be very different from the environment needed to lean how to play go. This technique while expensive, has proven to be extremely more effective than trying to teach the agents the “rules”. If you’re interested in learning more read Rich Sutton’s “Bitter lesson”.

Most of the AI products available today, especially agentic AI products, usually try to leverage the base intelligence and capabilities of the model, as well as tool calling. And this is in my opinion the correct path that most startups should follow. Most startups actually can’t and shouldn’t pursue paths where they need to buy massive GPU clusters to try to train their own specialized unique models.

But in a world where everyone has access to the same models, and has equal ability to build tools, where does one find differentiation? How do we build the most effective autonomous agents in a world built for humans?

3. User Experience in a world of autonomous agents

I spent the last 6 years of my life helping customers improve the user experience of their websites and apps, and what was true when I started is still true today: the best ways to differentiate your product is User Experience and Ease of Use.

There are two lessons I want to share with you from the best product person that has ever lived: Steve Jobs.

Steve understood that Apple’s edge comes from making technology easy and accessible. His relentless focus on simplicity, and his understand that innovation a lot of times means subtraction instead of addition, are the main drivers behind Apple’s huge success.

The best way to launch a successful product in the world of AI, is the same best way it has always been: relentless focus on the user experience.

The second lesson I want you to take from Steve Jobs is that it is not enough to have a great product, you need to have a sound business too. Margins matter, costs matter, growth matters. The gravy train of subsidized inference is going to stop at some point, so you need to have a good grip on your costs. You need to understand and forecast the cost of compute for a particular task, and understand as well what the ROI of that task is for the user. You need to understand what the users are willing to pay for and how much they are willing to pay and work within these constraints.

In the end, the answer to the question: “How do we build the most effective autonomous agents in a world built for humans?” is very simple: build for humans, and keep an eye on your costs.

I hope you enjoyed this essay, and if you haven’t subscribed yet to my newsletter please do to receive the upcoming essays directly in your inbox.

Every company aims to achieve sustained, high growth over time. But in a competitive and rapidly evolving market, growth is rarely a straight line. Instead, it follows a predictable pattern known as the S-curve, a natural trajectory observed in the lifecycle of products, services, and even entire markets.

Each S-curve represents a period of innovation that starts slow, picks up momentum, and eventually slows down as the product matures. To avoid stagnation, companies must strategically launch new products or services to stack S-curves, ensuring growth continues seamlessly. This approach requires a culture that values and supports “big bets” on bold ideas.

1. The Arrow of Ambition

This illustration depicts what I like to call the “arrow of ambition”. This is what every company dreams about when they think of their performance over time: an arrow that is always going upwards and to the right. This target represents the ideal trajectory of consistent growth that many companies aim for but rarely achieve in practice.

2. Performance is Never Really a Straight Line

In reality, growth follows an S-curve with phases of acceleration and deceleration. This illustration shows how real performance moves in waves, contrasting with the straight performance target line. The S-curve model describes three phases:

• Early Stage: Slow growth as the idea or product gains traction.

• Growth Stage: Rapid acceleration as market adoption peaks.

• Maturity and Plateau: Slowed growth as the market becomes saturated.

To achieve that straight-line growth over time, successful companies stack S-Curves on top of each other.

3. Zooming In: Stacking S-Curves

Now let’s zoom into the red box I highlighted in the previous section to examine how two S-Curves tack on top of each other. Anticipating when a current product will begin to decelerate and planning the launch of a new product to overlap with that phase is critical.

As one product reaches the later stages of the S-curve, the goal is to introduce a new product that is well positioned to begin its own growth stage. This way, the growth phase of the new S-curve kicks in just as the older product’s growth begins to slow.

Example: Apple’s strategy of launching new products like the iPhone, iPad, and Apple Watch illustrates this concept. Each device follows its own S-curve, and as one product reaches maturity, Apple launches another, creating a stacked growth model.

4. The Importance of Big Bets

Successfully stacking S-curves requires a company culture that encourages big bets on ambitious, high-stakes ideas. A culture of big bets goes beyond merely fostering creativity—it’s about taking calculated risks on high-potential innovations that could drive future growth. Here’s why this mindset is vital:

• Encourages Long-Term Thinking: Big bets often take years to fully mature, requiring companies to start early on the next big thing, even before the current product peaks.

• Reduces Stagnation Risk: A single S-curve approach can be dangerous if the market shifts or competitors innovate faster. Diversifying through big bets ensures resilience.

• Promotes Agility and Adaptability: With big bets, a company can respond better to market shifts, ensuring continuity even if one product line underperforms.

5. What Happens If You Don’t Launch New Big Bets

Without launching new big bets, performance eventually plateaus, and will almost definitely lead to a decline. Many people assume you can stay indefinitely in a state of plateau, but this is a naive idea that doesn’t take in consideration evolving competitive forces and technology advances. Companies that cling too long to one product or service may struggle to recover as market dynamics shift, and competitors introduce newer, more appealing options.

6. The Pitfalls of Moving On Too Quickly

This illustration captures the opposite risk: moving on to a new big bet before the current one has matured. The result is a series of incomplete S-curves, where neither project reaches its full potential. To avoid this, companies must find the right balance, allowing enough time for each product to peak while preparing the next big idea.

7. Creating an Ecosystem for Sustainable Growth

A culture of big bets is not about betting everything on a single risky project but about building an innovation ecosystem where multiple ideas are explored, nurtured, and scaled. Diversifying across projects provides flexibility and resilience. Key strategies to support this include:

• Cross-Functional Collaboration: Diverse teams generate more breakthrough ideas.

• Rewarding Smart Risks: Recognize and reward bold thinking, even if not all ideas succeed.

• Building a Flexible Innovation Pipeline: Keep several projects in the pipeline, enabling the company to pivot quickly.

Conclusion

Creating a culture of big bets is about more than just innovation—it’s a strategic approach to growth that anticipates the inevitable peaks and plateaus of the market. By stacking S-curves through big bets, companies can sustain momentum and agility, continuously adapting to stay competitive. The result is an organization ready to embrace change and prepared to grow, no matter how markets evolve.

Key Takeaways

• Growth follows an S-curve, and companies need to launch new products as others mature.

• Stacking S-curves enables smooth transitions and sustained growth.

• A culture of big bets fosters risk-taking, long-term planning, and adaptability.

• Understanding where a product is on the S-curve guides optimal timing for new initiatives.

• Building a diverse ecosystem of big bets ensures resilience and readiness for continuous growth.

In the end, cultivating a big-bet culture means choosing the right opportunities with the right timing and fostering an environment that values innovation and resilience. A company that embraces big bets can transform ambition into sustained success.

1. Put in the reps

Let’s just be clear: hard work is a basic requirement if you want to get ahead. If you want to prove yourself in a job, or build an impressive career trajectory, you will need to be more focused, more disciplined, and more importantly consistently good. I always tell people that consistently good is better than occasionally great, and just like in sports to be consistently good (and get better over time) you need a lot of practice and hard work.

I don’t subscribe to the ideology of hustle culture, working 100 hours a week, and all that non-sense. I believe in putting in the reps, and working hard, and yes sometimes this means you will have to work evenings or weekends, especially earlier on in your career, but I don’t think this is something you have to be doing all the time to be successful.

2. Be most helpful

Or in other words: jump on the biggest grenades you can find. One thing that has worked consistently well for me whenever I start a new role, is to do the workplace equivalent of punching the biggest guy you can find on your first day in prison: I find the hardest problem that everyone is afraid of touching, and I volunteer to solve it.

The diagram above is Daniel Debow’s Helpful Hierarchy:

Level 1: is when you tell someone there’s a problem, i.e. not very helpful.

Level 2: is when you tell someone you’ve found a problem, and some causes.

Level 3: is when you identify the problem, some causes, and offers a few potential solutions.

Level 4: is when you identify the problems, causes, and a few solutions, then make a recommendation on which solution to pursue.

Level 5: is when you go to your manager saying you identified a problem and its causes, determined how to fix it, and fixed it.

Get on that level 5 train as early as you can in your career, stay on, and you’re guaranteed to go very far.

3. Share your ambitions & ask for opportunities

This one is pretty simple and straight forward: Share your ambitions with people around you, especially your managers. If you want to become CEO one day, let everyone know you want to become a CEO. By being vocal about what you want, it helps you achieve a few things:

1. You make it real for yourself. Instead of this dream being only in your head as a pipe dream, it is now shared with others as the reality of what you’re trying to achieve. You have started day 1 on the path to getting to that dream.

2. When opportunities present themselves that could bring you closer to your goals, you’ll be the first person to come to mind for consideration. Many people get overlooked for opportunities simply because people don’t know that they’re interested.

But it is not enough to wait for others to identify opportunities for you. You need to ask for opportunities as well:

1. Ask for opportunities to practice or learn new skills, or to acquire new experiences. For example if you’ve never hired anyone before and you want to acquire that experience, ask to be included in the hiring team.

2. Identify bigger challenges and ask to take them on, even if they don’t necessarily fit into a specific role or promotion. See level 5 in the helpful hierarchy above.

3. Ask for promotions: If the bigger challenge you’re looking for translates to a bigger role, then take a risk and ask for that promotion even if you don’t feel 100% ready yet. Most people are figuring things out as they go, and it is okay if you do it too.

4. Take every promotion you are offered

Finally, my last piece of advice is targeted at a very particularly scenario that I have seen many times, especially when the person involved is a woman or a person from a marginalized group (i.e. people who have to deal with much bigger problems of imposter syndrome due to the average workplace hostility towards them):

Let’s say the person works as a software engineer and is doing a great job, and their manager comes to offer them the opportunity to be promoted to team lead / manager. What happens a lot of times is the person doesn’t feel ready, and passes on the opportunity with excuses like “I still feel I need to do more to prove myself first as an engineer before taking the promotion”.

The reasons why I say always say yes to a promotion are:

1. First of all, unlike what you might think, you can always get your old job back. No manager in the world would choose losing someone who has been a high performing individual contributor over giving them their old job, unless they are complete morons and in this case you’re better off finding another job anyway.

2. You will most likely do much better than you give yourself credit for, but the most important thing you will never know until you try.

3. Every promotion, regardless of how good or bad it goes, comes with “career capital”. Career capital refers to the abilities and resources you accumulate – whether skills, credentials, connections, or savings – that allow you to do more with your career in the future. It is like being at the casino, collecting chips, that you can cash in at a later stage in your career, but unlike the casino even if you fail you still win some chips.

I hope you found this helpful and I wish you the best on your journey. If you have any feedback or questions, you can find me on Twitter/X: https://twitter.com/_mohannadali

Subscribe now

* Criiiinge *

Every time I see one of these “going-back-to-the-office-to-hussle-harder” tweets, I cringe, a lot.

Let me start by getting a few things out of the way. When you say you tried remote and it sucked, I believe you! I’m sure remote sucked for you. I also know that you’re definitely doing it wrong and this is why it sucks more than usual.

• Your version of remote vs. back to the office: the office wins.

• Back to the office vs. remote work done right: remote wins, or at the very least goes head to head.

Now before I go on explaining why I believe that to be true, let’s just stop acting like there are no real examples of successful remote companies. Why do people keep conveniently ignoring companies like Gitlab? $5.6 billion market cap, $500+ million ARR, 2,300+ remote employees. I would also like to humbly use Hotjar as an example, with high 8 digits of ARR and 400 remote employees. Clearly some people know how to make it work.

Now onto the reasons why your remote experience sucked.

1. Lack of transparency & documentation

“There is no serendipity in remote work” - Alex Jost my colleague at Hotjar 2 weeks after I started.

Said another way, nothing happens by chance in remote work. Information doesn’t travel organically, so you need to have high intentionality about how you design and engineer communication flows.

This means you need to get into the habit of (a) writing and documentation, (b) having a high degree of transparency and access to everyone in the company. This means anyone can easily find and consume any information they need, with little to no red-tape, thus significantly reducing the need for information to travel through informal channels such as water cooler or kitchen conversations.

2. Invest in building relationships in real life

People think of remote workers as these anon avatars on reddit, trying to get some real work together, hiding behind their keyboards without ever meeting or building meaningful relationships with each other. Real remote companies invest time, energy, and money into making real relationships happen not only online, but also offline.

At Hotjar we do this in 3 different ways:

1. Every year we fly every member of Hotjar to a 1 week company retreat, where we align on strategy, have fun together, and more importantly build shared memories.

2. Every Hotjar team member has an annual “work together” budget of €2,000. This is a budget our team members can use to go work in any city where another team member lives. Our team members have done work togethers in amazing places such as Costa Rica, Cape Town, and Buenos Aires.

3. Team offsites are also organized regularly. This means that team are encouraged to meet in person to get things done and collaborate in real life on business challenges.

3. Invest in your team’s working conditions

When people are working in an office, companies have dedicated people to improving the workplace conditions of the team. From comfortable chairs, desks and any equipment the teams needs, to catered lunches, infinite snacks, and Friday beers. However when the team is working remotely, the working conditions of the team are often an afterthought. Companies don’t provide budgets for home office setup or equipment, they don’t provide budgets for beverages, high-speed home internet, or any other forms of comfort necessary for the team to do their best work.

If it’s painful to work from home, it is unlikely your team would be willing to put in the hours to get meaningful outcomes.

I can keep going and point more differences of how remote work differs from one company to the other, but I think you get the idea.

If you’re interested in remote work and want to learn more from our experience at Hotjar, get in touch with me and I’d happily answer any questions.

You can also come meet me at @CollisionHQ June 26-29. I'll be delivering the keynote on the remote stage on June 29th, as well as moderating the first half of the day. Hotjar will also be hosting a private happy hour on June 27th, so let me know if you'd like to be on the list

Subscribe now

My colleagues at Hotjar have probably heard me say the following statement at least one thousand times:

“Everyone, working on the most important things, every day”

This is my aspirational goal for how efficient I want us to be at Hotjar. We believe in a healthy work-life balance, and we demand 100% of our team - nothing more and nothing less. Yet we believe we can continue building a hyper-growth business without stretching our people too much. How you might ask? Superior prioritization.

Now let’s start by exploring some of the ways some of you are doing prioritization wrong, starting by the very obvious one:

#1 You’re not prioritizing at all

If you’re reading this, chances are you are doing some form of prioritization. I don’t think in this day and age there are any startups that don’t at least attempt to prioritize and focus on what is important. But let me tell you a story that will sound all too familiar:

You spend some time deciding on what your priorities are for the next quarter, but you’re an ambitious and hard working individual so you end with a list of around 10 “priorities”. The quarter kicks off and you start working on these priorities, but you uncover a few other problems along the way, so you start adding more things to your list. Half-way through the quarter your list of “priorities” now has 20 things on it, and you feel way too overwhelmed and you wonder if your problem is time management.

Let me start by telling you, don’t worry your time management skills are just fine - your problem is prioritization. First of all, there isn’t such a thing as 10 priorities because 10 is way too many. If everything is important then nothing is. I know some people go to the extreme of saying you need the one singular priority, but I find this to be too restrictive. What has worked well for me is a simple rule of thumb: just 3-5 things. The other thing to talk about is what to do when new things come up after you’ve set your 3-5 priorities. Mike Tyson has famously said “Everyone Has a Plan Until They Get Punched in the Mouth” - it is absolutely normal for things to change along the way and for new priorities to emerge. The most important thing is replacing priorities instead of adding priorities; in other words something needs to get off your list for something else to get on your list.

#2 You’re prioritizing based on gut-feeling vs. based on numbers

Most of us claim to be more data-driven than we really are. I’ve seen product managers for example believe themselves to be very data-driven because they spend time looking at metrics in Mixpanel or Amplitude, but when it comes to prioritization you end up with a RICE matrix that looks something like this:

I don’t have any particular problem with the RICE framework, although I don’t personally use it. The main issue that I have with this example is, how are you evaluating Reach, Impact, and Confidence?

For example Initiatives 2, 4, 5, all have a Reach of 5/5 - Does this mean they have very similar reach? or just different degrees of “very high” reach? If the latter, then it is not helpful, because when we are prioritizing we want to understand which one is “the highest” vs. trying to know which ones are roughly “very high”. Moreover, how are we evaluating reach in the first place? Is it simply our estimate based on gut feeling or is it based on actual figures we see in the data? If its based on figures we see in the data, then let’s please use these figures.

It gets worse in my opinion when we look at impact. How are we evaluating impact? Is it our impression of how valuable or important something will be? Or have we built a business case with real numbers and calculations to estimate the impact of our initiative? Many teams get bogged down by trying to be super accurate with their calculations and end up failing and reverting back to finger-in-the-air estimates. Your calculations don’t need to be accurate, they just need to be directionally correct. I always tell my team I want to understand the rough size of something more than the exact sie: Is this the size of this opportunity roughly $500,000, $1,000,000, $10,000,000, or $50,000,000? Build the muscle of doing back-of-the-napkin math to estimate impact and you’ll go a very long way.

#3 You’re not considering everything when prioritizing

When we are prioritizing most people are trying to answer this one question:

Out of all the things that I know, which ones are the most important things to work on?

But this is the wrong question. The real question is this:

What are the most important things that I should work in general?

Let me explain what I mean by this. When we are evaluating the importance of the initiatives we’re trying to prioritize, we shouldn’t just ask ourselves how do they measure up against each other, but we should also ask ourselves whether they are big enough to matter in the first place, and whether they help us achieve our goals.

For example imagine you’re trying to come up with some initiatives to generate $20 million in revenue. You also know you can only do 5-6 of these in any given year. This means that each initiative should contribute $3-$5 million in revenue. When you’re building your business cases, or reviewing the list of initiatives and ideas you have, it is not enough to just order them according to size, but you should also discard any idea that is worth less than $3-$5 million. If you end up not having enough ideas to generate the upside you’re expecting, you need to go back to the drawing board, and let me tell you it is much better finding out before you start working than finding out at the end of the year when you’re $10 million short.

If you liked this post, please share this newsletter and follow me on Twitter @_mohannadali

Share Mohannad’s Newsletter

In my experience working at startups, I found one thing in common among the most successful people and the most successful companies, and that is their ability to reason about risk in a way that others usually can’t.

Before we dive into this post, I want to get one thing out of the way. If you, dear reader, would stop reading here, there is one key takeaway I’d like you to have, and that is that almost everything that you consider to be risk, isn’t actually risky at all. Most people lean naturally to risk-aversion, and while there are no hard and fast rules, it is still very likely that you are exaggerating the risk involved in any given decision. People often tell me that I have a higher tolerance for risk but that isn’t true - I am just very pragmatic about whether something is a real risk or not and I don’t let my emotions control that decision. Now let’s dive into our topic.

Understanding risk: Risky != Speculative

One common mistake I see people making is conflating risk and speculation. Doing something risky is not the same as shooting in the dark. Shooting in the dark is reckless, whereas taking a risk is a calculated move. Taking risks is about understanding the different possible outcomes and their likelihood, and making a decision based on your willingness to accept failure in exchange for a potential higher outcome.

If the risks that you are taking are not grounded in data or a solid understanding of the parameters of the decision, then what you’re really doing is gambling and not risk-taking. Our jobs as startup operators sometimes need a little bit of gambling, but as we are going to see in the next section of this post, you should carefully manage the amount of gambling in your work and keep it to a minimum. I like to tell my team that my job is 90% chess (calculated moves), and 10% poker (gambling), so every now and then I’ll ask them to take a chance with me and try out something audacious, but the rest of the time we’re methodical, structured, and grounded in data.

Balancing risks: A diversified portfolio

When it comes to balancing risk, John Cutler does a fantastic job explaining this in the framework you see above.

You need to think of your risks as an investment portfolio. A good investor doesn’t put all their money into high risk assets, but they rather diversify. To construct a good investment portfolio, you will need a good balance of high-risk opportunities, medium-risk opportunities, and low-risk opportunities. If you’re an investor and you put 100% of your money in crypto, then you might see outsized returns but you might also see all or your investments go to zero. On the other hand if you want to play it safe and and put all your money in bonds, the chances of your investments going to zero are very low, but your upside is capped at a couple of percentage points a year. A diversified portfolio hedges your bets and gives you the opportunity to strike the right balance between potential returns and potential losses.

Take more risks: Continuous Validation

As we saw in the section above, every team should decide for themselves what is their equivalent of their “risk” quota, and in order to maintain a diversified portfolio, they need to maintain that quota and never really go under nor really go over. A “risk” is inherently time-bound too. Sooner or later, a risk should end up either materializing, or simply disappearing. The way this usually happens is after enough time passing, the moment of observation comes and we can clearly see whether what we thought might happen, in reality happened or not.

In the context of running a startup, you don’t want to spend a lot of time “waiting” to validate your risks, so this is why you need to tackle your risks (especially the bigger ones) early and often. Prototyping and experimentation are common ways for teams to validate their risks quite early in the process, and forcing that moment of observation without having to wait too long. Once you have validated you risks, you free up some of your “risk” quota, which you should then reinvest by taking additional risk and repeating that full cycle again. That’s how you keep innovating, how you keep pushing the envelope, and how you drive high growth in your business.

I’d love to hear from you, so if you have any thoughts you would like to share, please do so in the comments, or find me on Twitter.

Why do you need a sales team in product-led organization?

I think one of the biggest misconceptions about Product-Led Growth (PLG) is that you shouldn’t need a sales team. In fact many people go as far as considering having a direct sales team almost antithetical to PLG. In this post I will try to explain why that is not an accurate understanding of PLG, why you need a sales team when selling a bottom-up product, and the key differences between PLG sales and traditional sales.

The case-study of Atlassian

Atlassian are often cited as the poster-child of product-led growth, and the best example of that bottom-up flywheel that is 100% based on inbound interest in your product. Here is how Atlassian themselves explain it in their F-1 filing:

Our high-velocity distribution model is designed to drive exceptional customer scale by making affordable products available via our convenient, low-friction online channel. We focus on product quality, automated distribution and customer service in lieu of a costly traditional sales infrastructure. We rely on word-of-mouth and low-touch demand generation to drive trial, adoption and expansion of our products within customers.

All of this is factually very true, and it is a model that — when executed well — creates very efficient distribution that doesn’t rely on traditional sales organizations. That being said, what most people don’t know about Atlassian, is that do rely a lot on sales through channel partners and resellers.

In the 2021 Annual Report (page 20), Atlassian mentioned that they currently derive approximately 40% of their revenue from channel partners’s sales efforts. This has grown a lot since 2015, where channel partners and resellers generated 25% of Atlassian’s total revenue.

In addition to relying on sales teams at channel partners and resellers to drive such a significant portion of their revenue, Atlassian today employs a significant internal direct sales organization to sell to their enterprise customers.

Layering sales on top of PLG is a common playbook in SaaS

Atlassian layering sales on top of their strong product-led flywheel is not the exception, but rather the rule in the market.

Monday.com is another great example, of how a very strong PLG organization was able to accelerate their growth significantly by layering a Sales and Partners organization. The following image is from their S-1 filing, and you can very clearly see the inflection point after they created their sales organization:

But Monday.com aren’t the only other example. Every product-led leader we have even seen grow to become a large company have layered top-down sales at some point in time:

Why do you need a Sales team in a PLG Go-To-Market Strategy?

1. Meeting the needs of more sophisticated buyers: As your company grows and your bottom-up motion starts attracting bigger companies, you will often find yourself facing buyers who have a need to work with a sales person. Their needs can range from operational needs such as procurement requirements, to more product-related needs such as understanding whether your product can cater to some of their more complex use-cases or not. Product Managers often think that the product should be so self-explanatory that there shouldn’t be a need for a human-assisted buying process, but the product alone will never be able to compete with an empathetic two-way discovery conversation between a sales person and a customer.

2. Better conversion rates and higher ACVs: Sales teams very often beat self-serve in terms of conversion rates and ACVs (for bigger customers) due to the simple fact that they are able to tailor exactly the explanation of the value of your product to the needs of each particular customer. They also help by explaining which plan is the right one for them and avoid a situation where the customer ends up getting little value from your product because they are on the wrong plan.

3. Consultative selling leads to better retention: while customer retention is commonly a Customer Success job, when your sales team does consultative selling, they prime the customer to get value from your product on day 1. Consultative selling makes the relationship between the sales person and the customer that of a doctor-patient relationship as Mark Roberge puts it in this book the Sales Acceleration Formula, and that builds great trust in your product and your team.

What is the difference between sales in a Product-led org vs. a Sales-led org?

1. Allocation of budget and investment: Product-Led companies typically spend significantly more on R&D than they do on Sales and Marketing. Sales-led companies often spend between 50%-100% on Sales and Marketing.

2. CAC/LTV: Product-led companies need to layer sales in a way that doesn’t break their existing CAC/LTV ratios, whereas Sales-led companies would structure their CAC quite differently due to their fundamentally different sales motion, that is largely made up of outbound sales as we’ll see in the next point. The relationship between your CAC and LTV is really about finding the right fit between your customer acquisition channels, and your model (how you charge and the ARPA) as Brian Balfour explains it in great detail in this article on his blog.

3. Inbound vs. Outbound: Speaking of customer acquisition channels, product-led companies who have built a strong bottom-up motion have the advantage of being able to cast a very wide net in the market, which means they are able to capture leads of all sizes. This gives their sales teams a ton of inbound demand that they can prioritize and ultimately work with a large volume of high quality leads. In the absence of this high volume of inbound demand, Sales-led companies spend a lot of their time and effort on outbound cold-calling and pipeline generation with sales development representatives (SDRs).

I’d love to hear from you, so if you have any thoughts you would like to share, please do so in the comments, or find me on Twitter.

Egypt and the UAE are two of the hottest startup ecosystems in the MENA region, so in this post I compare the evolution of Early Stage VC in both countries.

10-Year Trends

Egypt and UAE are almost head-to-head when it comes to number of investments.

Between 2012 and 2016, the UAE startup ecosystem significantly outpaced the Egyptian ecosystem in terms of number of investments per year. From 2017 onwards however, we see Egypt catching up with a steep increase in the number of early-stage investments, to even outpace the UAE in the years of 2019 and 2020. Egypt in 2021 saw a decline in the number of early stage investments as we saw in my previous post, but the UAE saw very interesting growth in the same year.

UAE significantly outpaces Egypt in terms of total invested capital

When we look at the total invested capital instead of number of deals, we see a very different story with the UAE consistently outpacing Egypt over the last 10 years. This is driven mainly by a much stronger Seed-stage investment market, as well as — on average — bigger Pre-Seed and Seed investment rounds as we will see later in this post.

2021 Trends

UAE shows very strong growth in Seed stage investments

Egypt outpaced the UAE in terms of Pre-Seed investments (UAE lower by -57.68%), and they went almost head to head in Series A (UAE higher by 10.48%), but it is really in the Seed Stage where the UAE shines with a +318.76% difference compared to Egypt. UAE Seed stage startups raised a total of $161.7m, compared to Egyptian See Stage startups who raised only $38.6m.

Series A rounds in Egypt are bigger than the UAE

When we compare the Median Investment by stage in Egypt and the UAE we see Egyptian startups doing better in Series A, but we see startups in the UAE doing better in Pre-Seed and Seed:

• Series A: Egypt $15,000,000, UAE $5,966,202 (-60.23%)

• Seed: Egypt $525,000, UAE $1,200,000 (+128.57%)

• Pre-Seed: Egypt $125,000, UAE $290,000 (+132.00%)

The top Egyptian Startups are raising more capital than the top UAE Startups, except in Seed rounds

As you can see in the charts above, the top 3 Egyptian startups in both Pre-Seed and Series A have raised more capital than their counterparts in the UAE, while in the Seed Stage the UAE startups are doing significantly better than Egyptian startups.

Key Takeaways

Investors in Egypt need to write bigger Seed cheques, and they need to write them more often

I personally believe that the ticket sizes that we are seeing in Egypt are too small to enable any fast development of the ecosystem and they don’t give startups enough capital to find product-market fit. This is probably because of some legacy investors who started investing 5-6 years ago in EGP instead of USD (the currency exchange rate working against them), and have set the market standard for the average ticket size. The market dynamics have changed quite a lot since then, and the talent market is becoming increasingly more competitive, and investors need to catch up.

Investors also need to take more chances on more entrepreneurs. As I wrote in my previous post, this concentration of investment in a very small number of companies is even detrimental to the success of some of the VCs themselves who have a very small portfolio of companies.

Egyptian outliers are worth betting on

As we saw in the data, there are some remarkable outlier startups in the Egyptian ecosystem. They have done a great job in building successful businesses and were able to raise significant amounts of capital. This means that investors have a great opportunity in the future growth of current and upcoming Pre-Seed and Seed stage companies in Egypt, but for them to really find these outliers, they have to - you know where this is going - take more chances on more entrepreneurs and write bigger cheques.

Disclaimer: All information is based on data available on crunchbase.com

The Venture Capital Scene in Egypt, on the surface, looks like it has been growing significantly over the last 3 years. This is very clear when you look at the total invested capital by stage:

• Pre-seed: $5,707,417 (2019), $8,946,699 (2020 +56.76% ), $34,453,144 (2021 +285.09%)

• Seed: $20,533,000 (2019), $24,863,000 (2020 +21.09%), $38,620,000(2021 +55.33%)

• Series A: $14,540,000 (2019), $18,100,000 (2020 +24.48%), $199,898,491 (2021 +1004.41%)

It is understandable that 2020 saw modest growth given that it was the first year of the pandemic and investors were largely cautious due to the uncertainty of how things are going to unfold over time. In 2021 however, we have seen an explosion of venture capital around the world, and Egypt was no exception.

Have early-stage investors taken more chances on more entrepreneurs in 2021?

The surprising answer is No. If we look at the number of deals per year and stage, we see a very different picture. While we have seen a small uptick in Series A rounds, the number of both Seed and Pre-Seed rounds went down in 2021. Also keep in mind that the absolute number of Series A deals per year is very small (7 in 2019, 9 in 2021, and 11 in 2022).

• Pre-seed: 82 (2019), 57 (2020 -30.49% ), 50 (2021 -12.28%)

• Seed: 66 (2019), 72 (2020 +9.09%), 50 (2021 -30.56%)

• Series A: 7 (2019), 9 (2020 +28.57%), 11 (2021 +22.22%%)

More capital is concentrated in a smaller number of deals

In general it is true, that investors are investing more per company than before. The median investment amount by stage has grown consistently over the last 3 years, however the biggest growth we have seen is really driven by the top 25% of companies in every company stage.

Let me give you an example. Let’s look at the median invested amount in the Pre-seed stage.

• For the lowest 25th percentile of companies, the amount grew from $25,000 (2019), to $31,125 (2020), to $57,500 (2021).

• The 50th percentile, the amount grew from $47,300 (2019), $81,850 (2020), to $125,000 (2021).

• The top 25th percentile, the amount grew from $60,000 (2019), $192,500 (2020), to $1,000,000 (2021).

As you can see the top 25% have seen the biggest jump in the size of their investment rounds, with everyone else not really seeing a significant change in the amount of funding they are receiving. This shouldn’t come as a surprise with outliers like Rabbit raising $11m, Telda raising $5m, and Flextock raising $3.5m.

The top 3 companies in the Pre-seed stage (Rabbit, Telda, and Flextock) raised more than 56% of all invested Pre-seed capital in Egypt in 2021, with Rabbit single-handedly responsible for 32%.

We see a very similar story if we look at Seed stage investments, albeit not as dramatic. You find outliers in the Seed stage as well with Taager raising $6.4m, and Sympl raising $6m, and Cartona raising $4.5m.

The top 3 companies in the Seed stage (Taager, Sympl, and Cartona) raised 43.8% of all invested Seed capital in Egypt in 2021.

When we look at Series A investment rounds, we haven’t seen the top 25% of deals grow in size in 2020 at all, however in 2021 that changed with a whopping 681% increase. The biggest Series A deals in 2021 were MaxAB raising $55m, Trella raising $42m, and Capiter raising $33m.

The top 3 companies in the Series A stage (MaxAB, Trella, and Capiter) raised 65% of all invested Series A capital in Egypt in 2021.

What does this mean for founders?

It is becoming more competitive to raise capital in Egypt, but if you are successful you are likely to raise more capital than you would have in previous years.

I believe this is driven by the uncertain economic climate due to the impact of the pandemic, leading to a smaller number of investors participating in high-risk investment opportunities such as venture capital, especially when the public equities market has shown a very remarkable performance in 2021.

This is further highlighted by the fact that if we look at the last 3 years, 2021 was the lowest in terms of of the number of new emerging local VC funds.

What does this mean for investors?

There is a very big opportunity for investors to take more bets on more up and coming startups. As it is commonly said, venture capital returns follow a power-law distribution, and therefore the success of venture capital firms relies on taking more changes to find the one of two winners who would make the fund profitable.

I hope with the correction we are seeing in the public equities market, we would see more institutional investors taking the opportunity to invest in early-stage venture capital in Egypt.

Disclaimer: Availability of investment round figures

The data available in Crunchbase does not provide exact figures for the amounts invested in each company. The data available though provides a coverage that ranges between 55% and 83%, so I believe this is good enough to at least directionally provide some insights on the evolution of the early stage VC market in Egypt.