Eran Arkin
/therealarkin
Visit me to learn about Tech Careers, Product Management More. Experienced operator; Beginner online writer. Dad.
Eran Arkin
4y ago
I recently talked with a brilliant startup co-founder that relies on a highly specialized supply to deliver a differentiated service. Things are going well for them now, and they are looking to scale. The bottleneck, not surprising, is scaling the supply while keeping the demand's satisfaction level the same.
While from the dawn of history, tools & machines augmented humans and enhanced their productivity. Automation is increasingly peeking at jobs requiring more complex and less routine skills. It is only accelerating due to COVID. And, I think it's an exciting trend, I wanted to share a few lessons from experts -
Rory Sutherland: Map out the entire value created, don't fall prey to the 'Doorman Fallacy.'
In his book alchemy, Rory defines the doorman fallacy by giving this example:
"You define a hotel doorman's role as 'opening the door,' then you replace his role with an automatic door-opening mechanism… The problem arises because opening the door is only the notional role of a doorman…. in addition to door-opening: taxi-hailing, security, vagrant discouragement, customer recognition, and signaling the status of the hotel. The doorman may actually increase what you can charge for a night's stay in your hotel."
You need to carefully think about how the end-to-end experience creates value for consumers, including the functional service itself, the 'hospitality,' what goes beyond the minimum, & 'signaling.' Don't fall prey to the Doorman Fallacy.
Peter Thiel: Understand the strengths of humans & machines
Peter Thiel wrote in his (highly recommended!) Zero to one book: "men and machines are good at fundamentally different things… We form plans and make decisions in complicated situations… Computers are exactly the opposite: they excel at efficient data processing, but they struggle to make basic judgments."
My daughter recently started using Stitch Fix, which prides itself on combining data science and human judgment stylists to deliver a new shopping experience. In a Reddit AMA, stylists expressed precisely that:
The machines algorithmically curate a set of items from millions of options as a starting point.
Stylists apply judgment based on their familiarity with the customer, especially if they change style (i.e., going to WFH from business wear), and put together a Fix by "under 15 minutes" (!).
Rishi Mandal: Give Humans the Right Tools & Evaluate
Future built the largest body of professional trainers (after the US army), I have used the service for over a year, and I really enjoy it.
Rishi, Co-Founder and CEO, talks about some of the tools they are using to scale the coaches' interactions:
The Workout Builder saves coaches time doing the "IQ" part of coaching - recommending and making it easy for coaches to build a set of weekly exercises & goal tracking.
The command Center enhances the "EQ" part of coaching -
CRM - add & retrieve details like the names of the customer's kids
Communication Tools - enhanced chats with snippets from the customers' apple watch
Workflow & task management, i.e., reminding coaches what to do next
He also talks about coaches' evaluation system, which includes 'functional components' that are measurable like response time and evaluation criteria for the "EQ" part based on sample communication.
Based on 37 coaches' glassdoor reviews, with an average of 3.9*, things are going pretty well for them. Though not everything is smooth, of course, as one coach puts it:" "adding new tasks and responsibilities every time a minor improvement in tooling is made results in a net-zero improvement in work/life balance." So based on that review, the tools are getting better, but coaches asked to do more.
Eran Arkin
4y ago
I got an intriguing question following my why MBA essay - “If I don’t care about the fun aspects of an MBA, what are some ways that I can achieve the same benefits without going through the “trouble” of doing a two years MBA.”
Come to think about it; an MBA is a bundle that can get unbundled based on individual needs.
But first, the most critical process is to understand your gaps-
Where are you today?
Where do you want to be?
What are the gaps between the two?
For example: In 2009, when I applied, I was a software engineer in a Business to Government Israeli company and wanted to work on “the business side” in a consumer tech in Silicon Valley. A partial list of gaps -
I didn’t have work authorization in the US.
Didn’t have any credibility with US employers
Light to no business/consumer experience
And while MBA seems like the shortest path for me, it might not be the shortest path for you. Here’s how to unbundle it -
Benefits that have obvious alternatives
US Work Authorization: With remote work and Big tech massive expansion, you may not want it. But if you do - an internal transfer with a multinational corporation seems straightforward these days.
Coaching: MBA program provides you access to lots of career coaches services. You can hire great career coaches, and I’d recommend also doing that after graduating.
Academics: You can take online classes from top schools on any topic these days. Some even will host office hours for online students.
Benefits that have some alternatives
Signaling: ‘a proof you did something hard’. This is the hardest to replicate. Being accepted to a top school still has an enormous signaling value. All the eight PM/MBAs in my cohort at Google attended top schools.
Interesting projects: I did several real-world projects during the MBA. One of them was a go-to-market for HP labs. I learned a lot, and it gave me another interesting topic to talk about in the interview process.
These two go hand in hand. You can create or participate in interesting projects that can have signaling value and allow you to meet like-minded individuals and build your network. Starting a company, a side project, or getting involved in an open-source project can go a long way. Even creating and selling notion templates can be valuable ways to status signal.
Benefits that have partial alternatives
Alumni Network: My first manager at Google was a Berkeley Alum. I think it did play a role. Alumni also helped me with my transition to Facebook.
Peers: Huge value, its hard for late 20s something to suddenly not only boost their LinkedIn graph, but create lifelong friendships.
You can build your own alumni network by working in big tech in a non PM role. These tech companies are a great way to meet new people through interests groups and alumni networks. Cohort-based learning is also a good way. I’m now on #Ship30, and while it’s virtual and less structured, there are valuable connections you can make.
To recap - understand your own gaps, compare an MBA with alternative experiences and get on the career path you want.
Eran Arkin
4y ago
Humans evolve to multitask (i.e., breastfeed while walking), but multitasking isn't suitable for tasks with a high cognitive load.
There's contradictory advice in popular productivity books: Eating the frog popularized the idea of doing one thing at a time (with the most difficult first!). Yet Getting Things Done argues for doing any task that comes if it takes less than two minutes.
Research shows that you shouldn't try to do two things at the same "atomic unit of time" (participate in a meeting while attempting to finish PRD for another project), which will increase the error rate and waste time. BUT the reality of a product manager's work is that they are part of multiple project teams.
"Great things in business are never done by one person, They are done by a team"
Since our job is interdependence, it's impossible to do one project at a time. PMs often wait for other functions to complete their work or an experiment to end. Finally, there are some real benefits of working on a few things:
Allows your mind to process things offline and develop better ideas.
Some tasks are inherently easier than others, so switching provides a mental break from difficult tasks.
Here's how to be productive while driving multiple projects:
First, ruthlessly prioritize limiting the number of projects. Align with your teams & manager on these high-priority projects. If there are still too many, discuss who can drive the project. You'll be pleasantly surprised as teammates will step up!
Shreyas proposed the 3-5-7 framework: each individual has three main projects they are working on. Research shows that individual and team performance has benefited from a limited number of multiple team memberships, particularly for more experienced employees, due to an increased focus on working efficiently, and transfer of learning experiences.
Once you decide on the highest priority projects, here are a few more tips:
Tip #1: Preserve your state in writing. Much like memory-cache helps computers retrieve the information they are likely to need faster. Writing enables you to create your memory cache to minimize switching costs after you are done working on a specific project and BEFORE you switch jot down where you are.
Tip #2: Have a system to review all projects weekly. Set a time to check status (Friday afternoon is my favorite time since it allows me to preserve state for next week).
Tip #3: Set deadlines for dependencies Deadlines contribute to accelerating the pace of group work. Use them effectively to understand when to switch.
Tip #4: Allocate thinking time per project. When I worked on Google Travel, I ran two big initiatives with multiple projects. Over time I learned to allocate a few half days every week to each separately, which was immensely helpful.
To recap - As a Product Manager, you have to switch between tasks & projects effectively. I laid some ideas that worked well for me, and I would love to learn what worked for you!
Eran Arkin
4y ago
I frequently get asked this question. I graduated in 2012 with an MBA from Berkeley, and I believe it's a very subjective question.
Let's start with a Twitter tech myth: "MBA is a waste of time." Silicon Valley loves narratives, particularly the engineer dropout narrative, consider Musk, Page, Brin dropping out from Ph.D. programs, or Zuck drops from undergrad to start a company. It's a prime example of cherry-picking & don't forget that these 'dropouts' had strong networks partly because they were in those institutions in the first place.
It depends
It's not too hard to cherry-pick and find successful MBA entrepreneurs, not to mention executives. Consider Phil Knight (Nike/Stanford), Scott Cook (Intuit/HBS) Scott McNealy (Sun/Stanford). Or top CEOs like Satya Nadella (Chicago), Tim Cook (Duke), and Shantanu Narayen (Adobe/Berkeley), to cherry-pick a few. Even Elon's Tesla CFO holds an MBA. At Google, 5 out of my nine managers throughout the six years were MBAs, and three more had other advanced degrees. But correlation doesn't imply causation.
The truth is there are no right or wrong answers; it depends on your situation, goals, and skill gaps.
For example, in the middle of the great recession, I was a software engineer in Israel. I applied to a US MBA program because I wanted to work in Silicon Valley (screenshot from my essay) and experience top-notch education. There was no way I could work as a product manager at Google in 2012 without an MBA, 5x my pre-MBA salary. It opened the door for me to a US work permit and Google. Other huge benefits were more flexibility with my newly born daughter during school & a lot of fun in a supportive environment with classmates from 30+ countries who are good friends to this day.
Evaluate These First:
What are my career goals? Will MBA help you get there? Especially relevant for those who want to switch industries or geographies.
What is the actual cost of this? Can I get a scholarship? Consider the loss of two years of income & career progression.
How much do I value (additional) lifelong friends? This is one of the most significant benefits of an MBA IMO. In my class alone, we had seven marriages out of 240 students.
Can I realistically get accepted? The potential is bigger in the higher-ranked programs.
The world today is different than it was ten years ago. You can learn & network a ton on LinkedIn, Twitter & YouTube. But spending two years with top-notch classmates & faculty can still be an invaluable experience.
Eran Arkin
4y ago
Middle managers are, well, stuck in the middle. They don't drive day-to-day product building, coding, or marketing. And, they aren't yet executives with much power over resource allocation.
They mostly suffer from a bad reputation in Tech. Few examples I've read (and sometimes experienced) -
Waste their teams' time because of the value add disease: "Too often they look for something – anything — to change so they can feel confident they did their jobs. "@boz
There are plenty of middle managers in FAANG; what separates the exceptional ones from the average ones? Let's dive in:
They Successfully Start New Product Lines
Middle managers are closer to end-users, the latest trends in technology, and day-to-day operations than most senior managers. At the same time, they have the big picture understanding, organizational know-how, and credibility with executives to pitch, approve and lead new initiatives. Facebook's App Install Ads or Google Classroom are just two examples of industry-changing products, started in big Tech by middle managers.
They Help Drive Change
Change is inevitable. Our industry moves fast, and our org charts are always a work in progress. Middle managers engage with lots of 1:1s down, up, and across the org, which means they are the best source of organizational knowledge. They can help senior leadership conceive reorganization to better align with changing landscape and, more importantly, help bring people on board to such changes without significant attrition or churn.
But Also Keep Executive Accountable For Continuity
Your teams reviewed a quarterly roadmap, got everyone on board, including executives. A few days or weeks into implementation, a new shiny/urgent priority came in 'from the top.' What do you do?
Strong middle managers gather more context before rushing to action. Since they
Know their people & detailed work well
And, understand that both too little and too much change lead to underperformance
They ensure to balance continuity & tackling the new 'hot' priority. Much like taking a curve while driving.
Finally, They Truly Coach & Uplevel Their People
Most people conflate 'promotions' with 'progressional growth'; surely they are related, but not the same. I've seen, at times, managers fight hard in a committee to get their reports promoted without putting in the work to uplevel their skills. While promotions are nice, longer-term people will benefit more from managers who understand gaps between their reports career goals and where they are today and recommend actions to bridge them.
Being a middle manager is HARD, and being exceptional is even harder. I would love to learn from your experience as a manager and a report on what makes middle managers exceptional.
Eran Arkin
4y ago
Human behavior is a complex topic, and Fogg’s model is a great way to understand it. It’s simple yet powerful. It can help you design for behavior change and debug when users don’t
I was the first PM on Waze Carpool. It was all about behavior change, as 80% of people commute alone. I read a ton of research on behavior change, persuasion, etc.
The more I researched the topic, the more I wanted to learn directly from BJ Fogg. I attended his workshop and had a blast. Let me share the one model that I keep coming back to:
For a behavior to occur, three conditions must meet at the same time: Motivation, Ability, and Prompt
Motivation - person’s level of desire to do a behavior.
Ability - person’s level of Ability (or lack of) to do a behavior.
Prompt - No behavior happens without a prompt. It can be external or internal. An email with a call to action or your stomach telling you it’s time to eat.
Motivation and Ability work together and have compensatory relationships. The more challenging something to do (=lower Ability), the more Motivation you need to do it and vice versa. Tax filing is painful (no thanks, intuit), but 95% of US households jump through the hoops because we are highly motivated to avoid IRS penalties.
Motivation is hard to count on
Motivation is flaky. Many make New Year’s resolutions; however, 80% drop them by the second week of February. From my experience - most product experiments that target ways to increase motivation, make modest gains, or fail.
Instead, utilize times when motivation is high to ask users to do hard things that will prompt and make future actions easy. For example, if a user just fully watched a video and liked it, ask them to enable notifications (=Prompt) for that creator to learn about new videos. If you are motivated at night to exercise tomorrow morning, set the alarm clock (=Prompt) & and get your gym clothes ready by your bed (=Ability).
Make the behavior easier is usually the way to go
Start small, ask users for little. Reduce friction wherever possible. Save them time and money.
Removing required sign-up before checkout increased sales by $300M /year for this e-commerce website- “I’m not here to enter into a relationship. I just want to buy something.”
Think how Robinhood made it free to trade stocks on mobile, then reduced friction even further by letting them buy fractional shares.
Don't forget messaging. We tested messaging that emphasized the ‘easiness of the process (i.e., set carpool easily in 2 clicks) vs. messaging to increase motivation, and the ‘easier’ messaging won every time.
Try the model yourself to understand any behavior. Sign up for BJ’s free tiny habits course to learn firsthand about behavior design. Use the model, and let me know how it goes!
Eran Arkin
4y ago
I'll walk you through how AI product management differs from traditional product management based on my experience working on traditional and ML based projects at Google, Waze & Meta.
If you haven't yet worked on a problem requiring ML, you will likely encounter one in the next decade as “Software Is Eating the World, but AI Is Going to Eat Software”.
Making the transition isn't hard as it may sound. Let's dive in.
Online Data is Not Only the Output. It's a Key Part of the Input
Savvy product managers proactively utilize existing sources of data and find new ones. Think about how TikTok is designed: because there's exactly one video on the screen, the program knows exactly how long you watched and what actions you did on the video. This enables it to recommend the next video almost perfectly.
You Don't Control the Recipe, Just the Outcome
In traditional software, the user inputs, and the product team determines the exact steps to generate the outputs. In ML, you define the desired outcome, specific performance criteria to be optimized, and the program determines the steps. The beauty of it is that the program can discover some unobvious ways to solve problems [watch this video!], but....
Be Careful How You Set the Desired Outcome
Let's say you run a content recommendation feed and optimize for clicks. Your engineers launch a new model, clicks go up, and everyone gives high fives. You look at the top content dashboard, and suddenly you'll see a lot of bad stuff, clickbaity, low-value content, soft porn, etc. You didn't create the product experience or feeling you wanted, and advertisers may boycott you. Or you might push users too aggressively to harmful rabbit holes.
Junk food is immensely popular & addicting, but it might not be the restaurant you want to create.
Your XFN is Larger
To name a few common additions:
Specialized ML Engineers to build models.
Project managers & ops people to manage offline rating processes.
Privacy experts: you'll want to retain data longer since data is fueling model performance
You Don't Need to Know the Ins & Outs, But Understanding ML Concepts is Important
There are less than 100 car manufacturers in the US and more than 200M drivers. Drivers don't know how to create a car, but they know how to drive & read the dashboard.
Some key things you'll want to know (DM if you want a post on the best resources) before you are ready to drive -
Supervised / Unsupervised learning
Optimization & Loss function
Precision, Recall, False Positive, True Positive, False Negative
While There are few differences outlined above, the most Important PM skills, are the same: critical thinking, communication, and user empathy
Good luck in your ML Journey. Drop me a note or DM if you have any feedback. Thank you for reading
Eran Arkin
4y ago
"Software Is Eating the World, but AI Is Going to Eat Software." As NVIDIA CEO described 2017. And yet, not every product problem is a good fit to be solved by AI. As a Product Manager, you need to focus the team on the user problem and the business outcome. AI/ML is a powerful tool at your disposal.
Many jump on the bandwagon of ML without thoroughly assessing if their problem is the right one and if their organization is ready.
Use this quick checklist to assess if your product problem is solvable by ML:
The problem requires making a lot of decisions.
It takes a couple of editors to determine the order each day of articles in the printed version of a paper. However, it will be infeasible for Twitter to employ editors to sort its feed per person. Examples for such problems include:
Recommendations: People who bought this product also bought this, people you may know, feed ranking, etc.
Pattern recognition: Email spam labels, determining if a picture contains a face, etc.
Each decision is a low consequence.
If we get it wrong once in a while, no one will die. ML isn't perfect. While it performs many tasks better than humans, it can't drive a car on a snowy day or land an airplane. High stakes usually (still) involve humans.
Humans can agree on the "right" decision to make
Either the decision has a right answer, ("ground truth"). We can agree this email is Spam, and the other isn't, or we agree on the outcome of the decision. For example, by watching this video, the user is more likely to come back to our platform tomorrow.
Your organization has access to or can generate (or buy) a lot of relevant past data.
While there are some developments in this area, and new techniques like GAN generate data to learn from, you need a strategy to collect data in most cases. Options include:
Collect data in the product (Video completion on YouTube)
Human Raters (Google Search, TikTok Integrity); or Drivers (Waymo)
Buy data from other companies or trained models.
This is key; without data, your models are worthless.
Your organization is committed to this path; you have the talent (or can hire) & resources, and we understand this will take time to get right.
For example, at Google, we developed a model to determine how important each landmark in the world is for tourists (Search: "Things to do at <city>"). We had a world-class team and leveraged all google advanced AI and troves of data. Still, it took us a year to get a result we put in front of users. And, another year of iteration to a level where we proud of. We heavily utilized rates and travel editors for ground truth data.
On top of this, you need to retrain models periodically, keep your data fresh, and don't forget about technical debt.
I hope this note will help you determine if your product problem and organization ready to use AI. I'd be super grateful if you drop a line on Twitter for any feedback or share your learnings!
Eran Arkin
Headcounts Planning
4y ago
Since you need headcounts for the right reasons (Stop and read Ask Additional Headcounts For The Right Reasons first!), it's time to get into tactics on HOW to ask for them.
Unfortunately, many (including myself a few years ago) aren't aware of these tactics because it's often not widely talked about within organizations.
While following these steps will not guarantee you additional headcounts, it can certainly increase your chances. Let me share with you what I've learned:
Step 1: Understand Company Specific Considerations
Every company has a headcounts planning season bi-annually or annually. Ask around and understand when's the right time to ask. For example, if the company planning cycle is in November, and you are only ready with your ask in December, you almost certainly missed an opportunity. So plan ahead!
Besides timing, which is the most critical, there might be a template or other company-specific quirks such as joint funding across orgs. Talk with your manager/org leader to understand these considerations.
Step 2: Give Leadership Confidence That The Team Has a Capacity For Hiring
While you determine that you have a gap to fill, you need to make sure you have the capacity for hiring to close that gap. Headcounts give leadership optionality, granting you them to you means they are loosing optionality, so they want to make sure you can fill in the headcounts they'll grant you. Sometimes it's a 'use it or lose it' situation. Capacity means that you have the right hiring managers in place, and, for critical roles and clear job description written.
Step 3: Make Your Case and Give Options
OK, you are ready to make your ask! Since you did your homework (if not, read Ask Additional Headcounts For The Right Reasons). Provide "Menu of Options," as Jackie Bavaro writes here. Explain the potential impact of each option; if you determine that you need five more to establish the full potential, show what you'll do with only three and what you can achieve with no growth. Comparing three options in a table on a slide will do the work.
The prep work you did with the right reasons and the steps here will all ladder to get your headcounts request successfully. Even if you didn't get them, it will help your team focus and set the right expectations with leadership in case of no growth.
Eran Arkin
4y ago
I’ll help you determine whether you should ask for additional headcounts as a team leader. This advice applies directly to Product Leaders & Engineering managers in medium to big tech companies but can be helpful in other situations.
Growing headcounts for the right reasons are often essential to help current team members avoid burnouts and generate more impact for the business. The outcome of more impact and supporting a larger organization is career growth, as most tech companies and the ladder rewards larger orgs.
Unfortunately, some people don't critically evaluate whether they need more heads and grow their orgs for the wrong reasons (Shreyas, did you have Google+ in mind when you wrote this?), but more often than not, it will not lead to long-term success.
Step 1: Evaluate & focus your team's scope
Like gardening, you need to "prune a plant to encourage new healthier growth." Sometimes the reason you think you need more HC is that your team's scope is incoherent, or you work on too many fronts.
The hardest part of this step is the internal conversation. For example, when I started leading Facebook's Video Search the team was the primary owner of at least 2-3 other areas. I had to align first within the team that Video Search is the most critical area, and that other areas, while important, are less promising. Because one of the engineering managers wanted to develop one of his reports to lead the gaming search area, I had to show them that there are still tons of growth opportunities within a narrower area to secure buy-in. In my experience, once you align on the scope internally, it's usually easy to align with leadership to prune.
Step 2: Within the agreed areas of focus, where do you need to be? What is success?
So many people skip this step and make wrong assumptions. They don't align on success means before asking for headcount growth. Align on what success is in the most specific way explicitly, i.e., within two years, we get 40% of the market as measured by XYZ, or we will solve 90% of integrity issues without human intervention in 12 months, etc. Setting & aligning ambitious, exciting & meaningful success criteria is critical to successfully asking for more headcounts.
Step 3: What will it take to get to the desired end state? Map out gaps; they might be bridgeable without additional HC.
Now that you align with leadership, the scope & success criteria go back within the team and make a high-level plan to get there. Making a plan deserves a few posts, but let's assume you created one and identified the critical gaps. Brainstorm within your team how to do more with the current resources you have
Is our team working efficiently? Maybe we are just unproductive or underutilized?
Do we have the right talent profile? You may have the right number of headcounts but lack a specific skill within the team.
Are there some gaps solvable by leveraging resources within the company or buying from the outside?
If everything is optimized, you are ready to ask for more headcounts! In the next post, I'll show how to ask for Headcounts successfully.