Term 7 · Module 3 of 4

Choosing the right startup to join

Startup Dilemmas Founding or Joining

Practise this module in Claude, ChatGPT or Gemini

Set it up free
Mind map

Evaluating Startup Health: Why One Dimension Is Never Enough

Joining a startup means handing 2-3 years of your career to people who made their key decisions before you arrived. The earlier module asked how to judge a founding team from the inside (how relationships form, how roles are played, how rewards are divided). This module asks the same three questions from the joinee's seat, with far less access.

Evaluating Startup Health: Why One Dimension Is Never Enough

Case: is Zepto "healthy"?

The question is not about venture performance but about Zepto as a place you would want to join.

Evidence What it suggests
The Ansh Nanda equity dispute (informal equity that cost him a stake now worth billions) was eventually resolved: a documented settlement and a role split that has held for years. Palicha runs strategy and the outward face; Vohra runs product. A bonded founding team with clear roles, at least on paper.
A Reddit post, claimed to be from a former employee, described a culture built around the CEO's schedule: meetings starting at 2:00 AM because Palicha does not wake early, long working hours, experienced professionals avoiding the company, and a staff mostly early in their careers. Palicha did not deny the meeting times and explained them. He also admitted the app had used manipulative dark patterns to extract more money from customers, and said they have since been removed. Possible culture and ethics concerns.
  • Both stories are true and publicly documented. An outside evaluator usually meets only one, depending on which article or video comes first.
  • The lesson is not about Zepto. It is what happens when a diagnostic measures one dimension and treats it as the whole answer.
  • The honest answer to "is it healthy?" is that relationships, roles and rewards must be discussed with founders early. Founders must talk to each other about their positions, and employees must go and ask the founders.

The five lessons of the module

Joinee's question: is this team healthy?
  • Founder behaviour and power
  • Culture and employee signals
  • Learning environment and your role
  • Your own honesty as a candidate

The module asks: how do dimensions actually get made once a founder has real power? What can be observed from outside, including signals nobody puts in a pitch deck? Will the environment teach you something, even a hard thing? What are you being offered and what are you giving up?

  • A co-founder negotiates roles and equity before day one. A joinee inherits the existing structure and reads it from outside, through a handful of interviews and whatever is public. These are the only data available, so the skill is reading a healthy team accurately with less access.
  • Three of the five lessons are actionable: reading attrition, reading equity documentation, and reading the POSH committee. Each gives you something to search, verify, ask, or negotiate.
  • Looking at whether a company has investors, and who they are, is useful context, but a first-time joinee has no lever over it.

Key takeaways

  • A single-dimension diagnostic gives a misleading verdict; Zepto has both a healthy-team story and a culture/ethics story.
  • A joinee inherits structure and must read it from outside with limited data.
  • Apply the same three lenses (relationships, roles, rewards) from the joinee's seat.
  • Several lessons are actions (search, verify, ask, negotiate), not just observation.

Founder Behaviour and Power

Founder mode vs manager mode

Paul Graham's essay (based on a Brian Chesky talk at Y Combinator, and worth reading in full) attacks decades of conventional advice that once a company scales, the founder should hire good people and get out of the way.

  • This conventional view parallels the earlier effectual to causal shift: early entrepreneurship is effectual, and as a company scales it flips to causal, with professional managers running things.
  • Graham says it is wrong: founders who delegate too early and trust an org chart they cannot see into watch the company drift from its mission and vision.
  • Chesky took some decision rights back when he saw Airbnb drifting, and says the company became healthier, not worse.
  • Experienced operators push back: "founder mode" is just a fancier name for micromanagement. The debate split people who agree on most else. Some read the essay as licence to control everything; others as a narrow point about one stage, over-applied. Neither side is completely right.
  • Why it matters to a joinee: nobody calls themselves a micromanager, but many founders call themselves "in founder mode." The label cannot be trusted; behaviour must be observed from outside.

The test: does disagreement survive?

Functioning founder mode Dysfunctional founder mode
Founder asks hard questions, pushes back hard Yes Yes
Target of the challenge The decision Quietly becomes the person
Result People leave having heard something, even if they lost the argument The room learns to stop bringing bad news instead of arguing better
Likely signal Someone can tell the founder they are wrong and still have a job next morning Attrition rises

Ola vs Airbnb

  • Airbnb: the founder saw drift, took ownership of many decisions, and turned the company around.
  • Ola: Bhavish Aggarwal is in strong founder mode, personally involved in decisions at multiple levels. Reports cite a huge senior-executive turnover: roughly 13 departures over 5 years, sometimes 45% in one year, and a hired CEO who left within 4 months. The result is a "revolving door."
  • Same mode, different outcomes. The question is whether disagreement survives contact with the founder.

What to ask in the interview

  • You can ask directly: "Do you allow people to question you?"
  • A founder who answers with a specific story, including a time they were wrong and changed course, is telling you something different from one who gives a general philosophy about an open door and welcoming criticism.
  • A vague, undocumented claim suggests no specifics survived a confrontation with the founder. Specifics reveal how power is used day to day.

Key takeaways

  • Founder mode versus manager mode is the founder-involvement debate; the label tells you nothing, behaviour does.
  • Healthy involvement targets the decision; unhealthy involvement targets the person and silences bad news.
  • Ola's executive churn vs Airbnb's turnaround show the same mode producing different outcomes.
  • Ask for a specific story of being challenged and changing course; vagueness is a signal.

Reading Company Culture and Employee Signals

Using attrition to investigate a startup

Saying Ola's founder is dysfunctional in this way is not saying he is a bad founder: Ola built a very large business and is doing well despite current hiccups. It tells you what kind of company you are getting into. At your joining level it may matter somewhat less, but consider the logic: someone who gets shouted at by the founder and cannot shout back will pass it on downward. A dysfunctional org culture permeates.

Search LinkedIn: joiners and leavers, last 1-2 years
  • Where did leavers go?
    • Education: Natural progression
    • Another company: Talk to them
  • Left in under a year?
    • Yes: Red flag, scaled by company size
  • Check how founder reacts to ex-employee posts
    • Separation was not bad
  • What to do: use LinkedIn to see how many people joined, how many left, how long people stayed, and where they went. Two years of data is enough.
  • Leavers' destinations: young people leaving for education is natural progression. If they joined another company, see who the person is and consider talking to them.
  • Red flag: people leaving in under a year, read against company size. With 6 people, or even about 30, it is hard to judge; dig anyway.
  • Founder's digital footprint: check whether the founder comments on posts by past employees (a sign the separation was not bad), and look at what the founder says and supports on social media. Social media is somewhat make-believe, but patterns appear with deeper digging.
  • Even for a small company you can leverage the abundant available data and compare with others in the industry.

Evaluating startup culture as a woman

A more specific case, like couple-founders earlier: what does company structure reveal about how a company would handle a woman's complaint and career once inside?

  • POSH Act (India): every employer with 10 or more employees must have an internal complaints committee. Even if not required, ask whether there is an outsider (a mentor, advisor, or company member) a woman could talk to.
  • Interpretation caution: having a committee does not make a company great, and lacking one does not make it bad. It is a cue, not a verdict.
  • Context: much of the startup world has been a toxic, masculine, "geekish" environment. Uber is a case in point. Women often tolerated it because the company was good, excused it, and got used to it, at some cost to their mental equanimity.
  • Cues to watch: meetings at 2:00 AM, and parties deep into the night. These are not reliable indicators: a respectful company can have a partying culture. Culture is hard to judge, so be cautious.
  • Due diligence:
    • Check the website for a committee; ask about it and see how they respond; ask who the committee members are. It indicates seriousness and that the founder wants a healthy, safe environment for creating value.
    • Ask a woman who works or has worked there.
    • Look for a woman co-founder or leadership-team member; things are usually much better and she can be asked directly.
    • Watch gendered attrition: a low overall attrition rate with only women leaving is a cue.
  • Why gender matters: startups are stressful but also motivating (high dopamine from hard problems, e.g. space startups sending rockets), yet not equally comfortable for everyone. As in the couples discussion, society expects different things of men and women; when someone in the family falls ill, women report it more often falls on them.

Key takeaways

  • Attrition is researchable: LinkedIn joiners/leavers over 1-2 years, leavers' destinations, sub-1-year exits.
  • Interpret attrition against company size; small teams make it noisy.
  • A founder's reaction to ex-employees' posts and social activity reveals the quality of separations.
  • For women: check for a POSH committee (mandatory at 10+ employees), women in leadership, outside contacts, and who is leaving.
  • No single cue is conclusive; combine several and ask people with first-hand experience.

Learning Environment and Your Role

Edmondson's learning-zone framework

Instead of gut feel, use Harvard's Amy Edmondson's two axes: how safe people feel raising a problem, and how much is expected of them.

Low expectations High expectations
High safety Comfort zone: pleasant, but nothing happens Learning zone: intensity converts into skill
Low safety Apathy zone (called "empathy zone" in the captions): nobody is pushed, nobody cares Anxiety zone: real pressure; people hide mistakes instead of surfacing them, so nothing is learned
  • Google's internal research on what separates its best teams independently reached the same idea: psychological safety was the strongest predictor of high-performing teams, ahead of who was on them. It means being comfortable expressing views, and others being comfortable too. The debate is about views, not people.
  • A founder-driven, hard-charging startup can be either the anxiety zone (all pressure, no safety; teaches only how to survive stress) or the learning zone (real accountability; mistakes named and worked through, not buried). They look identical from outside but have opposite value for a first-time joinee.
  • Startups are always high-activity. Some leave you rejuvenated after a hard day; others leave you drained. That depends on founders and culture. A learning zone gives progress and a "warm glow" that snowballs into the startup's progress; an anxiety-ridden one involves blame for things even outside someone's control.

Cues to look for

  • How does the founder talk about a past failure, in an interview or podcast?
  • Do early employees who left describe specific things they learned?
  • A founder in Graham's genuine founder mode (deeply involved, focused on the decision, not the person) is much more likely to run a learning-zone team, since that involvement comes from fast, honest feedback.
  • Ola is suggested as a case where many teams would sit in the anxiety zone.

Rate of learning beats pay and title

  • At your stage, the rate of learning matters more than pay or title. Judging companies by salary ("what is your LPA?") is called one of the worst evaluation methods. The lecturer is unhappy about how popular media and campuses focus on LPA, and this is one reason for discomfort with placements in the programme.
  • You are a "time billionaire": optimise for better learning and better positions. Starting small does not mean you stay small; starting large does not mean you become CEO or partner. It depends on how quickly you learn, interact and get things done.
  • The manager matters: a great founder with a lousy manager teaches you little; a lousy founder with a great manager can still teach you a lot. Examine the scope of work, who you work with, what people say about the manager, and where the manager worked before.

Same title, different job

Two offers titled "product analyst":

Large company Five-person startup
Scope Owns one metric on one team Owns the whole onboarding
Decisions Waits for a decision through three layers of approval Makes the call
Feedback Slow Learns within a week whether something worked
  • Large-company coordination is highly standardised because nobody is indispensable: if someone leaves, a similar person replaces them.
  • Delta add: a founder once asked a PGP student choosing between his company and HP what her delta contribution to each would be: what she could push, own and create. She found this convincing. Her parents opposed (no experience, so join a large corporation), but she chose the small company.
  • Large companies do not give as much learning; what you do depends on how you play the cards. Joining McKinsey does not guarantee success, and joining a small startup on Bannerghatta Road does not preclude worthwhile work.
  • With about 40 years of career ahead in a very different, AI-shaped world, personality, ability to talk to people and ability to get things done matter more. "Knowledge is on tap."

Key takeaways

  • Edmondson's two axes (safety, expectations) give four zones; only high-high is the learning zone.
  • Google found psychological safety the strongest predictor of high-performing teams.
  • Intense startups may be learning zones or anxiety zones; probe how the founder discusses failure and what leavers say they learned.
  • Optimise for learning rate, scope, and manager quality, not pay or title.
  • The same title can mean very different ownership at a large firm versus a tiny startup.

Your Responsibility as a Candidate

Everything so far treats you as the innocent party. The other half: you may be the one misleading the startup. Many founders are naive and on their first venture.

  • Saying "I intend to stay three years" while having applied for a master's abroad is not fair.
  • Joining while planning to build something with a friend is not being true. State it upfront, e.g. that it is a completely different domain and done on the side.
  • It is not acceptable to be lax with truth and commitment because it seems tolerated. If you know from the start you will leave in four months, you bring disrepute to yourself and to the programme; employers may stop hiring from BBA DBE. Breaking a promise has repercussions.
  • Do not withhold information you are sure about. Disclose, to the extent possible, things like "I am doing something on my own," "I may get another offer," or "I plan to go abroad." The lecturer acknowledges that in India some of this is not received well, but a level of honesty keeps things on an even keel.
  • Symmetry: you ask founders honest questions (Did you fight? How did it go? Have people left, and why? Have there been disagreements and how were they handled?) and expect truthful answers; answer the questions they use to judge you honestly too.

Your Responsibility as a Candidate

Module synthesis

  1. 1Before joining
  2. 2Apply relationships-roles-rewards diagnostic from outside
  3. 3Read attrition, culture and learning environment as distinct signals
  4. 4Check yourself: am I as honest as I expect the founder to be?

Key takeaways

  • Candidates can mislead too: undisclosed plans (study abroad, side venture, planned short stay) are a breach of fairness.
  • Dishonest exits damage your reputation and the programme's.
  • Expect honesty from founders only if you offer it.
  • Be cautious about joining, but be honest about what you intend.

Interview: Founders of MyCaptain (Sameer Ramesh and Mohammed Zeeshan)

Two young founders, known to the interviewer since they were students, who came to NSRCEL (IIMB's incubator) and won its business plan competition. Their story is a case on problem-solution fit, equity, bootstrapping, culture, and exit.

Interview: Founders of MyCaptain (Sameer Ramesh and Mohammed Zeeshan)

Origin and first solution

  • Trigger: one night before an exam in a hostel room, three friends (an aspiring astronomer, an economist and a journalist) were all studying fluid dynamics in mechanical engineering. Three people from different parts of the country, passionate about different things, were stuck in the same degree. Something is wrong with how careers are chosen in India; seniors and batchmates voiced the same frustration. A personal problem, so it resonated; the cofounder, in Nagpur, joined on the first call without knowing the plan.
  • First solution: a magazine (an "unmitigated disaster" and, as one founder put it, "if your first solution is perfect, you waited too long"). Logic: schools pushed a magazine to students already (Hindu Young World was cited), so a more valuable one should also be pushed.
    • A virtual team of 50-60 volunteers (writers, designers in Pune, illustrators in Kolkata, logistics in Nagpur, the founders in Chennai) coordinated over WhatsApp around 2013-14, believing in the problem.
    • They printed 500 copies before speaking to any school or student, using money from a cofounder's mother given as a loan to be repaid. Nobody wanted to read a magazine.
    • After about 1.5 years of school visits across cities they had only about 123-124 subscribers.

Pivot, traction and NSRCEL

  • School events (initially to gain an audience) drew positive reception, even though the magazine did not. Sponsorships followed. The message connected with students in grades 9-12, validating the problem, so they pivoted to workshops and events in schools and colleges, then to online cohort-based courses (MyCaptain).
  • They won NSRCEL's business plan competition as, reportedly, among the youngest winners. The prize was ₹10 lakh, but they declined it to avoid giving 5% equity (the incubation required 5% equity); they had already pivoted online with early traction (about ₹4-5 lakh a month) and felt they could bootstrap, which became their DNA for the next 5-6 years.
  • Early product-market fit was present in the sense of clear problem-solution fit, with paying customers and word of mouth, around 2015-16 in their final year. Tech was stitched together on WhatsApp groups and Google Drive folders for the first roughly 2,000 paying customers: an MVP need not be a platform or website from day one.

Equity and co-founder decisions

  • Before the pivot, one founder asked the other to commit full-time with no guarantee of even a small salary, on one condition: no sitting for campus placements. The same expectation applied to any full-time co-founder.
  • Early hires who committed around graduation (about four in the group) were equity-based since there was little cash.
  • Equity framework: points were assigned for whose idea it was, contribution, when each person joined, the commitment (in years) they would give, and relevant skills. A weighted average emerged, then light negotiation; nobody hardballed because each ask was fair. A commonly accepted public norm: the first two co-founders split roughly equally (give or take 10-20 points) and the third to fifth co-founders usually hold under 15%. Later co-founders were diluted from this base ("diluting equally" in the ratio they had).
  • Who called the shots: one founder, the origin of the idea and the "CEO type." Responsibilities were split, and decisions were naturally left to him where appropriate.
  • One person did not join because the committed horizon was only one year, whereas the founders planned 5-8 years or whatever it took, even though that person had contributed significantly. They remain friends.
  • Co-founder conflict and equity division were common reasons for early-stage failure; similar conversations arose later with early employees as the co-founder group grew to about 4-5.

Hiring, culture and bootstrapping

  • First hires were sales, about 10-15 after the 4-5 core group, because they were bootstrapped and had a proven, scalable sales model. Logic: five sales hires can fund one operations hire. "Cash is king."
  • Why bootstrapped: in 2015-16 edtech was not fashionable (BYJU'S and Vedantu existed), raising money from a non-premier college was hard, cohort-based live learning was not seen as scalable, and investors preferred ops-light e-commerce, marketplaces and the like. Unit economics were very profitable, so they could sustain 2-3 years without outside capital. About 3 years after graduation (1.5 years after NSRCEL) they reached about a $1 million revenue run rate and began talking to investors.
  • Pricing and scope: courses at about ₹1,000 were compared to offline workshops (8-16 hours, same price): the online version offered a month of learning with a professional. Scope was horizontal (psychology, photography, writing, entrepreneurship, astronomy), not academic (JEE) courses, staying true to the original problem of pursuing what people want.
  • Recruiting for alignment: they could not match salaries (BYJU'S offered about 5 times as much), so they hired people who had faced the same problem. Recruitment was mostly fresher campus hiring, and the first interview question was "What did you want to do? Why are you here?"
  • Culture: a visible vision statement ("help every individual pursue a career they truly love") and a mission milestone (if a million people a year are forced into engineering via JEE, can a million be on the platform pursuing careers they love?). Internal talk was about purpose, not fundraising rounds. Peak bootstrapped learners: about 2.5 lakh.
  • At scale: the peak was about 375 people, 250 in sales across two product lines. Culture challenges arose in the large, target-driven revenue functions (typical of edtech sales), not in small specialised groups. Senior managers were themselves only 24-25 years old and carried culture forward. Culture began to shift when early believers left and new hires arrived.

Venture capital, exit, and vesting

  • Why they raised VC (2021-22): COVID made online live learning mainstream and attracted competitors; without raising and going aggressive they would lose market positioning. Raising changed the culture.
  • Why they sold: the edtech bubble burst and the industry was consolidating. Contrast: the first pre-series A (about $3 million) took roughly 2 months, mostly on phone calls; a roughly $2 million bridge took 1.5-2 years in small tranches. Options were to push 5 years toward profitability and an SME IPO at a business of about ₹100-200 crore revenue, or to sell and join a larger journey. They had received acquisition offers every 3-6 months and had 3 term sheets before signing with Ematicus. M&A, including equity discussions, was "brutal" and humbling.
  • Culture after the sale: the acquirer's leadership is 50-60% ex-founders, built inorganically, so each business function has its own culture and systems, shaped by its leader and stage. One founder still had a year of vesting left but moved on to the next venture, having judged his role complete.

How their thinking changed

Earlier Now
Founder A: missionary vs mercenary about 90/10 about 50/50 (more mercenary, still mission-driven)
Founder B about 80% missionary (stated earlier) about 60/40, still toward education (cautious about AI and ethical guidance)
  • Reason for the shift: "nobody is funding you for your mission"; with VC capital you must return it, and "no matter how impact VC you may be, you don't measure impact."
  • Selection filters: excitement first, with a clear outcome; then size of market (TAM) as a second or third filter. Scale equals impact. Exits need not be an IPO.
  • Changes they would make: do not make fundraising rounds the team's conversation; let founders and the finance controller handle metrics, while the team stays tied to the mission. Capital should drive the mission, not replace it. Cost functions take the hit when revenue functions stumble, and that meant hard conversations with close people. For product and program roles, hire from education nonprofits (Teach For India, SBI fellowship) rather than from edtech.

Advice for joiners and founders

  • When to join: as early as possible. After early traction and early signs of product-market fit is a good time, because you avoid the earlier mess. If you have extreme ownership and want founder-like involvement without founding, joining even before traction is fine. Not everybody should start ventures; early employees change companies' trajectories.
  • Missionaries vs mercenaries: assess which the founders are. Missionaries push through 1-3 years (about 1,000 days for a venture to figure things out); mercenaries leave if there is no traction in 6-12 months. Mercenaries are primarily driven by making money, missionaries by the problem. Their view is that the mix differs by geography (Bangalore is better than other places they saw, where education was treated as e-commerce).
  • How to find out: talk to founders, spend time with them, intern. You will sometimes get it wrong; after a couple of mistakes you develop a gut feel. Startups are about mistakes, including joining the wrong place.
  • If you plan to found: first work at a startup in that industry for learning and a head start.
  • If joining: keep trying; not every startup is equally good at everything. If the first fails to teach what you want, move to another.
  • If starting: just start, do not overplan; early traction matters more than any pedigree or deck.

Key takeaways

  • Their first solution failed, and the evidence (500 unsold copies, about 124 subscribers) pushed a pivot to events and then online courses.
  • Equity was set with an explicit points framework (idea, contribution, timing, years of commitment, skills), not ad hoc; commitment horizon excluded one friend.
  • Bootstrapping with sales-first hiring, purpose-led recruitment, and campus freshers built the early culture; scale and VC money strained it.
  • Exit choice: SME IPO vs acquisition, shaped by market conditions and repeated offers.
  • For joiners: join early, judge missionary vs mercenary, intern first, and accept some wrong joins.
Revise before your examAll 2 exam tips from this module in one list
  1. The real diagnostic question is "What happened the last time someone told this founder they were making a mistake?"
  2. Gendered attrition (low overall attrition but only women leaving) is a specific signal worth naming. A POSH committee is a seriousness cue, not proof of a good or bad culture.

Test yourself on this module

Questions from these notes, one at a time, with feedback. Your weak topics are saved.

Set it up free Takes about a minute. Free plans work.