创始人模式：Sajith Wickramasekara，Benchling创始人兼首席执行官

今天，卡罗琳和我在这里与Benchling的创始人兼首席执行官萨吉·维克拉马·萨克拉见面。欢迎你，萨吉。很高兴见到你。

Today, Carolyn and I are here with Saji Wikrama Sakura, the founder and CEO of Benchling. Welcome, Saji. Nice to see you.

我也很高兴见到你们，谢谢你们邀请我。

It's nice to see you too. Thanks for having me.

我们非常高兴见到你。你曾参加过2012年Y Combinator的夏季项目。

We're delighted to see you. You went through Y Combinator summer twenty twelve.

那已经是很久以前的事了。

It was a long time ago.

我们可是记得很清楚。

We know way back.

十三年了。

Thirteen years.

十三年前，你创立了这家公司。这期《社会雷达》特别节目非常特别，我们正在Y Combinator创始人静修营现场，周围都是这些了不起的创始人。我们正试图探讨什么是“创始人模式”，因为我们知道，所谓创始人模式，从定义上讲，就是创始人兼CEO所做的那些雇佣CEO不会做或做不到的事情。但我们还不清楚这在实践中具体意味着什么。

Thirteen years ago, you founded this company. And this is a special a special episode of of the Social Radars. We are here at the Y Combinator founder mode retreat, and we're surrounded by all these amazing founders. And we're trying to explore the idea of what is founder mode, because we know that founder mode, by definition, is doing things as a founder CEO that a hired CEO couldn't or wouldn't do. But we we don't know what that is in practice.

我们想收集一些案例，所以非常希望听听你在Benchling历史上使用创始人模式时的任何经历。但首先，我们想让听众了解Benchling是做什么的。

We wanna collect examples. So we'd love to talk to you about any experiences you had in founder mode in the history of Benchling. But we first wanna hear what is Benchling for our listeners.

当然。我是Benchling的联合创始人兼CEO。我们曾参加2012年夏季的Y Combinator加速计划。你可以把Benchling理解为为科学进步打造现代软件的公司。我们的客户是生物技术公司和制药公司的科学家，大约有20万名科学家在7000所大学和1300家公司中使用Benchling来研发药物。我们的软件帮助他们设计实验、记录数据、分析结果，并与他人共享信息。

Sure. So, yeah, I'm I'm the cofounder and CEO of Benchling. We went through the summer twenty twelve Y Combinator batch, and you can think about Benchling as we build modern software for scientific progress. So our customers are biotech are are scientists at biotech and pharma companies. It's about 200,000 scientists at 7,000 universities and 1,300 companies making medicines that use benchling, and our software helps them design their experiments, capture the data, and analyze it and and share it with each other.

你可能难以置信，但研发一种新药的过程极其艰难和复杂，成本高达25亿美元。90%的药物在临床试验阶段就会失败，而支撑这一行业的技术却是纸张、电子表格和电子邮件的混合体，相当落后。因此，技术是提高成功率、更快将药物带给患者的关键途径。

And you wouldn't believe it, but life science is making making a medicine is an incredibly hard, difficult process. Costs about 2 and a half billion dollars. Ninety percent of medicines fail when they're in a clinical trial, and the technology that powers this industry is a mix of paper and spreadsheets and and email, so it's pretty pretty archaic, and so technology is, like, a really important way to improve the success rate and and bring more medicines to people faster.

哇，我妹妹在一个相关行业工作，所以我对它的昂贵和复杂程度多少有些了解。

Wow. My sister works in an industry, and so I am sort of familiar with how expensive and complicated.

把人送上月球，可能都比研发一种新药容易。

It's it's it's probably easier to put a person on the moon than it is to make a new medicine.

没错，而且临床试验实际上分为四个阶段，整个过程既漫长又昂贵。你们能量化一下你们的软件在这20亿美元的流程中节省了多少成本吗？

Yeah. And there's Literally. Yeah. There's, like, four phases of clinical trials too, so it's just like the whole process is lengthy and expensive. Have you been able to quantify the dollars that your software saves in this $2,000,000,000 process?

当然可以。这显然是我们希望最终客户感受到的：买了Benchling，就能更快地把药物送到患者手中。我们实现这一点的方式是将问题拆解成不同部分，看看如何加速每个环节。比如，我们可以衡量科学流程中不同环节的效率：在特定时间内能完成多少工作流程，或者完成某个流程需要多少人力。我们通过将问题分解为更小的部分，证明我们的技术能大幅提速。实际上，在新冠疫情期间，有几种单克隆抗体就是如此。

Yeah. That's it's definitely obviously, that's a you know, we'd love for, at the end of the day, for customers to feel like, oh, I buy bench, and I get my medicines to patients faster, and I think the way we can do that is to break down different parts of the problem and how we can speed them up. You know, the different parts of the scientific process where we can measure how much actually science can get done, how many of a certain workflow can get done in a period of time, or how many people are needed to run a certain process. And so we have ways of breaking down the problem into smaller pieces and showing that they can be done way faster with our technology. Actually, during during COVID, there were a couple of the monoclonal antibodies.

但你知道，在疫苗出现之前，有几种抗体实际上是那些完全使用Benchling进行科学研究的公司开发出来的，你知道的

But, you know, before we had vaccines, a couple of those antibodies were actually made by companies who did all their science on Benchling and, you know

不会吧。

No way.

显然，我们并没有做过A/B测试，比如对比使用Benchling的客户和不使用Benchling的客户，但他们确实能以惊人的速度推出这些成果，你知道的，实验室里那些在做艰苦关键工作的科学家才是真正的前线人员，但正是他们加上机器人技术以及Benchling这样的软件，才让他们比其他人快了太多。

Obviously, there wasn't, like, an AB test of one of our, you know, noncustomer not using Benchling versus the one that did, but they were able to, like, get these things out in record time and, you know, come at you know, there's scientists in the lab who are doing the hard critical work. Those are the people in the trenches, but, like, that plus robotics plus software like benchlink allowed them to go way faster than everyone else.

你们现在有多少人？因为你们已经存在十三年了。

And how many people are you now? Because you've been around thirteen years.

时间真长啊。是的，我们大约有600人，总部位于湾区。我们在波士顿也有办公室，那里显然是一个重要的生物技术中心，同时在欧洲也有分支机构。

It's a long time. Yeah. We are about 600 people based in the Bay Area. We have offices in in Boston, which is obviously a huge biotech hub, and then also in in Europe as well.

600人。

600.

600人。

600 people.

很多。对，大公司。

Lot. Yeah. Big company.

所以从一开始你什么都不知道的时候，你就一直在做这个。

So you've been running this from the very beginning when you didn't know anything.

对什么都不知道。

Didn't know anything about anything.

顺便问一下，当你在2012年夏天参加YC时，你刚开始创业，是个全新的公司吗？就像，刚刚

Well, by the way, did you start at y when you did YC in summer twelve, brand spanking new company? Like, just

一个全新的公司。我当时是麻省理工学院的大三学生，主修计算机科学，但也在好几个生物实验室工作过。我一直觉得奇怪的是，所有那些软件人员——也就是我所在的群体——他们有非常好的协作工具。对，而生物领域的人员却只有纸和笔，again，纸和笔。

Brand spanking new company. I was I was a junior at MIT. I studied CS, but I worked in a bunch of bio labs, and it was always weird to me that all the software people, which was one of the crowds I was with, they had great tools for working together. Yeah. And then the bio people had just, again, paper paper Yeah.

这让我觉得太疯狂了。

It was crazy to me.

对，我记得我们对这个想法非常兴奋。我们特别喜欢程序员也懂一些医学知识，能够弥合这一鸿沟。

Right. I remember we were very excited about that idea. We love when programmers also have, you know, medical kind of knowledge and that kind of thing where they can bridge that gap.

这是一次很酷的跨界。

It's a cool crossover.

是啊，当时只有YC的人相信我们。

Yeah. YC was the only people who believed at the time.

真的吗？你去过其他地方吗？

Really? Had you been to other No.

投资者？即使在YC之后，我们依然觉得，如今，一些YC合伙人还会把我叫回去，跟新一批学员谈话，提醒他们：只要坚持下去，即使在演示日之后没有成功也没关系。

Investors? Even after YC, we were, I feel like I, nowadays, I feel like some of the YC partners all they they bring me back to talk to the batches to remind them, it's okay if you're not successful after demo day as long as you stick with it.

哦，那我们该快速聊聊这个了。你们在演示日时达成了什么成果？融资进展如何？

Oh, well, we should we should quickly touch on that. How how what did you what were you able to accomplish by demo day, and how did your fundraise go?

是的，当时有很多大学用户在使用我们的工具。实际上，PG给了我们非常棒的建议，他说：那些真正每天使用你软件的人，比如从事生物学研究的人，如果有一天你成了他们不可或缺的工具，那将变得极其重要。因此，我们有很多大学用户在使用我们软件的早期版本，但我们一分钱都没赚到。

Yeah. We we had lots of users at universities using our tools. And, actually, PG gave us really great great advice of, you know, it was like, hey. The the people who are if they are living and breathing in your software, and these are the people who, like, work with biology, like, someday that will, like, be really important if you are the de facto tool for them. And so we had lots of users at universities using their early versions of our our software, but we're making no money.

是的，每个投资者都说：生物学？这似乎是个很小的市场。而所有生物学领域的人又都说：这确实很酷，但我们做的是药物，不懂软件。所以我们当时处于一个很尴尬的境地，但非常幸运。

Yeah. And every investor was like, oh, biology, that seems like kind of a small market. And all the biology people were or all the all the biology people were like, oh, this is really cool, but we make drugs. We don't know, like, software. And so we were kind of in this weird place, and we were very lucky.

我们总共开了大约一百次会议，筹集了670万美元。我现在甚至都不告诉创始人我们的估值上限。

Like, we took about a 100 meetings. We raised $6,700 k of money. I even I don't even tell our valuation cap to the founders now.

他们根本不会相信，他们

They they wouldn't believe They

他们不会相信我。但这足以让我们起步，我们坚持了下来，最终从中打造了一家出色的企业。

they wouldn't believe me. And that was enough for us to get off the ground, and we stuck with it and eventually built a built a great business out of it.

多年来，你成长了，你也必须转变为一名管理者。

So over the years, you've grown. You've had to have turned into a manager.

有时候，是的。

Sometimes. Yeah.

给我们讲一个你想到的、关于你不得不进入创始人模式的故事吧。

Tell us a story that you you can think of that has to do with when you've had to go into founder mode.

是的，对我来说，创始人模式几乎等同于拥有主人翁意识，这才是关键区别。我希望公司里的其他人也能感受到这一点——我希望员工们也能拥有这种主人翁心态。对我来说，当一名主人意味着能够自如地在业务的各个层面运作，并对每一个细节都充满个人责任感，无论它多么微小。作为创始人，我希望我的高级领导者也能这样想：当然，你们是公司战略和愿景的守护者，但你们也愿意去修改新闻稿里的文案。

Yeah. I think, like, to me, founder mode, I I sort of think of it synonymously with just having an ownership mentality, and that's the big difference. And so hopefully, it's not something people at the company also feel like, you know I hope people at the company feel like they can also have that have that in them. Being an owner to me means acting sort of being comfortable operating at a lot of different levels of the business and and feeling a lot of personal care over every detail no matter how small. I I think, like, as founder, and I'd want my senior leaders also to feel this way of, like, yeah, you're obviously a steward of the company strategy and vision, but you're you're willing to edit copy on the press release.

或者你会在某个设计中注意到像素关闭了，然后你就会对此发表意见。这就像当你看到低于你标准的东西却放任不管时，你实际上是在说这没问题。对我来说，这种主人翁意识就是创始人模式的体现。曾经有一段时间我远离了这种心态，那正是我最糟糕的时期。

Or you're gonna want you're gonna notice when the pixel is off in some design, and you're gonna say something about it. And it's kind of like anytime you see something that's below your standard and you let it go, you're sort of saying that's okay. And so to me, it's it's that ownership mentality that that represents founder founder mode to me. And there have been periods of time where I got away from that, and those were the worst parts

好的，跟我们讲讲你曾经远离这种细致入微状态的一次经历吧。

of Okay. Tell us about one of those times when you felt you were away from that, like, being so detail oriented.

我想，确实有一些阶段公司人员规模增长很快，我们招聘了一些高级领导者。你知道，随着公司发展，我们雇用了高管。招聘高管是个颇具争议的话题，尤其是在这个会议上。我雇过一些对业务产生变革性影响的优秀高管，也雇过并不合适的高管。我认为最难的一点是，当你招聘一位新高管时，要分清哪些是他们带来的专业领域知识或独特技能，哪些是公司文化。举个具体例子，当我们规模变大时，有一段时间我和高层领导层逐渐与客户脱节了。如今，即使公司刚起步时，我每天都会和客户交流，有些周甚至会和十到十五个客户交谈。

I I think, like, there were there were definitely some periods of faster headcount growth, and I had hired you know, as we've grown, we have hired senior leaders. Think hiring executives is a kind of controversial topic, especially at this conference, and I've had great executives who have been transformative for the business, and I've had executives who haven't been the right fit, And I think one of the trickiest parts is trying to when you hire a new executive, trying to understand what is something that represents the domain expertise or the unique skills you hired them for, and what's the culture of the company. So just to give you a specific example, at a point when we we got bigger, there's a period of time where, like, I and senior leadership got kind of disconnected from our customers. Like, these days, I and even when the company's getting off the ground, I talk to customers, like, every day. Talk some some weeks, I talk to ten, fifteen customers.

我经常亲自拜访我们的用户，那些购买我们软件的高管。我花大量时间去获取第一手的真实情况。

Like, I'm constantly visiting our users, the executives who buy our software. I'm I'm spending a lot of time getting my own ground truth.

嗯。

Mhmm.

但曾经有一段时间，我觉得我手下的高级领导更专注于打造一个能与客户沟通的“机器”。从理论上看，你或许会说：好吧，也许我亲自与这么多客户交流确实不够 scalable，现在公司变大了，想达到新高度，这种方式已经行不通了。于是你就能合理化地想：也许他们不需要那么贴近客户。但问题是，我认为高级领导的行为会反映整个公司的状态，公司整体就会变成那样。

And there was a period of time, though, where I felt like I had senior leaders who were more focused on building a machine that could talk to customers. And you can, from a theoretical sense, could say, okay, well, maybe me talking to all these customers by hand was not a scalable thing, now that we're bigger and trying to reach this new height, it's it's just not gonna work anymore. So you can kind of rationalize, like, okay. Like, maybe they don't need to be that close to customers. But the thing is, like, I think people like, the behavior of the senior leadership reflects, you know, the the whole company becomes like that.

于是，我最终陷入一种状态：有好几层员工根本不再与客户交流，我当时就想：如果我们都不接触客户，怎么管理面向客户的部门？这种情况是缓慢而逐渐发生的，我后来学会更明确地表达我对高级领导的期望，并重新确立文化标准。而早期，我曾误以为：哦，公司变大后，这本来就是正常做法——让团队专门负责这些事。我完全错了。

And so I end up at a point of time where I had multiple layers of people who just didn't talk to customers, and I was like, how can you run our customer facing organizations if we don't talk to customers? And so that's the kind of thing where it happens slowly and over time, and I learned to sort of assert more of what I expect out of senior leaders and kind of dictate the cultural standards when earlier, I think I was confused for, oh, this this must be just how you do things when you get big. Like, you have teams of people who do that for you, and I was totally wrong.

有没有那么一天，你突然意识到：好了，这事必须停了。今天我就要去找那些高管，跟他们说：我们得重新贴近客户，深入一线。这算不算一个关键时刻？

Was there, like, a day where you're like, you know what? This needs to stop today. Like, today, I'm gonna go to these senior leaders and be like, we're talking to customers again, boots on the ground. Like, do you was that a, like, a moment in time?

我也在想同样的事。

I was wondering the same thing.

哦，我肯定觉得有过很多次内心惊呼‘天哪’的时刻，觉得这情况不妙。但当你招聘高级主管时，你已经经历了漫长的搜寻、严谨的流程、全面的背调，通常在三十到六十天内就能判断他们是否合适。最糟糕的是，我觉得在这些情况下，我没有及时采取行动，甚至可能在欺骗自己：‘我只是需要多辅导他们一下，就能把他们带回到正轨。’这并不是说他们就是差劲的高管。我认为人们往往会复制过去公司里让他们成功的那些行为，而你需要判断的是：这些行为是否适合你的公司？

Oh, I definitely I think there were many moments where on the inside, was going, oh, shit. Like this is not good, but you when you hire senior leader, you've run these long searches, you've had a rigorous process, you've done all the references, and you kind of know in thirty to sixty days if they're gonna work, and I think the worst part is, you know, I I think in these situations, I didn't act fast enough, and I was probably lying to myself of, oh, I just need to, like, coach them, and I'll be able to steer them the right way and so forth. That's not to say they were bad executives or something. I think people replicate the behaviors that made them successful at their past company, and you have to be a good judge of, do you need those behaviors at your company?

没错，没错。

Right. Right.

所以我经历过很多次‘天哪’的时刻，可惜我没有足够快地采取行动。

So I had many oh shit moments. I unfortunately didn't do anything about them fast enough.

惯性力量太强大了。

Inertia's powerful.

是的。我认为这是公司发展到后期阶段的人们最该吸取的教训之一：人们往往不能及时应对问题。

Yeah. I think that that's one of the biggest lessons that people who are at sort of later stages in their company can learn, which is people don't act quickly enough to address problems.

顺便说一下，我觉得作为创始人和CEO，你们下面的人其实都知道情况如何。公司里的人都知道这事正在发生。

And by the way, like, I think as founder and CEOs, like, your like, people on the ground know what's up. People at the company know it's happening.

是的，必须这么假设。

Yes. Have to assume that

坏消息是，下面的人都已经看明白了，有时候你得亲自去发现真相，从他们那里获取信息。你很容易跟实际情况脱节。

the bad news is all people have figured it out down there, and, like, you need to go find the ground truth and get it from them sometimes. Like, you can get a little bit isolated from from that.

如果层级太多的话

If have too many layers of

层级太多。大家都知道这种情况正在发生，而你往往是最后一个才知道的人。

family layers. Like, everyone knows when that's happening. You're often the last one to find out.

所以你觉得Benchling现在算是扁平化组织吗？因为这可是个相当强烈的词。是的。

So do you think benchling right now, do you would you call it a flat org? Because that's a pretty dramatic word. Yeah.

我们肯定不是扁平化的，我们有层级，我有高级领导。我觉得更多是我更愿意相信自己的直觉，去判断我对高级领导的期望是什么，而不是轻易接受“以前都是这么做的”或者“我以前就是这样做的”。

We're we're definitely not flat. We have layers. I have senior leaders. I think it's much more that I'm I'm way more comfortable trusting my gut on what I expect a senior leader to do, and versus I'm I'm a lot less willing to accept just this is how it's been done before. This is how I've done it.

而且，有时候我会这么做，但我更愿意反复问五次为什么，以便真正理解。对我来说，有些事情是不可妥协的。比如，如果你要担任高级管理人员，你就必须与客户沟通。是的，这纯粹是我的处事方式，我也接受这一点。

And, like, sometimes I will, but I'm gonna be much more comfortable asking why about five times to, like, really understand. And there's some stuff that's nonnegotiables for me. Like, if you're gonna be a senior executive, like, you gotta talk to customers. Yeah. And, like, that's just my way of being, and I'm okay with that.

那你怎么判断他们是否真的在和客户沟通呢？

And how do you gauge if they're actually talking to customers?

你必须愿意以身作则，这很明显。我一直在和客户沟通。每次开团队会议时，我的每周汇报都会包含我从客户那里学到的东西，所以，我应该从我的高级领导那里了解他们与客户的互动，就像他们也应该从我的经验中学习一样。我认为，如果你自己都不愿践行某种行为，就很难要求别人去做。我完全能理解另一种类型的企业，比如拥有百万级客户或面向消费者的业务，一切都靠数据统计，你可能不需要亲自接触客户，但我们有1300个付费行业客户，其中一些客户规模非常大。

You have to be willing to model the behavior yourself, obviously. It's like I'm I'm talking to customers all the time. When I run my staff meeting, my my weekly updates are always involving things I learn from customers, so, like, I should be learning from my senior leaders about their customer interactions just like they should be learning from mine. So I I think it's hard to ask for a behavior you're not willing to model. And I could totally believe for another type of business where maybe you have, like, a million customers or a consumer business and everything's, like, statistics, like, I could totally believe you don't have to do that, but we have 1,300 paying industry customers, and some of them are, like, really big.

我知道签下合同的客户姓名，知道那些管理员的孩子叫什么。这非常具体，非常细致。

And, like, I know the names of the people who signed their orders on the contract, the admins who I know what their kids' names are. Like, it's, like, it's it's a it's very, very specific.

听起来很像Ronco公司的方式。是的，这来自Ronco的套路。听起来确实有效。显然很有效。

Sounds very Ronco to Yes. That's from the Ronco playbook. Yeah. I don't Sounds effective, though. Clearly effective.

我从不要求别人做任何我不会自己去做的事。

I don't ask them to do anything I wouldn't do.

所以，是的。也许这就是最好的启示之一：不要要求别人做你不愿意做的事。

So Yeah. And maybe that's that's one of the best takeaways is don't ask people to do anything you're not willing to do.

或者你不知道该如何做，因为你真的应该了解公司每个部分的运作方式。

Or that you don't know how to do, because you really should know how every part of your company runs.

我会尝试。是的，而且现在我也尽量不让别人告诉我不要做某些事。比如，有人会说：‘你不该做这个’，或者‘我们有专门的团队可以帮你做’。我认为，我不希望任何外部的人来改变我经营公司的方式。

I try. Yeah. And then I also try not I don't let them tell me not to do things now. Like, that's, again, a, oh, you shouldn't be doing that, or, like, we have people, teams who can do that for you. Like, I think it's that's a place where I don't want any sort of someone coming in from the outside to change how I wanna run the company.

所以，你认为这些年来，你是否逐渐更加信任自己的直觉，并且更有信心去反驳那些没有按照你期望方式做事的人？

So do you think over the years, you've just developed more confidence in your own intuition and confidence in pushing back if someone's not doing sort of something the way you want?

我觉得是的。不过我依然不是那种特别喜欢冲突的人。我认为，准备好面对冲突是一种后天习得的技能。实际上，公司最初聘用的几位高管都非常出色，帮助我们提升到了一个全新的层次。因此，我有点被宠坏了——当我后来组建下一任管理团队时，我并不习惯那些带着不同做事方式来的人，而这些方式可能并不适合推动我们前进。

I think so. I'm still not not you know, I'm not a super conflict happy person. I think it's like a learned skill to be, like, ready for conflict, and Yeah. Actually, the first couple executives who hired the company were, like, really great and helped kinda take us, you know, to a different different level. And so as a result, I was a little spoiled that when I got my sort of next management team at some point, like, I wasn't I wasn't used to sort of people who might come in with a different playbook than, you know, what would be required to to take us forward.

所以

So

太棒了！这真的非常有趣。非常感谢你与我们分享你的故事，我们期待有一天你能回来，做一期完整的深度访谈——一次Benchling深度解析。

Well, awesome. This was really interesting. Thank you so much for sharing your story with us, and we look forward to having you come back someday for full episode. A deep dive. A benchling deep dive.

Benchling深度解析。

Benchling deep dive.

谢谢。

Thank you.

再见。

Bye.