Full transcript
0:00Margulan Kalievich , let me ask you a
0:01question . I’m nearly 40 , and I’ve
0:04decided to become a startup founder . I
0:07look at how startups were brought into
0:12Russia in the past , like VKontakte . It
0:17copied Facebook , right ? And Ozon ,
0:20Wildberries — that’s like Amazon
0:22coming here . So here is my question . I
0:27have always sold metal . I was in heavy
0:31metal , moving physical goods , doing
0:33things where the need is clear ;
0:36there’s an order from a company , we
0:38fulfill it . Right . But how do I
0:43transition to — well , I do program , I
0:46do a lot of web coding , I write a lot
0:50of things , but everything I create is
0:53like a car , you know , a drift car ? When
0:58you get into it , there’s this gear
1:01stick , a strange steering wheel , one
1:03seat , and just metal all around .
1:06Meaning , only I know how to drive it .
1:08If anyone else gets in , that’s it , it
1:10just doesn't work for them . Right . So
1:12how do I shift from this standard
1:15mindset , where you see a clear need , to
1:18a startup mindset , to start making
1:21money from ideas ?
1:23Well , you’re an IT guy , aren't you ?
1:25I’m not an IT guy , but I’ve always
1:28been connected to marketing ,
1:30development , and somehow through IT .
1:32I used to build websites too . My
1:36approach was always that when a new
1:38technology came along — for example ,
1:41when the first websites appeared — I
1:43sat down , learned HTML 20 years ago ,
1:46built a site , put it on the internet ,
1:48and I got a large number of clients .
1:52Uh-huh .
1:52Right . And every time something new
1:55appeared , I always grabbed it and
1:57mastered it on some intuitive level of
2:00my own .
2:01Well , look . How do you change your
2:06mindset and psychology ? I’ll start
2:08with the basics , okay ? The basis is
2:11always what is actually happening in
2:14the modern world . The first
2:16indisputable thing , you could call it
2:18an axiom , is thinking from first
2:20principles . The first thing that is
2:22indisputably happening is that the
2:23world is accelerating . Do you agree ?
2:25All processes are moving very fast .
2:27Nowadays , a week doesn't go by without
2:28an update . My Hermes agent had 157 new
2:32commits . I said , " It’s only been a
2:34week , how do you manage to fix things
2:36and add new features ? " So , the world is
2:39accelerating , all processes are
2:40accelerating . Do you agree ? Okay .
2:43Second . The world is becoming more
2:47complex , because if you take the same
2:49level of complexity but do it at speed ,
2:50it automatically becomes more
2:52complicated . Let's say , for example ,
2:55when you are driving at 30 km / h , you
2:57look at the trees , leaves , birds , the
2:59sky , and the clouds . You see those same
3:02clouds , those same birds , those same
3:04trees . Hit 100 km / h . That's it , it's
3:08hard to notice the birds , hard to
3:09notice the leaves , and you'll be lucky
3:11if you even notice the clouds . But if
3:13you go 150 , you won't have time for
3:15clouds . The road will consume all of
3:17your attention . Agree ? Uh-huh .
3:20And this is what’s happening ; the
3:22world is accelerating , and because
3:24it’s accelerating , it’s becoming
3:25more complex . The problem is that the
3:29complexity here isn’t just from the
3:31speed , but from the actual complication
3:33of processes . Right . Right . Moving on .
3:38If any process accelerates and becomes
3:40more complex , it increases the number
3:43of risks . Agree ? It’s one thing to
3:47drive at 20-30 km / h , and another to
3:49drive at 150-200 . With the same
3:52reaction speed and the same abilities ,
3:55your capacity to avoid risks drops
3:58significantly , right ? So , in this way ,
4:01the risks only continue to grow . Right .
4:05And what do we do intuitively in such
4:07cases ? That is precisely what
4:10distinguishes humans . We immediately
4:12slow down . Imagine you are speeding
4:15along and suddenly hit a patch of fog
4:16at night . Before that , the road was
4:18clear and bright , and then — fog . What
4:20do you do ? You abruptly reduce your
4:22speed . You switch to low beams and
4:26basically drive by feel . Agreed ? It’s
4:31the same when you arrive and walk into
4:33a swamp . In a swamp , you take a stick
4:36and start probing . Because just because
4:39there was a path here yesterday doesn't
4:41mean it's there today , since the swamp
4:43is constantly moving . Today the path
4:45might be two meters to the left , but
4:47you can only find that by feeling it
4:49out . And now , if we translate this to
4:52business , how do we do it ? The question
4:54isn't about changing your mindset or
4:56the technology of how to invent a new
4:58road or path . No , no . The question is
5:00about changing the entire understanding
5:02of processes . To do this , you start
5:04moving . I created a system , it's called
5:07the slug strategy . The slug strategy is
5:10based on the fact that Japanese
5:12scientists took a slime mold . A slime
5:14mold is a single-celled animal-plant
5:16creature . You know , it’s like , well ,
5:18a slime mold . And they put it into a
5:21maze . But on the other side of the maze
5:24, they placed a fragrant bait . And what
5:26did it start doing ? It started
5:27extending tentacles , really , outgrowths
5:29throughout the maze . And wherever there
5:32was a dead end , it stopped sending
5:34nutrients there , and its tentacle would
5:36retract . But where there was no dead
5:39end , nutrients were supplied , and that
5:42branch kept going . As a result , the
5:44slime mold unerringly found the exit
5:47from any maze , despite having no brain .
5:51It's a single-celled organism , it has
5:52no brain . Then they , uh , simulated this
5:55famous experiment on a map of Tokyo .
5:59And the slime mold drew the most
6:02optimal logistical routes . And when
6:05they placed it on a real map of Tokyo ,
6:06it turned out that there was no need to
6:08involve engineers to build railroads
6:10and highways . They could have just made
6:12a city map from the start , let the
6:14slime mold loose , and it would have
6:16created the optimal route . Just imagine
6:19, Harvard engineers did it , and here is
6:22a brainless creature . The result is
6:24exactly the same . So , when we face such
6:28uncertainty , the only way — look , in
6:32this uncertainty , what we can state for
6:37sure is that nobody knows the future .
6:42Right ? And , accordingly , in such a
6:45situation , you have two options . Either
6:48you make the decision yourself , but
6:50then you take on all the risks , or you
6:53make the decision based on data . Yes ,
6:56but how do you make decisions based on
6:59data if there is no data ? And that’s
7:02where I call it the " slime mold
7:04strategy , " where you formulate many
7:07hypotheses , and your task is to run
7:09many cheap , quickly verifiable
7:11hypotheses . And then you run 10
7:14hypotheses ; eight don't work out , two
7:17do . What do you do then ? At the end ,
7:19you use a traffic light system : red ,
7:21yellow , green . If the green hypothesis
7:24works out , you give it resources , money
7:26, attention , and work by the principle :
7:28strengthen the strong . The hypothesis
7:31that didn't work , you shut it down , and
7:34you keep moving like that . Therefore ,
7:37if you want to create an app that will
7:40succeed in current conditions , don't
7:43try to create an app that will succeed .
7:48You need to create a pipeline for
7:50testing hypotheses , and then , without
7:52overthinking it , strengthen what works .
7:56Because what is a hypothesis ? It’s
7:58like tentacles or sensors for
8:00collecting data from this world . The
8:03world itself will tell you what it
8:04needs . You take a hypothesis and say , "
8:08If I release this app , I believe that
8:10within 30 days , with such and such an
8:12ad budget , this many people will click ,
8:14this many will convert , and this many
8:17will leave a request . " That's it , you
8:21allocate the money , make a landing page
8:23, launch it , watch it , and check the
8:25metrics . Yes , are people interested ?
8:28Did it hook them ? Did it not work ? You
8:30say , " I believe that if I change the
8:32headline , the hypothesis will succeed . "
8:34You change the headline , run it , it
8:36didn't work . Next , you say , if I change
8:38the color , it will work . You change the
8:40color , it didn't work . And that is how
8:43you test .
8:44Understood . Thank you . And here’s the
8:46thing .
8:46But you know what’s good about this ?
8:49It’s definitely a foolproof way to
8:51move forward . But the problem is that
8:55it often contradicts our ego .
9:00Unfortunately , the main obstacle here
9:02will be our ego , because we will have
9:04to admit that we are
9:08at the level of a single-celled
9:10organism , a slug .
9:11Making a decision ,
9:12right ? But essentially , in other words ,
9:16if we speak metaphorically , right ? We
9:18just all need to learn how to talk to
9:21this world . Learn how to with this
9:23world . I am a believer . I say , we need
9:25to learn to hear what the Almighty
9:28wants to convey to us . This is feedback
9:30. It’s from Him . He says : " Don't mess
9:33around with this nonsense . Nobody needs
9:35it . But this right here , it will take
9:36off for you . Focus on it . " Mhm .