Grigory Petrov on sales at conferences
From the text you will learn
What do numerous conference appearances give you? What profit do you get from them for your company?
I am a full-time DevRel, or Director of Technical Marketing as it is commonly called. Evrone is essentially a two-way service company. On the one hand, we hire developers, take care of them and make sure they feel comfortable working with us – “outstaff with a human face.” On the other hand, we offer companies jobs as programmers and make sure that competent people work for them and the companies do not have headaches with management. That is, we care about the work of these companies.
And in order to organize such two-way care, we need to talk about ourselves. We need to show our technical expertise, we need to show that we really understand IT very well, we need to show that we have a high engineering culture.
I speak myself, and I prepare our speakers. That is, when my colleagues speak at professional technical conferences, this is strong social proof. That, firstly, we care about the developers, they find it interesting, pleasant, and convenient to work with us. Cool enough for them to prepare for speeches and submit papers at conferences. You can't force a person to speak. And secondly, we show potential clients that we are quite experts. Because weak reports will not systematically be included in the programs of the best conferences.
Why do I need this personally? Today Alexey Paevsky, the founder of the Neuronovosti portal about neurophysiology, a very cool scientific journalist, spoke in my scientific section. And you know, 20 years ago, when I was developing Radmin, Remote Administrator, I had a small problem. I trained developers to the level I needed in about three years, and they changed jobs in about a year and a half. My credit and debit did not match monstrously.
And after that, I thought for more than ten years: what can I do, how can I train developers better? Five years ago I started studying neurophysiology, realized that there were some answers there, and I now use my public speaking to test hypotheses. For example, in a month at CodeFest I will be giving my own talk, this is “Cognitive complexity of microservices.” I will tell you how the human brain understands the code, how we can cram a million lines into our heads, despite the fact that if you ask a person to remember 6-7 new names, then it will be difficult. But an experienced developer can understand a million lines of code. I will tell you how the mechanisms that provide this can work.
When I talk about programming, I constantly train: what formulations are best to use, how to convey complex things to conference guests, what research can help me with this. And you know, over the years it gets better and better, to the point that now I’m already writing a textbook on Python, where I directly integrate all my work. I will publish this textbook at the beginning of 2025, and with the help of it I will test on a large number of people whether I can teach “middle standards” faster than in 3 years.
Would you like to take part in testing the textbook? Write to Gregory in telegram @grigoryvp.
You have written many articles. Do they help with sales?
I can’t say that they directly help in sales; rather, they indirectly indicate our high technical expertise. I write many articles about what we do, about our projects, our stories and the like. No matter how cool you are, you understand inside that you are cool, you have a cool company, you do cool things, but until you tell people about it, no one knows about it. And it’s important to tell. It seems like the basics of management that it is important to praise people because telepathy does not work.
If you do cool things, it's important to talk about it, because telepathy doesn't work.
Do leads come more directly from conferences than from writing articles?
Oh, this is a very painful question: how to measure leads in general, especially in IT, especially when you offer some kind of service. Relatively speaking, I came to you from the after-party, where I talked with the developers, gave them chevrons with NFC tags, talked about us, talked about the conferences. I don’t have a direct goal of selling them anything. I tell them what benefits we bring, what perks the outstaff has: that we do management, conduct one-on-one sessions with our developers, do grading, train them, invite them to corporate events, organize parallel career growth, make sure they don’t burn out. This is convenient for businesses because they do not have to take over management.
But I'm not saying, “Hire developers from us.” I'm talking about the benefits. But for some companies such benefits are less interesting, and for some, interest may appear only in the future. And a person can write to me in six months, in a year, in two years. It's very difficult to track the correlation between what my colleagues and I do at conferences and the leads we generate. But we know that if we don’t tell people about us, then no one will come. So we do conferences in the hope that we're doing cool stuff and it's useful. But I don’t know how to measure this.
For example, I give out chevrons with NFC, I can count how many I gave out at each conference (Evron chevrons are the company’s branded merch, which is used instead of business cards – editor’s note)how many clicked on the link, how many wrote, but I will not be able to interpret these metrics, and I am not sure that they will show anything relevant.
It turns out that it is difficult to track this?
As one person said: “I know for sure that 80% of my PR budget goes nowhere, but I don’t know what 80%.”
As a generalist who has been writing code for a quarter of a century and has participated in various projects, I am asked: “Grisha, what programming language to choose, what stack to choose, what team is better to organize?” And there is a textbook answer that I think is good. What you need to choose is the stack in which you are an expert. If you have extensive experience in plus development, then you can even make a website using pluses faster than in Python if you retrain plus developers in Python. Or if you hire Pythonists or Rubyists without having expertise in this stack. And the same with PR, that is, do the PR that you know how to do.
We have been organizing Ruby Russia for 15 years, we are known in the community as a company that organizes a top conference on Ruby and regularly brings the author of the language, Yukihiro Matsumoto, to Russia. Many hundreds of offline rubyists come to the conference. And during Covid, we collected up to 2000 online registrations. This is what we get. As one of the leaders of the Russian Python developer community, I host podcasts and organize conferences. Every day I prepare up to four speakers, that’s up to a hundred speakers a month, that is, up to a thousand speakers a year for different conferences. We are also good at this. From all the variety of possible PR activities, we do those that work for us. And we believe that they benefit us.
You're an expert in several areas and continue to be an expert in Python, even though you haven't written production code for many years. How do you do it?
The fact is that every working day begins with me reading the news for a couple of hours. Using a huge RSS feed, I see what’s happening in the world of Python development, Ruby development, a little Go, front-end, neurophysiology. I'm learning all the new things that happen in the world of Python and Ruby. I write a little code to “touch it with my hands.” Then I spend 3-4 hours preparing speakers, and then I work for several more hours on some of my reports, research and other activities that I need for work.
So it's essentially a daily routine of studying the news, trying new things and keeping your finger on the pulse. Plus, if you look, 5 or 6 years ago at KnowledgeConf I read a report about a personal knowledge base. There I tell you how I have been creating my own personal knowledge base for 15 years. (Well, now it’s already 20, of course). I have this knowledge base, this is an extension for VSCode, it contains my own markup language, my own techniques for maintaining it. I write down quite a lot of things. If you look at my GitHub account, then you can see that it is all green with holes every few months. The day before yesterday, in my opinion, there were no commits, because it was a very difficult flight.
There is little code there, mostly just commits and updates to this knowledge base. I write down everything important, constantly re-read what I have already written down and improve the structure of the knowledge base. Almost like Zettelkasten: I, practically without knowing him, came up with something similar and over the course of 20 years polished it to suit my Devrel needs.
The answer to your question: I read the news 2 hours a day, try everything, write everything down.
You mentioned neurophysiology several times. Name a few cases when you successfully used it in your work
Well, first of all, the speakers. This is probably the most successful application of neuroscience in my work. I still don't know how to properly train programmers, but I've gotten pretty good at teaching speakers. This is aided by theories about how the mind is supposed to work, how attentional dynamics work, and how people are aware of the presentation and slides. I put the training of speakers at IT conferences on track. For example, an architect or technical director of a company comes to me who wants to speak at HighLoad, PyCon or Ruby Russia. And he says: “Grish, I have such and such ideas, I want to perform, what should I do?” I tell him how people understand the world around them, how they understand a report, why slides are needed. I explain the principles of how he or she should approach preparing a report, how to make slides, how to practice. Again, I developed the training methodology myself based on what I was able to read and understand about the dynamics of attention and the formation of long-term memory. The technique contains several interesting points. Well, basically this is, of course, spaced repetition system, but slightly modified specifically for reports. The reports of some of my students become the best reports of the conference. Plus, as conference organizers, we look at ratings. Which are growing. This is the first big application of neuroscience, what we're doing right now.
The second is my Python tutorial that I'm writing right now. I have a group of early adopters, about 100 people in telegram. Every week I publish a new piece of the textbook for them. They read and share their impressions. And every Friday we call each other and discuss what we’ve written via video call. The textbook is not ready yet, so I cannot say that this is a successful use of neurophysiology in practice. The bulk of the material, of course, has already been written. Now I am rewriting the chapters over and over again, and at the beginning of 2025 I will publish the resulting Talmud in paper form. And then it will be clear how this textbook competes with others. Will those who study it become slightly better programmers? And how long will it take them?
What do you think is better for learning new material – texts or videos? For example, there are many “Python in 7 hours” courses on YouTube. How effective can they be? Or is text better??
This is an interesting question. Look, programming is complexly structured information, of which there is quite a lot. All our lives we learn to consume complexly structured information in the form of text. We are used to reading textbooks, we are used to reading text, we are used to understanding complex things from the text, because the text can be re-read. When we comprehend a text, we have a visual field across which our eyes can jump. We are very well accustomed to understanding complex things through text.
There is no evidence that any complex skill can be learned any better using visual or vocal information than from text. People learn in the same way – both “visual”, “auditory”, and “kinesthetic”. At the same time, please note that people may have values, beliefs and emotional preferences. That is, a person may prefer to learn using audiobooks. That is, such a person will learn worse by reading audiobooks than if he reads a text textbook. But at the same time, he or she will not read a text textbook, because he or she doesn’t like it, but will read audiobooks because he or she likes it.
For example, if we take a person who identifies as a visual learner and only learns from comics. And we will force such a person to study from comics and a textbook. Then we will see that he or she learns better from the textbook. But since he likes to learn from comics, he himself will never learn from a textbook. So here it is very important to take into account that in addition to the fact that all people learn the same way, people may have preferences, values, and beliefs. A person may think that he is an auditory learner – and only audiobooks.
Let's return to the topic of sales. Should people who sell understand how the development process works, be immersed in it and communicate with the client in the same language?
How do you get a programmer to be a salesperson? I think that in an ideal world, it would probably be cool that the sales person who sells outstaff services is also a programmer himself and is well versed in this. But this is something out of fantasy, it doesn’t happen. Therefore, a salesperson must have good interviewing skills, good inquisitorial skills, in order to properly interrogate developers, clients, ask the right questions, interpret the result correctly, conduct the assessment correctly.
Moreover. The fact that I am a strong developer myself often hinders me. Because my expertise is not endless. If Pythonists come to me for an assessment of backend, then I can know their language, stack, and architecture better. And I can very well evaluate the project, the timing, and the risks. But if ML engineers come, then, surprise, I don’t understand Machine Learning at all. Despite the fact that he is a neurophysiologist, because “wet” neural networks differ fundamentally from artificial neural networks. And they come to me and say: “And here we highlight the signs, and here we have embedding, and here we have inferences…” For me, this is some kind of elvish, I don’t understand what they are talking about. Unlike our manager, who is not a programmer, but is equally good at interrogating any developers.
I see it’s the same with sales people. In my opinion, a person who helps companies and developers find each other must have good communication and organizational skills. Moreover, he or she does not have to be a programmer at all.
I am convinced that communication and organizational skills are much more important. Good programmers have little interest in becoming sales people; this is a non-scalable story. How can you teach a salesperson programming? Will he or she have two jobs? Spend eight hours selling, and then spend a few more hours learning to program? You need to invest years to even learn programming at all. Sort of like that.
PR, marketing, accounting, sales. What link can be sacrificed in the process of selling outstaff?
If you remove PR, then no one will know about you except word of mouth. It is theoretically possible to live on word of mouth. If you remove marketing, you will not be able to evaluate your services, you will not be able to understand who your client is, you will not be able to position yourself in the market. As a result, the business will most likely not work. Accounting is generally our pride, our management is practically what we charge for our small premium. Because we care about developers, because we organize processes.
Accounting is an integral part of outstaffing. And sales… Well, if you remove sales, then, in fact, how will you get the client to account? At the same time, an interesting question is what to call “sales”. If the client has already arrived, and you have fixed prices, and you can land the client directly in the account, then this can work. I think, depending on the company, you can donate either PR, then you have word of mouth, or sales, then you should have something like McDonald's.
So that the potential customer to whom the PR comes does not need a sales person to assess costs, needs and all this. So here are two options. Either we throw away PR and remain on word of mouth, or we throw away sales and turn into McDonald's.
Describe your ideal sales team, including presales. Who should be included?
The ideal sales team? Well, I think the ideal sales team includes one sales person who has a background as a developer. But, as we have already said, this does not happen.
The ideal team that can actually be assembled is, I think, a salesperson who knows how to communicate with the business about payroll, about processes, about framework agreements, about interaction with outstaff. He should be helped, firstly, by a technical presale, because it is the technical presale that will answer the customer’s technical questions. And he should be supported by an account that will provide direct access to developers.
For all this, we have an Evrone career, our internal ERP, a spacecraft control mechanism that we have been making for more than 10 years. There, sales, for example, can “book” a developer for a potential client, but when the client has already agreed, all this should be transferred to the account, and the account should already take specific actions so that the guys and girls start working for someone.
Release of the podcast “Something in Programming” with Grigory Petrov watch on Doubletapp YouTube channelaudio version listen on a convenient platform.
Watch other episodes And subscribe!
Photo: Doubletapp, Evrone, {SPEACH! (ex.IT-People)
