Insolence and enthusiasm for rapid growth from Aviasales
Aviasales Product Director Anton Baizur at the Epic Growth Conference , which was organized by Mobio , Getloyal and Appsflyer with the support of myTarget , Appnext .
In his report, Anton tells how the processes within the company have changed and adapted so that the team can maintain the dynamics of development and growth of Aviasales.
Full text transcript can be viewed at the link . Or read the short notes below.
In 2017, Aviasales increased by 60%. The service helps partners sell tickets for $ 1 billion per year. For 10 years of the company's existence, its team has faced a huge number of various problems associated with growth. Often wrong. Anton told about the culture, daring food and marketing experiments, which gave a multiple growth to the first meta-search of airline tickets in Russia.
')
Since Aviasales is a technology company, the main thing for it is people. The process of hiring people has not changed during the entire existence of the company. 90% of the people who work at Aviasales were hired after the first interview.
Young professionals want to go to Europe, to the valley. Southeast Asia does not quite look like a valley, I would even say, it is not a valley at all. In Aviasales, we began to experience some difficulty finding people. At the same time, we are still looking for ourselves, almost never used consultants for hiring. And it seems that the only person the company has found with the help of HR consultants is me.
We are faced with the problem of internal vertical growth within the company. Unfortunately, they could not solve it completely. We are good at it with technical experts, but unfortunately, it is impossible with product managers. There are very few professionals in grocery marketing on the market.
Decision making pyramid
How did the processes within the company change? We call it the decision pyramid. Despite the fact that we are a startup, we are still a company that has attracted investment. We have shareholders, a board of directors, we have to live within certain rules.
Moreover, as a transactional company, we receive money from each ticket that is found and sold through our service. Therefore, it is quite simple for us to set goals for ourselves for several years ahead. The strategy for 2-3 years, as the top of the pyramid, is measured in simple terms - these are tickets and money.
Inside the product, we naturally run everything through A / B tests. But this is not a story about the fact that we, like everyone else, are testing through A / B. I want to share a number of rules that guide us in A / B testing.
We never test technological changes and UX changes at the same time. Once done so, when they launched a new interface and technology in the search issue of tickets. Launched in A / B and lost 40%. Engaged in this 4 months. Received the result of defeat in 40%. In the end, after this, we suffered for four more months and made them catch up with each other.
Now we are trying to carry out all the experiments in the Cheap & dirty mode.
In Aviasales for two years there is no technical director. What did it give us? First, we have vertical growth among developers within teams. We have the role of Dev-head within each team. This is not a story about a team leader. We never had tmlidov, developers have always been at the same vertical level.
Marketing
Aviasales know our jet, wild, crazy marketing in social networks or email newsletters.
First, it doesn't always work out well in marketing. This does not mean that we did not get into someone's understanding of the sense of humor. Among the million events, jokes, reactions in social networks, some just go unnoticed. There is no recipe by which we make successful cases in SMM or mailings.
We constantly conduct some experiments, and these experiments are even expensive, in which we, quickly making a decision, invest money and see the result.
Analytics is a failure that you shouldn’t talk about at all.
What to do with this huge amount of data that does not fit? It is impossible to turn to him, wherever you put him, he costs cosmic money. We constantly wondered why we needed these indicators at all. They suffered terribly, and objectively few in the market have such a super-efficient analytics team that earns money based on this. Yes, there are options for how to use the data: for advertising, case studies, retargeting, and so on.
But we could not figure out how to get significant benefits from analytics in finding tickets. As a result, during the hackathon it happened by chance. For the sake of experiment, the guys decided to take a huge array of data and try to predict the prices of air tickets.
In 48 hours, they wrote an elementary model that predicted prices for air tickets in the direction of “Moscow - Phuket” with an accuracy that was much better than the cache we had. We drove it according to previous data, we looked - it worked out.
Subscribe to the Epic Growth Channel in the Telegram, and be the first to know about top content in product marketing.
In his report, Anton tells how the processes within the company have changed and adapted so that the team can maintain the dynamics of development and growth of Aviasales.
Full text transcript can be viewed at the link . Or read the short notes below.
In 2017, Aviasales increased by 60%. The service helps partners sell tickets for $ 1 billion per year. For 10 years of the company's existence, its team has faced a huge number of various problems associated with growth. Often wrong. Anton told about the culture, daring food and marketing experiments, which gave a multiple growth to the first meta-search of airline tickets in Russia.
')
People
Since Aviasales is a technology company, the main thing for it is people. The process of hiring people has not changed during the entire existence of the company. 90% of the people who work at Aviasales were hired after the first interview.
Young professionals want to go to Europe, to the valley. Southeast Asia does not quite look like a valley, I would even say, it is not a valley at all. In Aviasales, we began to experience some difficulty finding people. At the same time, we are still looking for ourselves, almost never used consultants for hiring. And it seems that the only person the company has found with the help of HR consultants is me.
We are faced with the problem of internal vertical growth within the company. Unfortunately, they could not solve it completely. We are good at it with technical experts, but unfortunately, it is impossible with product managers. There are very few professionals in grocery marketing on the market.
Decision making pyramid
How did the processes within the company change? We call it the decision pyramid. Despite the fact that we are a startup, we are still a company that has attracted investment. We have shareholders, a board of directors, we have to live within certain rules.
Moreover, as a transactional company, we receive money from each ticket that is found and sold through our service. Therefore, it is quite simple for us to set goals for ourselves for several years ahead. The strategy for 2-3 years, as the top of the pyramid, is measured in simple terms - these are tickets and money.
Product
Inside the product, we naturally run everything through A / B tests. But this is not a story about the fact that we, like everyone else, are testing through A / B. I want to share a number of rules that guide us in A / B testing.
We never test technological changes and UX changes at the same time. Once done so, when they launched a new interface and technology in the search issue of tickets. Launched in A / B and lost 40%. Engaged in this 4 months. Received the result of defeat in 40%. In the end, after this, we suffered for four more months and made them catch up with each other.
Now we are trying to carry out all the experiments in the Cheap & dirty mode.
Development process
In Aviasales for two years there is no technical director. What did it give us? First, we have vertical growth among developers within teams. We have the role of Dev-head within each team. This is not a story about a team leader. We never had tmlidov, developers have always been at the same vertical level.
Marketing
Aviasales know our jet, wild, crazy marketing in social networks or email newsletters.
First, it doesn't always work out well in marketing. This does not mean that we did not get into someone's understanding of the sense of humor. Among the million events, jokes, reactions in social networks, some just go unnoticed. There is no recipe by which we make successful cases in SMM or mailings.
We constantly conduct some experiments, and these experiments are even expensive, in which we, quickly making a decision, invest money and see the result.
Analytics
Analytics is a failure that you shouldn’t talk about at all.
What to do with this huge amount of data that does not fit? It is impossible to turn to him, wherever you put him, he costs cosmic money. We constantly wondered why we needed these indicators at all. They suffered terribly, and objectively few in the market have such a super-efficient analytics team that earns money based on this. Yes, there are options for how to use the data: for advertising, case studies, retargeting, and so on.
But we could not figure out how to get significant benefits from analytics in finding tickets. As a result, during the hackathon it happened by chance. For the sake of experiment, the guys decided to take a huge array of data and try to predict the prices of air tickets.
In 48 hours, they wrote an elementary model that predicted prices for air tickets in the direction of “Moscow - Phuket” with an accuracy that was much better than the cache we had. We drove it according to previous data, we looked - it worked out.
Subscribe to the Epic Growth Channel in the Telegram, and be the first to know about top content in product marketing.
Source: https://habr.com/ru/post/354270/
All Articles