Why not everyone is capable of making products?
The ongoing battle between adepts at product development and outsourced employees is not likely to slow down, in my opinion. Logical arguments are becoming increasingly rare, while the disputes between two groups of IT professionals (different in size, but equally important) are becoming more reminiscent of the controversy between Apple and Samsung fans. Probably nothing is wrong with that. But instead of confessing to love one of the approaches, I will try to stick to the facts.
At one point in my life I was a co-founder of a product company, and I’ve made many mistakes. Then I worked on outsourcing sites, in fact I constantly switched between the two types of companies in search of a golden mean between the approaches. Now I have been observing the integrated orientation in DataArt for almost a year, leading the system’s transition to the product approach in development and I would like to tell you about it.
I wrote an article along the lines of my speech at IT Talk in Kiev, which is a series of internal DataArt conferences, to which we often invite anyone who is interested. It was practically impossible to add all the facts to the report, which confirm its main theses. Nevertheless, I hope you find this article convincing enough.
Dreams about products
In many countries the development community is formed by outsourcing. A soldier who does not dream of becoming a general is a bad one. This also applies to an IT professional who doesn’t want to create their own product and become a billionaire. Or at least a millionaire. This is a healthy desire.
Why does everyone dream about making their products? My favorite explanation for this is the survivorship bias. Stories are written by successful people who are often just lucky. They do not want to admit this, and they only see their hard work as success. The product gives us large scale and a quick profit. Therefore, everyone wants to create an analogue of e.g. Snapchat. By the way, if you identify how many readers of this article use this service, I think the percentage will not be that great. This is useful information while considering the needs of adolescents and the chances of creating something worthwhile for them.
Outsourcing is a service business that always scales only in size. Growth is needed here to process ten times as many customers. Sometimes for this you need 10 times more people, because the scaling here is non-linear. However, it wouldn´t matter much for the developer.
There are two main motivators that make working in a product company more attractive: money and fame. A programmer would much rather say that he is a Facebook employee, than to explain how cool it is to work in an outsourced project. Service companies alone are rarely famous, but with a successful product it is quite possible to share success. And it wouldn’t matter how interesting or boring the work of the developer who participated in its creation will be.
In service companies, the sense of belonging and ownership of the results of work is generally much weaker. Internal start-ups and business incubators, on the other hand, consider the sense of ownership as essential, since the dream of developing their product is what drives them.
Ready to roll
If we want to transition a service company to a product approach, we must understand that this is a new business for us. Now, other motivators will work for the team, and the requirements for understanding the business for engineers will also change. Everyone must realize how the product created by them earns the client money, from the company management and project managers to developers and QA.
Imagine approaching any business owner. Let’s say a mobile coffee business owner. Ask them how much they earn, how many clients they serve daily, what is their average profit, and what are the operating and unexpected risk costs. The owner probably has answers to all of these questions. And when they decide to get a few more coffee trucks, the owner will train and expect the workers to understand all these areas of the business. In fact, they should be taken into account by everyone who works on the product - decisions on the future of the product is made based on such indicators.
Relationships between engineers and the business are built in different ways. Outsourcing companies usually lay many layers of managers between them, protecting developers from having to think about business aspects. The product approach puts them at the forefront of communication with software customers. At the other extreme, far away from the traditional service model, there is Booking where there are no testers nor managers who decide to "release or not."
Here you can deliver a new feature in one move. It's not that they do not have people capable of handling everything, just the code first goes into the rotation of AB tests with millions of options. The main parameter is the order conversion rate. If the number of orders with new features increases, the robot starts to increase traffic. If it decreases - the robot removes it from the rotation, and the author receives an email: "Thank you, your idea has failed." Thus business indicators become a measure of decision-making. You know that inaction often proves to be a better strategy than writing unnecessary features.
Criteria for success
Now imagine that you are the owner or a CEO of the company with those same coffee machines. For some reason, you decided to outsource the development of an IT system for your coffee houses. What information do you need to convey to the executors so that the business remains effective? Business vision and target audience: who buys coffee from you and what are the places with the highest traffic. And also indicators that signal that the product needs to be adjusted or changed. In this case, such an indicator will be the average number of purchases per day. Experiments can be conducted rather quickly, i.e. spend all day on the spot and take into account the weather conditions. It seems that one doesn’t have to be Einstein or Copernicus to be able to do it.
In order for the development of this product for coffee houses to be successful, we need to answer two key questions:
- Who and how can determine whether the project is successful or not?
- Who will ask if the project is successful?
If the service company often does not have people who can answer these two most important questions, then there is no product approach in it. For the sake of justice, product companies don’t always have the answers either. But it is wrong. And in such cases, THAT type of project manager, beloved by all, appears on the stage. You know, the one who likes to calculate not completely understandable indicators, i.e. compares the price of one "Tavria" with monthly electricity bills, etc. It is not very difficult, and with certain efforts it allows you to look smarter than the others for quite some time.
Coffee machines and coffee "on wheels" is an operational business that operates under understandable schemes. In IT we are engaged in the creation of innovative business products. At least, you and I are pleased to think that way each time we produce something new. But to convey the exact idea of business indicators to us is no less, and maybe even more important. While in most companies, in my opinion, this function is not covered by anyone.
The problem is aggravated by the fact that most of the products that we create have a very long feedback loop. With coffee, everything is more or less “right here and now”. In large online games, the feedback loop is already significantly longer. Users do not leave immediately if they become bored. They begin to gently fade out, and the fall is first expressed in fractions of a percent. And you need to understand whether this is a seasonal phenomenon related to external factors, or internal. And you can fix everything in a month and a half at best, or even twice that much.
Outsourcing companies perfectly follow the technical part, but they often do not know how to monitor the business. At the same time, the majority of clients who turn to them do not have a direct relationship to IT. That is, in fact, they also do not really know how to monitor their own high-tech product from the business point of view. When it all comes together, then there’s trouble ahead.
Dive into the business
There’s nothing wrong with not making products. Outsourcing is a pure and honest occupation. Another thing is that, only looking at your project from the position of product management, you can understand that the optimal solution can lie outside the software development. Moreover, such solutions are highly valued in product management.
From my own experience I know this much: when you create software, you most appreciate architecture, reliability, functionality, etc. When you create a product, you begin to appreciate time most of all. As a rule, a project manager considers the work completed, when the tasks are closed, and the required quality indicators are achieved. For the product manager, the quality criteria are reduced to business indicators. Once you’re in their place, you understand why the MBA program includes finance or marketing.
The transition to product development looks attractive primarily because it opens the way to the global market. If you offer a ready IT product, the buyer does not care where you created it in Belgium, Kenya or China. If your offer is a service, there are a number of restrictions: language barrier, time zones, etc.
Outsourcers have a long time to prepare a good technological base. But this reserve proves to be not just a plus, since the companies have trained engineers, but a minus as well, because they are accustomed to solving all problems through the production of software. And the bigger your team is, the more difficult it will be to change the approach that is customary for the colleagues.
Making products without understanding business is impossible. And if you are aiming at the production, you will have to teach this to all those whom you plan to use in the product development. To ensure that the common way of doing things does not put too much pressure on those who will have to radically change, accelerators are usually advised to be in separate offices. The physical separation from the traditional company here can really help.