Deep analytics

Deep analytics will require close contact, you will have to coordinate every little thing. Up to the behavior of the application when scrolling the screen: whether the content itself will be immersed or you will need to click on the page numbers.

After working out all the details, Dima will be told the exact amount of the project up to a penny, they will tell the exact time when the developers will send a link to a working service. They will give the TK, with which Dima will be able to contact other developers — you can compare options or hold a tender.

Long and expensive, but it will be clear exactly how much money you need to cook. A good option if the customer wants to know the budget and the deadline to a penny, and is willing to pay dearly for the preparation. Dima was not satisfied with this — for him to wait for a conditional three months only TK is too long, I don’t want to pay hundreds of thousands for it.

Finally, Dima got to us. We asked him a little about the project and suggested a golden mean — to conduct a quick and shallow analysis, as we call it, the Discovery phase.

It’s like deep analytics, only not so detailed, long and expensive. For example, at Discovery, we will identify key business processes, solve complex issues like the logic of receiving cargo or forming an order, and determine the list of services with which to integrate.

If we see services whose integration is difficult, long and expensive to evaluate, then we will say so. For example, with Dima it turned out like this: 80% of the project volume was estimated for the exact amount and time. Integrations with services like tracking cars by GPS were offered to be done by the hour, with an approximate forecast on time and a fork in the price.

The result is a mixed style: in a couple of weeks, Dima received a TOR in which the bulk of the work is estimated, and for the remainder of the project there is a fork in the cost, which he calmly puts into his plans.

Which work option to choose

I don’t want to say that quick analytics with the help of Discovery is the coolest way to prepare TK. Although, judging by experience, it is more suitable for our customers. If you look objectively, then you need to choose a work model depending on the scope of the project and the comfort of the customer. In short, so:

Without TK — when speed and flexibility come first. It is necessary to start as soon as possible, and the plans and vision of the project may change along the way. The customer bears the risks of cost growth and is ready to be involved in the project. Suitable for startups.

Deep analytics — the study and evaluation of the project can take several months, it is expensive, but then everything is accurate, the budget can be approved and there is no doubt that it will change. A good option for large corporate or government projects.

Quick analytics, “Discovery” — the study of the project is detailed enough to draw up a technical specification, but without scrupulous elaboration of all the subtleties. It costs less, takes a couple of weeks. Involvement is also needed, but not as often as in work without TK. There are risks on the budget, but minimal, the boundaries are usually prescribed in advance. Suitable for atypical projects where it is important for the customer to start as soon as possible, but with an understanding of the budget.

Dima for the logistics project, we offered to hold a “Discovery” — he has an unusual and great service, he wanted to start as soon as possible, but the development budget was limited.

Discovery is a flexible method. For Dima, we made an assessment of the project as a whole, and we decided to integrate with the payment system later, with hourly payment.

This saved time and money — to accurately assess the integration, you would have to pick through a lot of bank documentation. Moreover, such integrations are additional features, because the project could have started without them.

“Not every project will need deep detail. The customer needs an absolutely accurate budget — OK, let it be a long and expensive analytics, but then no changes. But in most cases, this is not necessary — because with the help of Discovery, you can fairly accurately estimate most of the project and save on analytics and startup speed”

Discovery. Stage 1. Business goals

I’ll tell you how we conduct Discovery. I will try to describe it so that you can try to repeat the process yourself and see the benefits of this method.

First, we sit down with the customer and ask him to tell us in general about the future project. We are trying to find the key problem that the service will solve and the goal that the customer pursues.

Another customer, Lyosha, came to us with an urgent request to solve problems with documents. He has a brokerage business in which managers accepted and sent clients a bunch of different documents by mail, in several letters. The problem is that it took a very long time, managers sometimes got confused in papers.

The solution is to make a service that will accept documents from different sources and validate. If something is wrong, it will be sent back for revision, if everything is fine, it will be transferred to the manager. All actions are recorded in the cloud, no one gets confused and nothing is lost.