HADI cycle with examples: how to generate a hypothesis and make your life easier

ntile
5 min readApr 22, 2019

--

Hypotheses are only the pieces of scaffolding which are erected round a building during the course of construction, and which are taken away as soon as the edifice is completed.
Johann Wolfgang von Goethe

Even before the words CEO, startup, digital marketing, SpaceX, Tesla appeared, the great German writer had uttered the quote that gives a clear understanding of how modern companies work. Now not so many people continue doing what they are able to and benefit a lot from that. Now it is not enough to be skillful only. A customer makes their choice based on how you present your product, what level of service you provide, how favorably you compare with your competitors, etc. Each company is always trying to attract more and more customers, increase their loyalty, the number of their recommendations to other potential clients, improve customer satisfaction and gain a good reputation. To understand better how to make the customers happy, millions of hypotheses are generated every day.

What is HADI cycle?

HADI cycle is a practical way to use data. Acronym HADI formed of the core program steps’ names — Hypothesis — Action — Data — Insight. It’s scheme is quite simple: first you are posing and testing idea, then collecting data and drawing conclusions and finally making improvements which help to achieve the best results. Namely you are learning from your mistakes. Using of HADI method helps companies in organizing test results and gaining inner visions about their business activity.

What happens to hypotheses after generating them?

  • Many of them remain not implemented. The reasons for that are lack of resources (human, technical and material), inability to prove the value and appropriateness of your statement, and so on.
  • Hypotheses may do harm. If a hypothesis ends up being a failure but has taken too much time and many resources, it may even lead to the failure of the entire business.
  • Hypotheses come true and get successful. There are not so many of such hypotheses, but their main distinctive feature is: you increase a small-scaled implementation and get a profit because of that.

It may be concluded that ideas are a very powerful force, but it is essential for them to have a proper tool, which will filter them, monitor their implementation and signal about the consequences. Such a tool is a HADI cycle.

What does HADI consist of?

  1. H — Hypothesis. It always starts a cycle.
  2. A — Action. To accept or reject a hypothesis , you have to take actions. This section is necessary for description of actions and their implementation afterwards.
  3. D — Data. Then it is time to collect the data. In this section, there is a description of the parameters which have to change under the influence of actions, and ways of registering those changes.
  4. I — Insights. The last iteration of a cycle is Insights. Based on the gathered data we can judge if we have achieved our goals or not. Perhaps, the hypothesis is useful but needs some improvement. It is necessary to add this information to the relevant section and try to iterate the cycle.

Methodology or personal experience?

Some people, when first hearing about this method, are quick to say: “I’ve thought of that on my own” or “Everyone does so, but that doesn’t work”. Why may it not work? As it was mentioned before, the value of HADI cycle is not only in controlling how the hypotheses are implemented but also in filtering them. Therefore, when you come up with an idea, you should assess it according to two parameters. The first one is the team’s faith in the idea. We are going to measure it in percentage: from 1 to 100. The second parameter is the implementation complexity. It will be measured with a regular scale: from 1 to 5. After the assessment you should start thinking: is it worth implementing the idea which gets 4 out of 5, and the team’s rate of faith is only 30%?

Let’s consider one more type of ideas which would be better to postpone. For instance, we have launched advertising, we are attracting people to our site or service, but the number of registrations isn’t rising. And then one of the teammates has an idea: if to invert two blocks, it will be easier for a user to navigate on the site after registration. The idea is great, not hard to implement, and the team believes in it. But the trouble is: while implementing the hypothesis, we are not solving the main problem, which is to increase the number of users. Meanwhile, advertising is reducing our budget, and people are still getting stuck at the midpoint and cannot see our great service with perfectly convenient navigation. So don’t forget that even HADI is not a panacea from misplaced priorities. However, don’t let your ideas be gone. To avoid that, you can assign a type of priority to a hypothesis and implement the most relevant ones at a given moment. Not to forget about older hypotheses, you should keep them carefully and update their priority regularly according to the project development phase.

How long does one cycle take?

The main value of this method is a quick implementation of hypotheses. That’s why you shouldn’t spend more than one week and one sprint in order to check a hypothesis. Of course, there are exceptions to all rules, and we cannot forbid you to act differently. But it is worth keeping in mind that our purpose is not to implement the idea fully, but just see what it is capable of. After we have got convinced that the idea is viable, we can get scaled and spend more time and resources doing that.

There is one more aspect regarding the time. It is the number of hypotheses during one cycle. You shouldn’t hurry much and try to implement several ideas at once to impact the same parameters. It will completely destroy the experiment fairness and complicate the possibility of determining if a hypothesis is successful or not.

Want to read about tools for HADI cycle implementation and its examples?

Visit https://b.ntile.app/hadi-tools

--

--

No responses yet