How to design real business hypotheses with analytical thinking in 5 steps.

We talk a lot about hypotheses daily, especially in the corporate world, and what I learned is that most of the things we talk about are just guesses, speculations, or assumptions, if we consider that the hypothesis is a scientific method.

In product management or life, a hypothesis is a scientific method for reducing uncertainty and possibly solving a problem.

Being a structured set of arguments and explanations that possibly justify data and information, but which have not yet been confirmed by observation or experimentation.

It is the positive, negative, or conditional statement (not yet tested) about a particular problem or phenomenon.

Every problem-solving process follows a cycle of Abstraction (to motivate) and Tangibility (to see if it worked). The whole process of solving a problem, be it a disease with medicine to the management of a product.

Strategy (Abstraction) — KPIs (Tangibility)
Opportunities (Abstraction) — Hypotheses (Tangibilization).

Step 1 — Problem Statement (Tangibilization)
Step 2 — Zoom out, Context, Scenario, Descriptive Analysis (Abstraction)
Step 3 — Zoom in, Diagnostic Analysis (Tangibilization)
Step 4 — Predictive Analytics, Forecasts (Abstraction)
Step 5 — Hypothesis, Prescriptive Analysis (Tangibilization)

Before going into detail about the five steps, you need to remember step ZERO, Be humble, assume you don’t know, and fall in love with the problem.

We can say that a problem is what distances you or even the gap between your point A (where you are) and your point B (where you want to go). It usually presents itself as a pain, a desire, or a need. A problem has no face.
To identify the problem you can use several ways, but bet on the simple “Toyota’s 5 Whys”.

Write the problem, clarify the ambiguities.

Example problem statement: My patient has a fever and muscle pain.

From the moment we have the problem statement, we start a process of zooming out, getting to know the territory, understanding the context, reading the scenario, anamnesis

“According to medicine, a good diagnosis is in a good anamnesis”.

A good anamnesis is composed of some essential elements, gathering everything you need to know about your patient (client) in one place, and keeping the focus on the main complaint (QP), family history, and previous pathologies.

  • Who is the target audience? Why does he buy it?
  • What’s happening? What data do we have for analysis?
  • Macroeconomics, Market share, number of customers.

We can also talk to the patient, investigate the history, have a verbal and non-verbal interpretation, cross-reference information, and understand the history of the current illness (HDA), systematic or symptomatic interrogation (IS), life habits, and psychosocial history (conditions socioeconomic and cultural).

With a good understanding of the problem context, we proceed to establish a diagnosis. In the diagnostic phase, we try to understand why this happened, and what are the correlations and causalities of the problem.

In this case, a cold, flu, dengue, migraine, or any other infectious disease in which symptoms can include fever and muscle pain could be diagnosed.
In product management, a tool used for diagnosis is the JOBS TO BE DONE.

At this time we analyze current and historical facts to make predictions about future events. What will happen if the problem is not resolved? What is the trend over the analyzed DATA?

What solutions can be implemented? What are the expected returns for the solutions presented? Or rather, describe the scientific method.
The hypotheses have characteristics such as:

  • To have uttered, to be a declarative sentence;
  • Have a relationship between two or more variables (parameters);
  • Be testable, and verifiable, through observation and/or experimentation processes.

And what is the hypothesis for the case?
(If the patient takes the medications and hydrates himself with 3 liters of water daily for 5 days, we expect the fever and muscle pain to disappear).

  • Affirmative — positive hypothesis (The search result must prove the statement);
  • Affirmative — negative hypothesis (The search result must prove the statement);
  • Conditional hypothesis (The search result is conditional on the results).

In results-oriented product management, when testing and validating a hypothesis, we need a few points :

  • A described solution;
  • Time factor;
  • Expected results with metrics.

We return to lesson ZERO: humility, because every hypothesis, even if confirmed, is associated with a time.

Much is said in the corporate world about hypotheses as an intellectual way to distill opinions and guesswork.

But just understand that the word hypothesis came to help us create a more data and experimentation-oriented organizational structure. Helping us to reduce uncertainties and reduce risks.


Read the full article here

Leave a Reply

Your email address will not be published.

How to evangelize UXR in your company. Run a UXR Roadshow.

How to evangelize UXR in your company. Run a UXR Roadshow.

You are the only UX researcher in your company?

How much money does a Senior Product Designer make?

How much money does a Senior Product Designer make?

Table of Contents Hide An answer backed by unsolicited offers in their inbox

You May Also Like