Embreo - Mobile Application Development Studio
Mobile Application Development Studio
01.png

Blog

How to design an experiment for idea testing

17_embreo_blog-03.png

Hi all, welcome back to part 3 of our ideation series. In this article we explore about how to design an experiment and testing ideas. 

To recap our previous articles, the result of testing your ideas and assumptions is the collection of the following,
1. Unbiased Data
2. Customer Insights
3. Actionable Steps

But how do you go about testing to receive them?
Before, we jump the gun and start tackling how we should design experiments, let’s first take a leaf from researchers and scientists when it comes to experimenting and testing. If we do not understand the thought processes, it will be difficult to know what outcome we should be looking out for from our experiments. 

Moving along to the philosophy of logic and reasoning…
Beware! This may be quite dry.
There are 2 main schools of thoughts namely, INDUCTION and DEDUCTION.

 

I N D U C T I V E    R E A S O N I N G 

17_embreo_blog-01.png

Inductive reasoning works from a “bottom up” approach. This means that we first begin with specific observations and data about a particular phenomenon, after attaining these observations, we start to look for patterns and regularities, from these patterns and regularities, we come up with a theory or hypothesis about said subject matter.

With that said, conclusions drawn from inductive reasoning while credible in nature are inherently uncertain and is subject to biases, as well as limited to current knowledge.

New information or knowledge of subject matter can invalidate induced conclusions. Take this statement for example,

Adam is a womanizer. Adam is a millionaire. Therefore, all millionaire’s are womanizers. (Which may or may not be true 😉 )

CASE STUDY : 
(Clients and proprietary information has been heavily modified to ensure the counter-parties involve are unidentifiable)

Adam is a client of ours and looks to provide a solution within the insurance industry as to what he perceives to be a problem. He believes that the following is true, that
1.  Insurees are being sold too much coverage. 
2.  Insurees are uneducated about the appropriate coverage for their current positions.
3.  There is not enough transparency within the insurance industry.

If he had used induction reasoning to come to this conclusion, his thought process would look something like this.

OBSERVATION :
1. According to the latest available data, between 2014 – 2015, total revenue for life insurance across all direct insurers have increased by an average of 17.7%.

2. Between that very same period, the total number of life policies (Business In Force) had only increased by 2.49%.

3. Sales of new life insurance had only increased by 0.54%.

4. Termination of existing life policies had increased by 10.37%

5. Claims of life insurers had decreased by 47.23%

6. Insurance premiums for all life policies of annual premiums, single premiums had increased across the board.

7. Sum insured across all policies have increased by 14.2%

From the above, which are all statistics that were calculated based off MAS’s annual insurance statistics that can be found below here.
http://www.mas.gov.sg/Statistics/Insurance-Statistics/Annual-Statistics.aspx

Adam noticed the following patterns and behaviors.

PATTERNS NOTICED :
1. Insurers are making more revenue and profit without substantially increasing sales.
2. Insurers are making more profit despite a lower amount of claims.
3. More people are terminating their insurance policies yet sum insured and premiums are increasing.

From those patterns and behaviors that Adam has noticed, he draws the following conclusions aka theories.

HYPOTHESIS / THEORY : 
1. Insurees are being sold too much coverage. 
2. Insurees are uneducated about the appropriate coverage for their current positions.

As such, he decides to pursue the following venture

SOLUTION :
Developing an app to help track insurees coverage to improve transparency within the industry

 

D E D U C T I V E     R E A S O N I N G  

Deductive reasoning approaches the topic from a “top down” approach. This means we first begin by creating a hypothesis or theory that we think is true. Thereafter, with the said hypothesis, we begin our testing and observations, consistently checking if the results of the tests or the observations confirm our hypothesis.

It should be a consistent cycle of which every assumption that has to be true in order for the hypothesis to be valid, be tested consistently for results that confirm the hypothesis. Failure of the observations to prove the hypothesis means that the hypothesis is not true.

CASE STUDY :

We shall use the same back story in order to showcase clearly the differences in processes. Adam is a client of ours and looks to provide a solution within the insurance industry as to what he perceives to be a problem. 

Firstly, he would decide that he wants to build an application that allows users to track and view their total insurance coverage in order to improve the transparency within the insurance industry.

SOLUTION :
Developing an app to help track insurees coverage to improve transparency within the industry  

He decides to build it because he hypothesizes that the following is true.

HYPOTHESIS / THEORY :
1.  Insurees are being sold too much coverage. 
2.  Insurees are uneducated about the appropriate coverage for their current positions.
3.  There is not enough transparency within the insurance industry.

In order for his proposed solution to maximize its chances of success, his hypothesis needs to be true. However, in many business scenarios, unlike science, subject matters tend to be subjective for everybody has different preferences and perceptions, which makes it impossible for there to be an absolute standard or definition.

In such a context, a business only cares whether there are paying customers towards a solution. Paying customers, that will pay enough for them to make a profit and that will be the primary objective of all these testing.

Hence, with that objective in mind, he can design experiments in the following ways to maximize his success,
1.  Define a success criteria for each hypothesis and test (Low – High Accuracy)
2.  Get sales (Highest accuracy)

The most obvious path towards validating his hypothesis and solution. While datasets and experiments can alleviate the uncertainty surrounding the solution and hypothesis of business ideas. Being able to sell and get sales is the most direct way of having your solutions validated.

For at the end of the day, as a business, your objective is to solve a problem that customers are willing to pay for. Having paying customers already fulfils the objective of a business and indicates the market demand.

Over here at Embreo, we believe that a combination of both inductive and deductive reasoning is required. We utilize both reasoning philosophies to strategize the most effective ways for developing startups and work with our clients in implementing these development strategies.

In our next article, we will be sharing more about our ideation systems and how we merge both inductive & deductive reasonings to help our clients strategize on their ideas.

Stay tuned, follow us on our Facebook page here or sign up for our newsletter.

If you’re interested to work with us and find out more about our processes, contact us here. 

leoda embreo