<img height="1" width="1" style="display:none" src="https://q.quora.com/_/ad/1103c3a45d2e46399d3d99b48950095f/pixel?tag=ViewContent&amp;noscript=1">

Everything You Need to Know About Conducting Lean Startup Experiments

Conducting Lean Startup Experiments, planning experiment, startups, founder, tech startup.

Megan McLatchie

Marketing Executive

Updated 04 June 20

Share:

  

(3 Minute Read)

 

When building a Lean Startup, the more you can learn about your customer the better and learning through your own hypothesis-driven primary research will provide you with the most important things you need to know about your startup business; exactly what your customers want and whether your startup is meeting their needs.

 

In the words of Paul Graham, “How do you figure out what customers want? Watch them.”

 

Getting started

The objective when gathering primary research is to determine something called an ‘insight’ or a ‘breakthrough’ that proves our initial hypothesis to be either true or false. Quite simply, an insight is a new piece of learning that you previously did not have gained through conducting an experiment. 

 

If we assume what our customer's needs are, this makes the risk of failure very high as we don't know for certain whether or not what we're building is right for them. Conducting primary research to gain unbiased insights from users, helps to mitigate the risk of building something that nobody wants and instead build products people love.

 

To do this, first we create an assumption, then predict an outcome, conduct an experiment, and finally compare findings.

 

Create an assumption

Otherwise known as a hypothesis, we make an assumption about the current behaviour of our customer. So let's use a busy Veterinary Clinic as an example. An assumption could be...

 

Assumption: Pet owners rush to a vet consultation whenever they detect their furry friends might have an illness or injury. Vets believe the majority of these pets did not need medical assistance. 

 

Predict an outcome

Next, you need to predict the ‘consequence’ from the hypothesis. Think about what the knock-on effect will be as a result of the customer’s behaviour. For a busy Vet this might look something like this...

 

Consequences (The result): Vets do not have the capacity to take on more appointments resulting in long wait times in practises.

 

Conduct an experiment

We use two experimentation tools:

- Problem Interviews

- Problem Surveys

 

The goal is to gain an accurate and deep understanding of how our customers actually behave and what the problems/impacts this has on them and why. The point is that we go and speak to your users directly to see whether our hypothesis and predicted results are correct or incorrect. So for our Vet example, an experiment might be...

 

Experiment: We surveyed 200 Pet Owners & 80 Vets - 80% immediately call a Vet to book an appointment and 77% of Vets think their consults were a waste of time and not necessary.

 

Compare findings

As you can see, we discovered that our hypothesis was true and gained an insight into the problems of pet owners and Vets and their behaviours! (mini fan fair) From here, we understand that there is a problem - the next step is to figure out why this is happening? Using the insight you just gained, you can go on to build another experiment for further testing and validation. If your hypothesis and prediction were incorrect, you’re in a position to make a validated judgement on whether to build, tweak or pivot your startup.

 

 

The difference between the struggling and the most successful startup businesses is that the successful startup teams make finding and using feedback a habit. If you have any questions about conducting experiments for your startup, put them in the comments below and I’ll be happy to answer them.

 

 

Have you got an idea for a tech startup? Apply to our mentorship programme today.

Share:

  

Comments

Sign up to our free mentorship programme!