“Our success at Amazon is a function of how many experiments we do per year, per month, per week, per day.” – Jeff Bezos, CEO, Amazon
Google conducted 50 to 200 experiments simultaneously back in 2008.
Booking.com runs 25,000 tests a year.
Mark Zuckerberg confirmed Facebook conducts over a thousand experiments at any given time.
Experiments lead to innovations. However, most companies shy away from experimenting. Why? When you experiment, you can fail. Companies see failure as a wastage of time, money, and effort.
How can your company be different? What are the ways to inspire the employees to experiment? If you are asking the same question, this article is for you. In the next five minutes, learn all about how to nurture a culture of experimentation in your web development team.
Promote Curiosity
Author and Motherboard senior editor Briant Merchant shared how Apple invented the iPhone in his book The One Device: The Secret History of the iPhone. If you think it was Jobs all along, you are wrong.
Apple engineers began to work on the iPhone as an experimental project without Jobs’s knowledge. Afterward, his executive staff prodded at him to make it official. Jobs was reluctant in the beginning as he thought cellphones sucked. However, he trusted his team and told Scott Forstall to create the iPhone operating system (iOS). Need we say more?
Motivate your employees to experiment. Stand by them even when you do not know how the experiment will turn out. The result can be as revolutionary as an iPhone.
Allow employees to experiment even when you are skeptical
December 2017. The holiday season was around the corner. The Director of Design of Booking.com suggested having a new layout for the home page of the company website. The travel deals, options for hotels, and vacation rentals would be replaced with a small window. The window would feature three options — flights, rental cars, and accommodations.
The CEO of Booking.com, Gillian Tans; and the head of the core experimentation team, Lukas Vermeer opposed the idea. However, they still experimented with the idea. Why? Booking.com allows all its employees to experiment without management’s permission.
Offer your employees the freedom to start and stop any experiments. Encourage debates. Furthermore, employees must feel safe to reach out to colleagues if they need help. When you trust your employees, they will go the extra mile to innovate.
Data over opinions
- Rohn Johnson became the CEO of JC Penny. Afterward, he removed clearance racks and coupons and replaced them with branded pieces. Why? Because he wanted to. He never ran any test to validate his ideas.
After 17 months, sales dried up. The company suffered immense loss and Johnson was fired.
Test everything you can test. Prioritize data over opinions. Thus, if the test says you remove the pop-up from the home page of the website, remove it.
However, are there no exceptions? Yes. You can reject the data if you feel implementing the feedback will alienate an employee or hurt their feelings.
Furthermore, set up an experimentation program based on in-depth research and data. Question every opinion subjectively and validate insights with data. You will see the experimentation culture sprouting in no time.
Accept failures as a part of experimentation
“Failure and invention are inseparable twins.” – Jeff Bezos, CEO, Amazon
When Amazon released a new version of the game Air Patriots, the seven-day-user-retention rate dropped by 70%. Any other company would have fired the developers or took the game out of the market. But not Amazon.
The team discovered users left because the game’s difficulty level increased by 10%. Amazon tested four new versions and launched one of them. Users played 20% longer and revenue increased by 20%. A failure resulted in further success.
Use failures as a way to experiment further. When you as the CEO adapt to such a mindset, your employees will emulate. Thus, everyone will feel free to discuss ideas, share feedback, and work toward innovation.
You will stay ahead of 90% of your competitors irrespective of the outcome if you simply try.
Create environments that support experimentation
You asked your employees to experiment. Thus, you nurtured a culture of experimentation, right? Wrong.
In 2015, the IT department of IBM offered to run tests. The management agreed. However, the head of the experimentation team rejected proposals one after the other because the newcomers proposed those experiments. IBM ran 97 tests that year.
The next year, Ari Sheinkin was appointed as the head of the experimentation team. He allowed employees to run experiments around the world. He built a center of excellence to offer support, installed easy-to-use tools, and offered training for everyone. Furthermore, he offered free online tests to all business groups as well.
By 2018, IBM conducted 2,822 tests a year.
Telling your employees to experiment is not enough. You must create environments that support experimentation.
Learn to give and listen to feedback
Employees want to listen to feedback from their managers. However, you must give specific feedback. Here is how:
- “You must improve your performance.” — wrong.
- “You needed to share 3 ideas for fixing the screen-freeze bug. I haven’t heard from you yet. Do you need more time?” — right.
Furthermore, you can use the SBI (Situation, Behavior, Impact) framework as well.
- Situation: What is the issue? When does it need to be solved? In which area must the employee work?
- Behavior: Which behavior of the employee is causing or adding to the issue?
- Impact: How is the employee’s behavior impacting others?
Furthermore, learn to listen to feedback as well. If you as a CEO say, “Here is the problem. Here is what I think. But I am interested in everyone else’s opinion as well,” you made a mistake.
When you offer a solution beforehand, no employee will want to speak over it and share their ideas. Thus, follow what Simon Sinek said,
Learn to be the last to speak.
Your employees will feel valued. You will learn about new ideas and why the employees shared the ideas. It is a win-win situation for both.
Realize the importance of small changes
Not all changes need to be building an iPhone. Consider this example from Microsoft.
In 2008, Microsoft ran an experiment. Whenever a user clicked on a Hotmail link, the link would open on a new tab instead of the same tab. When Microsoft tested on 900,000 users in the UK, engagement increased by 8.7%. In 2010, the company tested the same on 2.7 million users in the U.S., the result was the same.
Thus, allow small changes. Test the changes on a small number of users. If you see positive results, implement the change. If not, analyze what is not working and you can end up creating something better.
Respect ethical boundaries
- Facebook ran an experiment where they rearranged the newsfeed. The experiment tested whether viewing less positive news made people post less positive posts. Furthermore, the experiment tested whether the reverse happened when people saw fewer negative news posts as well.
When the researchers published the results in an academic journal, Facebook faced public outrage. Users felt Facebook exploited and manipulated them.
Now, let’s consider LinkedIn’s policy. LinkedIn has an internal guideline that states the company won’t run experiments that would alter the user’s moods or emotions or deliver a negative experience.
Thus, think about whether the experiments you want to run are ethical. Ignore and you can face backlash from users. Imagine if your company can survive such an outrage.
If you are ever in doubt, remember this; do what LinkedIn did, not Facebook.
Set up guidelines
New employees can have doubts about how to run experiments. How to set up a test? Must they work with a specialized team or by themselves? Where do they pull the resources from?
Create demo videos or detailed documents to explain the process. Furthermore, create a FAQ section and an experimentation checklist to answer common questions.
You can create a dedicated team to answer these questions as well.
Celebrate and learn
Celebrate the success. Announce incentives or throw a party (even a pizza party counts) to motivate employees. Furthermore, when the experiments fail, learn from the failure. Analyze what the underlying causes were and it can steer your company to a new success.
