Minimum Viable Data
This article is written by Milambo Kabeleka, a Contributor Author at Startup Turkey.
Alex Salkever is a writer, futurist and technology leader. He is the co-author with Vivek Wadhwa of “The Driver in the Driverless Car: How Our Technology Choices Can Change the Future”. Alex has served as a senior executive at a host of Silicon Valley startups in green technology, data science, open government, and cloud computing. He has been a senior leader at respected brands in technology, most recently at Mozilla where he served as a vice president. He was present at Startup Turkey Demo day, where he talked about getting minimum viable data.
According to Alex Salkever, Minimum viable Data is intentionally designing processes that will collect and store as little customer data as possible to be successful and grow.
One would ask why building for Minimum Viable Data is a better path to growth when more data seems to equate more users. Startups want to grab as much data as they can even if it’s not for immediate use. What if it’s better to get as little data as possible for the growth of the company and also a way to reduce the long-term risks for the company?
The Hidden Costs of Too Much Data.
Storage costs; these are not necessarily the costs of keeping it. These are costs incurred in moving the data in and out of storage.
Unknown risks: how often do people leave the company, someone could leave and not reveal how they stored some of the extra copies of data. It could be on a hard drive that no one knows about. Every data you collect creates more unknown risks.
Maintenance: data is not a static entity that you collect and forget about. Working with data for years will show you that it grows and has a life of its own. You may have to change the way your data is structured over time. It requires maintenance.
Technical debt: the concept that over time technology decays and has to be replaced. Same goes for data debt because it is actually tied to technical debt.
The biggest risk of these is the unknown risk.
The best reason to have as little data as possible is because it is the right thing to do. If you were a customer, what would you prefer, someone to collect as little information about you as possible, or as much? This wouldn’t be a problem if everyone told you what they were doing with your data, but they don’t. Once the data is collected, there’s usually no way of taking it back.
It is easier to start with Minimum Viable Data than to add it on later on as a process. Design from the beginning for your business to run with little data as much as possible will allow for an easier upgrade later on. You will be dealing with something much simpler to work with. This will be part of the DNA of
Six steps to Minimum Viable Data
This is a checklist that will help you build a Minimum Viable Data structure for your company.
Decide what data you need and why
Test to make sure it’s enough data
Do a data audit to understand where your data lives. Always ask your partners and service providers.
Create a data policy with playbooks and assigned roles. Make sure this includes a plan for data breaches or crises.
Design your systems to comply with the policy
Reevaluate twice a year to tweak and improve.
Minimum Viable Data is not an end state; you will have to continually improve it.
Most people do not discuss the dangers of data misuse, yet the effects are evident in the data leak scandals that we hear about in the news. Startups are so used to collecting big data without planning for its use, storage or maintenance as they grow. Minimum Viable Data, is a concept that everyone is encouraged to adopt today, for a safe future.