What Does Grandma Have to Do with Big Data?
Recently, I logged on to my bank (to remain nameless), and I was momentarily shocked to see that all of my accounts were at a balance of exactly zero. Now, if it were my daughter’s checking account, zero would have been better than something in the negative range, but to see nothing but a big, fat goose egg, gave me a moment of pause–but just for a second. I noticed it was just after midnight and realized that the system was probably down for routine maintenance or a backup or something like that. I also, in my heart of hearts, knew that my data was safe somewhere, backed up on some SAN or NAS or dare I say, tape drive. I just didn’t know where. Now, even though I knew my data was safe, I was still frustrated and downright peeved that the system wasn’t available. I couldn’t transfer the funds I wanted to or pay the bills that I needed to because the system was completely unusable. Scarily enough, this isn’t the first time this has happened with this bank. You would think that they would have learned by now. That was the most frustrating realization of all.
Considering I have been in the technology industry for twenty years, I had the history and experience to know that there was likely nothing wrong. Then, it occurred to me that not everyone has been in technology for twenty years. I thought about the emotional impact an experience like that would have on my Grandma if she had gotten online only to see that everything was zero and her account was unusable. Her reaction would have likely been vastly different than mine.
Now I don’t know about you, but with my Grandma, trust was a big word. That’s what all this comes down to. It took years before she was even comfortable getting an ATM card. What do you think she would have done if she had logged in to a zero balance? Big data comes down to trust: if I am going to give you my money, my identity and my business, I have to trust you.
Over the last twenty years, I have seen technology morph in ways I couldn’t have even imagined in 1993 when I graduated from college. The pace has been astounding, not just with cool new gadgets but with how quickly technology has permeated its way into every household, including Grandma’s. In 1977, in the early days of the RDBMS revolution, it would have been difficult to imagine the scale and scope of the transformation technology has had on the way we do business. But many companies today are still powering their businesses with technology that, while marketed as new, is layered on decades-old code built with old assumptions and a design that was never intended to scale to modern requirements. They are encumbered by investments that have gone way beyond their prime. Whether we acknowledge it or not, whether we know it or not, we pass the cost of these investments on to our customers every time we fail them. We forget that the scope of this transformation has a human face. It has the face of not just those of us in technology but the face of, yes, Grandma.
Back in the days of the VHS and Beta wars, the only thing we had to worry about was whether or not the video we wanted was in stock. Today, companies like Netflix have to consider the impact of weather, system outages, disk failures and other unforeseen events on our ability to watch our favorite shows. If Netflix has an outage, they lose credibility, not with one customer at a time, but with millions.
To overcome the challenges of the 21st century, Netflix had to drop the biggest barrier of them all–the shackles of the technology and ideology born and bred in the 70’s and 80’s. They had to embrace a new way of thinking so that they could preserve the relationship and trust with their customers. The choice they made was DataStax and Cassandra.
Today, Netflix cannot only withstand a simple node outage but they can withstand an entire data center outage and still provide service, uninterrupted, to their customers. By the way, they can do this at a fraction of the cost of legacy technology.
To make the leap, Netflix had to open their minds to a different approach. They had to leave behind the dogma espoused in 1977. They had to ignore the fear, uncertainly and doubt that entrenched vendors are pushing in the market. They had to leap, and leap they did.
Now let’s go beyond entertainment and consider my experience with my bank. It could have been avoided entirely. Without the constraints of old technology, I would have never had to experience what is equivalent to a full outage. My trust would not have been shaken and my patience would have been fully intact.
As technologists we have to consider the impact on the average person who is not a technologist and doesn’t understand the disruption that a backup or a node failure or some other kind of unintended event causes. We have to seriously consider the trust that goes hand in hand with technology. Trust isn’t just about the expectation that our user ids and passwords are secure or that someone isn’t stealing our identity. It also has to do with the expectation that when I pay for a service, I trust that when I want to use that service, it will be there. It has to do with emotional impact that a failure has on the average person.
As technologists, we also have to come to the realization that our hesitation is mostly emotional. New approaches to modern problems mean we have to learn again. We can’t rely on just the experience we have had but use that experience to come to the realization that traditional approaches do not work anymore. If you have a system that will touch the average person, you have to think differently. Failure is not an option any more.
We are yet again, at an inflection point. It is a transformation driven not only by transactions per second, but by trust per second. In a world where one mistake can generate millions of tweets, do yourself a favor, and at least consider that there might be a better way. Re-think what you are willing to sacrifice for a failure. Do it for yourself. Do it for your customers. Do it for Grandma!