Collection No 6
In order for future workers to find the data they need and make sense of the results, tools should aim to simplify and human judgment must not be overlooked.
As businesses around the world increasingly depend on data to inform their operations, the imperative to have effective data strategies only grows. From the tools that surface the data, to the strategies that turn information into insights, we must design toward a future where workers spend less time wrangling data and more time fostering human connections. What follows is an edited transcript of a conversation between frogs Patrick Kalaher (VP Technology Strategy) and Tjeerd Hoek (VP Creative) as they anticipate how this future will unfold.
1. We have heard frogs talk about how, in the future, data will be free from its container. What does this mean?
Patrick: That statement is true in a lot of contexts. One way to look at it is through the lens of customers having increasing access to data. An example is a guy who found a quirk in the way airlines do pricing, and then created a website to exploit this quirk. The idea is that if you want to go to Las Vegas and it would cost $1,000 to travel from Boston, you could instead book a flight to Los Angeles with a connection in Las Vegas that would only cost $600. So you could book a flight to Los Angeles and then get off in Las Vegas and not complete the journey. For United Airlines—who is now in a lawsuit with this guy–that information, pricing strategy, and data is now not only accessible to the public, but people are making companies around the information. In that context, data has already been released from the container, and it is pretty shocking for the airline companies.
This example also speaks to a bigger trend, which is the democratization of access to data within companies. In many cases the CIO no longer has a privileged perspective because lots of people inside the organization can get access to the information that runs the business. This is an inversion of the control hierarchy that we’ve always had, and it represents a larger way in which data will continue to be freed in the future.
Tjeerd: From a design perspective, we constantly need to free data from a single container and make it widely accessible, because human beings do not have singular needs. So the human needs force us to architect databases in a way that supports flexibility and freedom. These databases are less clunky and less single purpose than older technology, and they allow data to support things that human beings actually do in the world.
2. In addition to changes in the way data is controlled and accessed, data is now being captured in new ways. How will sensors and other technologies that collect data impact the workplace?
PK: We can compare it to changes we have seen in sports, where the referee used to be free to make a judgment that was trusted but is now evaluated based on measurements recorded by sensors. In baseball, when the ball goes over the plate, they now measure whether the ball was in the strike zone. And they’re also judging the umpires based on their ability to make an accurate call. It’s causing many of the umpires to change the way that they work.
Workplace sensors will create a similar kind of scenario, and people will change their behavior accordingly. If I’m a truck driver and I’m being graded on a particular metric, I’m going to change the way that I drive, because I’m going to be conscious that I’m being watched and evaluated on a specific data point. I don’t think that we know what the final outcome of that is going to be yet.
3. What is the effect of these trends on jobs—like data analysts and knowledge managers—that focus on bringing disparate data sets together?
TH: Experts will need to be hired to make sense of the information. Software can be designed to create to-do lists, timelines, or other kinds of smart notification systems that are incredibly clever and relevant to what a worker needs in a specific moment. But people have to create algorithms that surface that information. And people have to make judgment calls when looking at large data sets and coming to conclusions, particularly when you want to look at data through the lens of regional variances or other aspects of the human condition. Those types of issues are often a grey area to a machine system, and designers have to be honest about these things. They must know when an algorithm is going to generate a useful set of content, and when human data analysis will be required.
PK: I agree. There are always going to be situations where we’ll do sophisticated data analysis, which will drive a decision that is ultimately made by a human. Whether it’s modifying a supply chain or executing an order on a trading floor or deciding that something is a crisis that needs an emergency response, there are many scenarios that will involve human oversight well into the future. The job of analyzing data may actually end up getting a little bit harder as the amount and complexity of the available data increases, even though we’re supposed to be making it easier by creating more sophisticated automated analysis. It will require a sophisticated skill set to know when and how to intervene.
4. How should search functions evolve to keep up with this increased complexity?
TH: They have to improve. I remember when we worked on a large search engine for a tool that was built to search publications, and it could not understand the umlaut. It would just return a zero search result. Sometimes you’re just baffled by how stupid the search engines are.
Now, obviously, these types of problems rarely exist. Newer semantic search uses string matching to anticipate what people need as they are typing. But at some point, even with semantic search, you’re going to get so many results that the data is unusable. Those results will have to become more controlled and organized; otherwise you’ll never get through it. This is going to lead to a much, much better world.
PK: We also can’t think that the results we see from search are the whole truth. The average person seems to think that when they Google something, the results represent a snapshot of the entire Internet. But it is generally accepted that the deep web is 3,000 to 4,000 times the size of the web that has been traversed by Google. I think as long as we’re aware of that, and we’re constantly questioning everything, we’ll be okay. It is when we sit back in our chair and say, “Okay, I’ve got this search tool, so therefore whatever comes back to me is the canonical result” that we start to get in trouble.
5. Search is already benefitting from semantic properties – what about voice recognition? Does it play a role in the management of data and information in the future?
TH: I’m interested to hear from Patrick about how the more personified versions of voice – like Siri and Cortana – are trending the United States. Here in Europe more and more people seem to be using it daily and really relying on it.
PK: I’ve had the pleasure of working with the new Amazon Echo, the Pringles shaped thing that sits in your house and listens to you. You can ask it questions at any time and it will answer you. To me the interesting thing is the combination of the personification Tjeerd mentioned, and its link to the concept of persistence.
The fact that you can count on the Amazon Echo product to be there when you need “her” is oddly comforting. Another example is Siri – I think it is comforting to people that Siri is always there even as they move around the world. They can ask her anything, and even if she’s wrong, she’s wrong in a predictable way. She also has a sense of humor, which you’ll know if you’ve ever played around with this on an iPhone. I don’t think this technology really works for everybody, but it is very useful for a lot of people.
TH: The format of interaction is definitely important, but the interesting question for most people will be can it really do something for you? How it deals with failure is also an important point that Patrick mentioned. The notion of graceful failure, that is actually kind of human, can make you a lot more forgiving. Right now speech controlled systems often fail in such a dismal way, so many people often just give up after awhile.
These systems have to get more sophisticated and human, and address scenarios that add real value. Then we can have the search engine being accessed exclusively through voice, or a personal digital assistant that will really help you organize your life. But first our software systems must move beyond mechanical little tools and focus instead on how we function as people in the world.
6. What do these changes mean for the future worker? How do they advance the human experience?
TH: The purpose of these developments we’ve been discussing is to support people and increase the meaningful contact that they have with each other. But in order to realize that reality we need to design, both in the mobile space and in the workspace, software that allows for the connection to exist. Tools should not suck you in completely, but instead let you be in the world as you are using them.
I hope that with better tools we can stop the constantly increasing amount of complexity and distraction that comes at workers today. Software should do some of the work for you, surface what’s really important with respect to your intentions, purposes and current task, connect you to the right people, and then allow you to continue on your way.
Patrick is a vice president of technology strategy at frog. He has an extensive background of product ideation, product strategy, and implementation in a number of business domains across North America, Europe, and Asia.
Tjeerd is a creative leader focused on experience design for software and technology-amplified products and services across frog. He started the frog studio in Seattle in 2007, moving back to his homeland to join the Amsterdam studio in 2009.