For example, to prevent credit card fraud, a bank needs to make the right decision in real-time, at point-of-sale. Once the transaction has gone through, the money has gone! A great example here is the work SAS does with HSBC globally. They manage to protect every credit and debit card transaction in real-time.
Uber has recently introduced a real-time security check on drivers in several countries. The driver is prompted to upload a selfie and Uber's software compares this with the photo held on file. If there's any doubt that it is the same person, the account is temporarily suspended. This real-time ID check means that users are safer. They can be confident that the driver who picks them up is the person they claim to be.
Real-time analysis is also good news in healthcare. For someone with an acute condition, it is no good waiting half an hour before responding to a change in condition. Healthcare professionals need to react instantly as chronic conditions can also deteriorate frighteningly quickly. Getting rapid feedback to people is important! It is so much the better if an algorithm can initiate the necessary response in real-time, even if people may have more than one condition, such as diabetes and blood pressure.
Real-time decisions are also important when it comes to digital marketing. If a company wants to use web activity to drive an offer, they may want to do it in real-time whilst the person is still on the website!
Is real-time always the right time?
These are cases when real-time analysis really is changing things -- real-time is also very much right time! But is this always the case? Will insights always be better for analysis being done in real time? Will the return from implementing real-time analytics always justify the cost and complexity? I think the answer to both those questions is "no." Sometimes the business opportunity is delivered by analysis every month, or week, or day or every 15 minutes! It is totally dependent on the different business scenario being addressed.
But I would also go further. Sometimes real-time may even be the enemy of right time. By rushing to analyses and decide in real-time, rather than waiting until the right time, we can miss insights, and fail to take the right actions.
Right time decision-making is about providing the right information to the right people when they need it to support decision-making. Some of this information may be in the form of real-time data, and some of it may not. It may also be important to bring together real-time information with other, stored and aggregated information to get the right insight, or store some data obtained in real time until later, when you can combine it with further information.
Right-time data, in other words, is data that is used at the best possible time to maximize its impact on customers or the organization.
Different times and different context
Consider online shopping. Plenty of people put something in their shopping cart, then wander off elsewhere. They might come back later, or they might not. Real-time analysis will tell retailers when customers have abandoned their purchase, but that's about all. Right-time analysis will also add information about customer behavior from previous transactions, or even the behavior within their social network that may take time to occur. Perhaps all this information is required to help the retailer determine both what to do next, and when to do it.
The same type of data can also be used both real-time and at the right time, but at different times and in different contexts. Social media data, for example, such as mentions on Twitter, could be from either happy or unhappy customers. You need to monitor it carefully, because it is important to respond to unhappy customers rapidly -- if not quite in real-time, pretty close to it -- if you don't want their unhappiness to spread far and wide. You need a rapid automated (but real-sounding) response that will get them off social media and into the personal messaging space, when you can deal with their complaint privately. Happy customers, however, provide useful feedback that may support further marketing or sales campaigns: responding is not urgent, but the data is extremely useful because it tells you about customers wants and needs.
The bottom line
The answer is that sometimes you can or must respond immediately to data, sometimes you need more information and sometimes there is no additional business value in making an immediate decision. Context is all-important. Even in healthcare, urgency can be the enemy of good: The same indicator may be bad news for one person, but good news for another, with a different condition. Real-time data is definitely good news, but companies need to think carefully about how they use it, to make sure that it is also the right time.
This content was reposted from SAS Hidden Insights. Go there to read the original post.