Augmented analytics is when you take what was traditionally a very manual workflow and automate it. This gives you the ability to analyze data far more rapidly and to package up changes for humans to interpret. Essentially you’re augmenting a human experience, so rather than spending all your time looking for a needle in the haystack, the machine finds the needle and gives it to you. By bringing the human and the machine together you can create something very special and deliver that to an end-user.
There are three trends that I’m seeing in augmented analytics at the moment.
1. Most of the value is in discovery
There are four primary use cases for augmented analytics today - data preparation, profiling, augmenting data with knowledge, and continuous discovery.
Data preparation gives you the ability to prepare data faster, which is a one-off experience. For example, if you bring in Marketo data and use automated data preparation to augment that, you do it faster but you don’t need to do it again.
Once you have your data in a tool you can automatically profile and discover the shape of that data. A lot of vendors use augmented analytics to profile data and create charts from that profile. While this is interesting because it shows you the shape of the data, it doesn’t really help you discover insights in the data. Profiling also doesn’t have real-time capability, so it can’t continuously monitor your data or drive your business further.
The third use case is where people augment data with their own knowledge. They do this by taking charts and providing context around that data through a story or presentation, for example.
Finally, there’s the automation of continuous discovery. This is where the machine constantly looks at your data to see what's changed, how it’s changed and then alerts you to those changes. If we look at the industry today, the most value is around this automated discovery use case. It’s the only sector of augmented analytics where we see independent vendors bringing unique solutions to market.
The discovery use case is also where machines are most effective because the algorithms for discovery are well known along with the concepts of what people are trying to discover and how they approach discovery. The manual data discovery process is not valuable at all - you learn nothing by dragging elements onto a chart over and over again, but once insights have been discovered they’re very valuable. That’s the appeal of automated data discovery - you get instant insights.
2. Vendors are diverging
In the market, we’re seeing a divergence in vendors. Some vendors believe that augmented is the future and others definitely don't. It’s like the horse and buggy conversations of the past where you were either on the bandwagon looking forward or you weren’t.
Vendors that offer augmented analytics are getting market traction thanks to early adopters. In fact, there are now specialized vendors emerging who only do augmented analytics. The market for augmented analytics clearly exists and in the future people will demand it.
On the flip side, some vendors in the BI and analytics space are somewhat in denial. They're not actually reacting to these changes or building products with augmented analytics. They just seem to be hoping that automated discovery will simply disappear but they may find that they’re left out in the cold.
3. The UI is converging
For those vendors who are going all-in on automated discovery, we’re starting to see convergence in the way in which they're solving the problem. Similar to what we saw with dashboards in the early days, there’s a commonality in the way in which content is being delivered to businesses which is fantastic. This is highlighted in the UI and user experience. Customers know that they’re getting a feed or trigger. For example, they may see a snippet of a chart that tells them what’s happening and when they open it they can see a lot more detail about it. If they’re interested they can then analyze it further.
Having a common language makes it easier for customers to adopt the technology because they start to see similar patterns and use cases. This makes it easy for people to understand the discovery use case, which is really important if you’re going to compare vendors in this space.
What makes augmented so attractive is the automation piece. If you’re crushed for resources and don't have enough people to analyze your business you can automate the process. If you trust the automation process then you've got nothing to lose by turning it on. You can just start to gain insights straight away and that's the appeal for early adopters.
Get instant insights with Yellowfin Signals
Imagine if insights found you. Then you'd always know when, what and, importantly, why something happened. That's what Yellowfin Signals delivers - powerful, AI-driven, automated data discovery so you get instant insights.