And so, they enjoy great big helpings of hype. That hype is not only justified, but good, as it is through the widespread attention focused on the concepts and technologies that the trough of disillusionment can be traversed, allowing organisations to start levering their way up the slope of enlightenment.
That much emerged in a broad discussion with Matthew Quinn, Tibco CTO, and his colleague Michael O’Connell, Chief Analytics Officer at the integration and intelligence software company.
Kicking things off, the outspoken Quinn (who hails from Australia originally) said there are ‘a number of trends that have popularised analytics’.
“You do have to be careful as BI is a very broad space; The broader BI market hasn’t grown all that much, as it contains the bog-standard reporting and KPI stuff. While people like Gartner are projecting growth in visual analytics and discovery, analytics as a segment has grown, driven by the ‘democratisation of data’,” he said.
The latter term is, according to Quinn, “Just a wanky way of saying people can access the tools to empower themselves through data formerly hidden in silos. But it is still the people operating systems who have the best ways of spotting patterns, so giving them these tools enables them to take advantage of that data.”
Back in February this year, we had a look at Gartner’s expectations for analytics market growth this year; it said ‘the market is in the final stages of a multiyear shift from IT-led, system-of-record reporting to business-led, self-service analytics. As a result BI&A platforms have emerged to meet new organisational requirements for accessibility, agility and deeper analytical insight’.
Continuing, Quinn said the advent of big data, against which traditional reporting wasn’t able to provide insights, has resulted in what he describes as ‘huge hype’. “That, along with the democratisation of data, and tools maturing to produce the pretty graphs and in a compelling way, has shocked the system. People don’t want a spreadsheet; they want to deliver an interactive data experience to tell a story to colleagues.”
Quinn said it is this that the excitement in the analytics space is about. “It is the way people are being empowered with the tools to tell a better story, to explain what is going on in a way that people ‘get’ rather than staring at boring spreadsheets.”
Communicating the insights
It’s an interesting point. Gaining an insight is one thing, telling everyone else about it, another. His colleague and Tibco Chief Analytics Officer Michael O’Connell said the trick, as always, is to get the right insight from interactive analysis, and then to collaborate with smart visuals. “Businesses have a lot of data; this is why the focus on analysis is so prevalent, because it is almost like there is too much data and not enough action from it.”
That aligns with the results of a recent Teradata study, which found that while data gathering is on the agenda, a quarter of ANZ businesses are doing exactly diddly-squat with it.
While O’Connell said Gartner, on the one hand, tends to look at desktop and visual BI, “Forrester on the other has identified the ‘insight platform’, where companies want to take action on insights with visual analytics; sourcing the right data with predictive pieces and implementing those insights into software that can take an action, then capturing the result of that action for optimisation.”
It’s something of a lifecycle approach, where actions can be taken, assessed and fed into a Kanban-type cycle.
Quinn said he believed the ‘dip in big data’, which as long ago as 2013 entered Gartner’s Trough of Disillusionment, is because the ‘tooling was seen as a science experiment’. Now, we’re seeing a second wave in terms of operationalising the decision-making, making it add business value. This is the right place for it to have gone. Big data, from being a popular press topic, will drop off the radar as it goes from ‘sexy cool new’, to everyday technology – and it is here where a lot of the value is created.”
Indeed, Gartner’s most recent hype cycle doesn’t even include big data; writing at LaserFiche, Sharon Fisher noted, “That’s because, according to the company, Big Data is so pervasive that it can’t really be considered an emerging technology any longer. Gartner analysts Betsy Burton and David Willis telegraphed this last year during the release of the Gartner Hype Cycle 2015, when they announced that the company would no longer do a Hype Cycle for Big Data technologies.”
While hype of anything is pretty annoying (and our industry is never short of hype; indeed we are to blame for it much of the time), it is something of a necessary evil, agreed Quinn. “I’m not pooh-poohing the hype. We have to get people excited, that fuels the necessary steps from idea to adding value.”
And what could be more hyped right now than machine learning and artificial intelligence. Indeed, added Quinn, everyone is talking about it now. “The broader space needs the hype so people can take advantage of awesome technology, which was amazing 25 to 30 years ago, but needed advances in computing and databases to be workable.”
Quinn said it is a ‘hugely important time for AI’. “The recent influx in interest will fuel the value – because, even after 30 years, it hasn’t really yet cracked it in terms of how it can be put to real use for business.”
Augmenting rather than replacing
Where Quinn sees the greatest impact is in augmentation, and specifically, augmentation of the intelligence offered by analytics tools. “Machine learning and AI can’t replace a data scientist, but it can provide recommendations which can cut an hour of work here and there that a skilled data scientist might otherwise have had to do manually. And AI is also likely to help with user experience; as technology continues to evolve in complexity, AI will assist users to make decisions in a user interface, directing attention to the right places.”
O’Connell said that machine learning is a small part of the equation in achieving meaningful and useful AI. “Fundamentally, it is still a combination of getting the right data to address the right business problem. You want AI to help the person doing their work in their workflow; AI needs to embed into daily workflows, so it can do maybe 80 percent of the effort.”
Echoing Quinn, he too added, “It is about augmenting, rather than replacing.”
For some readers, the hype around AI, machine learning and, indeed, big data (in the 1980s, big data existed, it just referred to the overwhelming megabytes which had to be coped with) might have a familiar ring to it. The concepts aren’t new, after all. What remains to be seen is if this time round, the expectations, heightened or otherwise, can be put to meaningful use.