In today’s economy, analytics, automation, and artificial intelligence (AI) provide the fuel for new, digitally enabled business models. But if we look at how most organisations operate, analytic processes are based off legacy tools, waterfall methods, and manual steps that can’t keep pace.
ITP.net sat down with Alan Jacobson to talk about how businesses and processes are evolving, and the role data and AI play in that transformation.
What are the main challenges to adapting? What needs to change? The processes or the tech?
It’s certainly well known that change in business is hard. The reason why is deceptively simple. It’s because change as a human is hard. The larger the change, and the more people who need to change, the more difficult that process is to execute. Whether it’s changing your personal habits to exercise more, or learning a new tool or process to do your job, change typically requires three key elements:
- Awareness and excitement: People need to understand what the benefit is for them personally. They need to be excited and motivated to make that change. Sometimes that motivation comes in the form of an existential threat i.e. the company they work for being at risk of going out of business if they don’t learn something new.
- Enablement and training: People need to be given the necessary education to monopolise on that initial excitement.
- Support and sustain: Once you have excited and trained people, the priority becomes supporting and encouraging that action. In business, this would take the form of mentorship – helping people gain their first success, deal with their initial setbacks or challenges, and continue to hone their confidence. Supporting desired behaviour is key to long-term successful change.
Companies frequently perform well on the second step of this process, but most are weak at the first and third. So, while you need to pick strong technologies, that certainly isn’t enough by itself. Companies that are world-class at analytics and AI have strong leaders directly driving that transformation change effort across their enterprise.
The difference is that most analytically driven organisations use ‘analytical thinking’. This empowers them to understand the data sets available, and the governance considerations required to use them. A five-year download of live production data would not be useful to a modern data team, but a cleaned and ‘analytics ready’ dataset is a hugely valuable tool and a significant time saver. In business, the domain expertise and the context behind the analytic process is essential to deliver insights in a more efficient way.
Clearly, creating insights is only part of the equation. Businesses need to put more emphasis on human input, upskilling, and creativity – all working in tandem with advanced technology, such as AI, to find useful patterns and meaningful correlations. Our experience in helping our customers on their digital transformation journey consists of upskilling your domain experts – the people who own and are improving the processes – so they have a direct stake in the outcomes. It is far easier to teach an accountant some data science than it is to teach a data scientist accounting. What needs to change is the attitude towards technology upskilling.
Recent Alteryx commissioned research with YouGov around data analytics in the UAE and KSA found that 97 percent of Gulf data workers believed that more training is needed to unlock the potential of AI technology in the Middle East. The results consistently showed that legacy processes, legacy technologies, and siloed data were stalling that crucial journey to becoming a data-driven business able to take advantage of AI.
What’s wrong with the old way of doing things?
In many cases, current analytic and automation tools allow domain experts to build solutions faster than they can write a request for someone else to do it. If you have knowledge workers needing to request data analysis or process automation instead of operating in a self-service environment, then their operational pace will be completely uncompetitive compared to modern analytically mature organisations. In addition, if you hire analytic experts and they are asked to do these very simple day-to-day tasks, then they will likely leave.
According to Gartner, by 2023 data literacy will become an explicit and necessary driver of business value. The world’s most analytically developed organisations already have a large percentage of their core workforce able to leverage analytics technology beyond the legacy spreadsheet. While legacy systems undoubtedly still have an important place in business — namely for general administration — they aren’t advanced enough meet the needs of business today and deliver reliable insights. Employees who can analyse data and extract insights to drive their businesses faster and further is a huge competitive advantage today.
What does cloud data allow us to do that we couldn’t do before?
Jacobson: Cloud can mean a number of different things to various organisations. For some, cloud implementations allow companies to have different datacentres that are geographically friendly to where their users and/or their data sits. This then reduces the amount of data movement – saving time, but also providing necessary legal compliance depending on the use case.
For others, it is making full use of the ever-decreasing cost of data storage compared to the same storage on premise. In general, cloud technologies typically provide greater hardware elasticity needed to scale up and down compute resources when needed. The alternative is designing systems around peak load and paying to keep that infrastructure running even when the resources aren’t needed.
This scaling capability of the cloud can also be quite important when analysing large datasets – particularly where significant compute resource is needed for a short period of time, but peak load is not needed consistently. Cloud-based analytics not only facilitates powerful data processing but also enables insight generation at speed and at scale.
Do we have too much data? Often when I speak to people, they point to absolute massive amounts of data that we don’t necessarily know how to use.
Today’s cost of storage is incredibly low and continuing to decrease, making the value proposition of creating and storing data an easy equation. While digital transformation undertakings often focus on technology, however, a good leader will focus on the human element and the cultural change required to enable that factor. Upskilling workers in data literacy and democratising the access to data are core to getting the most out of any digitally enabled business model.
The importance of data cannot be overstated. While many businesses certainly have more data than they can use, I wouldn’t classify it as “too much” data. Instead, it’s a question of potential. Globally, we’re forecasted to see 180 zettabytes created each year. In 2020, that figure was 64.2 zettabytes. The USP of the five most profitable companies in the world isn’t cash reserves or real estate – it’s the ability to collect, analyse and act on the data they hold – at scale – to refine the value and make effective decisions.
More accessible AI is helping businesses make timely and contextual decisions that lead to more competitive outcomes, and that is only possible due to the ever-increasing volumes of data that businesses have to work with. AI is, in short, one big pattern recognition machine – albeit one shrouded in jargon. You train the AI model by feeding it data, and then ask it to predict outcomes based on that data.
The question now isn’t around having too much data – it’s about what insights can be gleaned from this hugely valuable resource and then used for competitive advantage.
