Business today is more than simply matching traditional competitors, it’s about exploiting digital technologies to create new opportunities, and being able to repeat this.
The economy is quickly going digital and Australian businesses must evolve into Modern Digital Businesses (MDBs) which strategically use intelligence assets to improve operations and deploy new products and services, in order to stay competitive and create value for their customers.
A group of digital business leaders recently gathered at ThoughtWorks Live in Sydney and Melbourne, to share their insights into how organisations can take advantage of data to adapt and thrive in the digital economy.
This report includes strategic and practical advice taken from the event for any business leader – regardless of their organisation’s digital maturity – on best practices for taking advantage of data and driving change.
A Continuous Intelligence (CI) framework starts with the process of acquiring data and, with the help of analytics and machine learning, derive insights from it to be able to make confident decisions and actions – which are in turn reviewed and validated, to ensure the organisation continuously improves its decision-making capabilities.
Steps organisations can take to apply CI to building an MDB, which is agile and technology-driven are also covered.
If a business can continuously try new things, across new verticals, it can find new revenue streams and new efficiencies to support growth.
The makings of a Modern Digital Business
Most organisations have modernisation programs to ensure products, services and operating models are updated and not holding back new development potential.
To remain relevant, Australian businesses must evolve into MDBs, which means modernising their technology estate; harnessing data assets; building digital ecosystems; focusing on customer-centred products and services; and continuously evolving digital assets.
This will result in an agile tech-enabled organisation which places its customers at the centre and is business outcome oriented.
Without modernisation and digital, businesses will limit their capacity to service customers and bring new products to market.
Data is a core pillar of an MDB, and by building a continuous intelligence capability, an MDB is always taking advantage of the data it has available to it.
Figure 1: How continuous intelligence turns data into outcomes.
CI ultimately enhances a company’s ability to deliver value to customers and drive performance. CI is not just a conceptual framework but a capability that is continuously exercised – through an organisation’s systems, process and culture – to deliver value.
A good example of how a MDB can create services from data is PSMA Australia. It delivers access to authoritative national mapping data and operates as the hub of relationship management between data custodians, data managers and clients. Its clients can develop innovative location-based services.
Another example is Arkose Labs, which globally delivers modern digital services to prevent fraud during e-commerce and other online transactions. Its success shows how data science and machine learning are not bolt-on ideas, but integral to its business.
These companies use new technology, processes and services, including machine learning, to acquire data, process it, and, ultimately, exploit the insights for new business creation. To make this a core and repeatable capability, they have evolved their structure, processes and culture to match.
ThoughtWorks has developed a maturity model for machine learning in organisations which their consultants use to help clients identify practical next steps for every stage of the CI journey.
Keeping up with data acquisition needs
To make informed decisions, today’s enterprise has to acquire data from various internal and outside sources, including the Internet, customers and partners, smart ecosystems and connected devices. Access to this data and information demands the right foundations and capabilities – such as data architectures, platforms and advanced analytical capabilities.
But much of the legacy infrastructure that organisations carry today wasn’t built to cope with today’s data volumes.
Speaking at ThoughtWorks Live, Jo Abhayaratna, CTO at PSMA Australia, said a core part of the organisation’s data acquisition strategy was to modernise its technology platform.
“This included adopting cloud and microservices,” Abhayaratna says.
So, improving your IT estate by replacing ageing, incumbent systems with more flexible options will help control stage one of the continuous intelligence capability: acquiring data.
In today’s hyper-competitive market, companies need an IT infrastructure capable of managing the CI cycle in days or hours, not weeks or months. That means applying continuous integration and continuous deployment techniques to your data capabilities.
To get a technology landscape ready for more data acquisition, the structures and cultures must also change and digital platforms are not just about upgrading or cost savings – they are about enabling employees to innovate.
Take advantage of new tech
Businesses looking to modernise and improve how they acquire data should look to a range of new technologies which will help expedite the journey.
Dr Gerald Hartig, Principal Data Scientist at Arkose Labs, spoke about applying continuous intelligence to defeat online fraud and abuse.
He recommends embracing DevOps and cloud and not building what is not critical to your business. And if things are working as expected, don’t be afraid to pull the plug.
During the modernisation journey, including the data acquisition phase, companies should investigate the application of Machine Learning (ML) technology as a core component of their infrastructure.
“ML is not a bolt on, it must be baked in,” Hartig said.
Turning data into information
Before it is useful in a business context, data must be transformed into information through cleaning, curating and classification.
In the CI cycle, raw data describing business domain events is turned into information which, in turn, is used to identify and exploit new business opportunities before the competition.
Better data processing can also help organisations prioritise investments, automate their best decision-making processes and use human intuition where it matters most.
Many business leaders are still sceptical about investing in data processing; however, Gartner has highlighted how it is now fundamental to business transformation and has a greater role in generating business value.
Many businesses are still trying to make sense of how they can best use data, but business leaders must be more circumspect about their data initiatives.
At PSMA Australia, CTO Jo Abhayaratna believes demand for good quality data is growing across all industries and at the same time more data is becoming available mainly due to new capture technology.
“Our vision was to capture the built environment which had never been done before and there was a gap in geospatial data around locations,” he said.
This was achieved through continental mapping and the fusion and processing of satellite image processing, machine learning and crowdsourcing.
“The challenge was to think differently, and now the business is built on data processing and is democratising data services.”
PSMA Australia is an excellent example of how previously unknown insights are now available thanks to advancements in data capture and processing.
“Data from suppliers is largely unstructured,” Abhayaratna said. “We make it ready for solution partners.”
Start with what you have
Most companies are not at the scale of a Facebook or Google when it comes to data processing, but that should not deter them from getting started.
“A low volume of data should not be an impediment to finding more about your systems and how they interact,” Hartig said. “Let the data talk and start with the value it can provide. You can always find some value with the data you already have.”
Garner insights to enter the digital economy
Data that has meaning and is fit for consumption and analysis can now begin to deliver some real insights on the business and other external trends.
In the insight stage of the CI cycle, data science models of the real world have to be created, trained and tested.
ThoughtWorks’ Technical Principal, Machine Learning, Mat Kelcey, recommends performing experiments to gather insight into which of your available data is most valuable.
With more insight, the business can work towards moving from descriptive to prescriptive analytics, which delivers the ability to make better decisions faster.
Build ecosystems with collaborative insights
Data can provide immediate value to an organisation, but it also provides value to a partnership ecosystem, where groups of businesses collaborate to deliver new products and a better customer experience.
Within the CI framework, when moving from information to insight, companies should assess how relevant their information is to their partners, or how partner information might be useful to them.
In the UK, a consortium of businesses have collaborated to streamline the post-trade process using blockchain technology. This level of business efficiency would not be possible if each organisation was not willing to share their information.
The collaborative power of the digital economy is demonstrated well by PSMA Australia, which, according to Abhayaratna, is part of a rich data ecosystem.
“We found out how to manage our platform as a product,” he said. “In future anyone can be a supplier and build an ecosystem.”
To become part of a data ecosystem, Arkose Labs’ Hartig said try and “plug in as much as possible”.
Even with a small amount of data “you can become a component in someone else’s workflow,” he said.
“You might need to redo the backend of your product to allow ML to be integrated into your system and it might also change where in your organisation you are making decisions and where your product is and how it is delivered.”
Speaking about the role of ML in modernisation, Karen Davis and Mat Kelcey from ThoughtWorks advised reviewing machine learning as part of your tech modernisation program as it will coexist with other services.
“By getting into an iterative process with ML, you can apply it to use cases which deliver the most insight sooner,” Davis said.
Even with emerging data services such as machine learning, Davis said there is never ML “by itself, it will relate to other ecosystems”.
Better decisions reduce risk
The quality of decision making depends on the quality of information and how it is modelled and processed. Key to quality modelling at all times is the capability to rapidly and confidently update predictive models in production environments in response to changing factors.
This leads to the decision stage of the CI cycle where models have to be deployed into production to serve a business purpose.
Moreover, the execution of these models leads to new data which might be in the form of sales, client behaviour, supply chain updates or environmental changes, and may be another source of information and decision-making in turn.
Stage four in the CI framework relates to planning and prioritising actions to prepare them for production.
Use case: Customer-centric decision making
Given the immediacy of digital, from industrial systems to smart watches, customers are more demanding than ever and are front and centre of every touch point.
Data can be used to make your services more accessible and gather customer feedback wherever you can. With more information from customers (and partners) you can make better decisions and therefore changes.
For example, if apps are difficult to use, customers will quickly seek alternatives or vent their frustrations on social media.
At PSMA Australia, Abhayaratna said the team had to shift its thinking to becoming more customer centric – an important step in its CI cycle.
“We deliver data to customers in self-service interfaces, so we have a more rapid feedback loop,” he said. “We found out how to manage our platform as a product and learnt this from ThoughtWorks.”
For example, when customers started using PSMA Australia services the team received requests for features that needed data not yet available.
Being customer centric is as much about value as it is user experience.
According to Dr Hartig, the big picture is if you have that customer focus, and the value you bring to clients, is an excellent starting point.
“Our business focuses on customer experience and we aim for zero disruption. For example, about 15 per cent of people will give up on a ‘Captcha’ and that is bad if you want to keep customers.”
Customer value can also be improved through automation and machine learning.
To be most effective, ML needs different people from different parts of the organisation to collaborate on the different ways of working, according to Kelcey.
“ML can improve customer experience, because with automation you can take a repetitive task and replace it with a robot,” he said. “You also want augmentation to make people better at their job.”
Time to action data-driven intelligence
Stage five of the continuous intelligence cycle is where the best intelligence is actioned into production.
By planning and prioritising actions and testing hypotheses, you can enact change with greater confidence and at a much faster pace.
As more products and services go digital, they will generate more data, and more data means more potential for intelligence, so it is vitally important to continuously evolve digital assets to avoid data becoming limited in its value.
Building and running intelligent systems is not a one-time thing, but a continuous effort. Digital platforms are not just about upgrading or cost saving, they allow the business to innovate and continuously try new things.
CI shapes the way the business operates and how well it responds to change. It also enhances a company’s ability to deliver value to customers and drive performance.
CI in action
According to Abhayaratna, his organisation has applied its continuous intelligence to improving its services and now has a “platform continuum”.
“We use continuous delivery for continuous customer insight,” he said.
In the case of PSMA Australia, it is building a platform that will enable them to diversify the range and sophistication of spatial data products it provides.
At Arkose Labs, the team applies continuous intelligence to client services, to ensure high levels of customer satisfaction are maintained.
“CI is humans and machine learning working together, with ML doing the heavy lifting and humans providing the oh-so-valuable context,” Hartig said.
At ThoughtWorks, both Davis and Kelcey advise taking small steps with ML to get CI for separate products.
“There will be small incremental benefits in each product,” Kelcey said. “Be continuous with both research and ‘moonshot’ initiatives which will bring a little influence everywhere.”
Next steps to building an agile tech organisation
Building a MDB takes courage and determination and it must be driven by those with ambitious goals for their business, from the top down.
Moving to a data-driven mindset will involve developing strategies across the entire cycle of CI. And developing a culture of “it is alright to fail and learn” is an important first step.
ThoughtWorks’ Davis said start by “bootstrapping” from some system and get an iterative process working.
“When you look at the outside of a digital business, people have a false sense of ‘everything went to plan’,” she said, adding in reality digital businesses try and fail on many fronts.
This sentiment was echoed by Abhayaratna, who said his organisation went through iterations of success and failure to get to where it is.
“Even we have tried to do this and not got it right,” he said. “This time we started small with an experienced, focused team and used that to deliver. A key learning was the use of an evolutionary architecture so we could take incremental steps to achieving our vision.”
He said on-going commitment is important.
“If you set it up as an ‘experiment’ then people don’t really commit to it. Realise and do retrospectives to see what is working and what is not and make it clear to everyone why you are changing.”
Cultural change takes time
Dr Hartig agrees and said making change involves a series of steps you can go on, and it is tempting to try and run before you can walk.
“But if you are talking about cultural change it happens gradually with considerable effort,” he said. “You can move something so long as there is constant pressure. That will then give you the target.”
“The big picture is if you have that focus and value you bring to clients that is an excellent starting point.”
Help from the right partners can also improve an organisation’s ability to change. The CI cycle can be time consuming and fractured, therefore the focus of a partner should be on helping clients speed up this cycle with automation.
Check out the videos, slides and presentations from ThoughtWorks Live 2019 here.
ThoughtWorks is a global software and digital transformation consultancy with 25+ years of experience partnering with clients where technology is a key differentiator of their business strategy. They design and deliver custom software solutions to transform organisations into MDBs. Their diverse and passionate technologist thrive on solving complex problems with an engineering mindset, a culture of innovation, and delivering value quickly.