Category:
Podcast
Series:
#2
Episode:
#7
Date:
April 2023
Stream on Spotify
Back to the beginning.
When we started our journey in 2019, something was missing. We were helping retailers make connected, customer-centric retail decisions across key business areas like marketing, supply chain and pricing. But we didn’t have the right name for our cutting-edge retail software. Was it a personalisation engine? A customer data platform? Nothing quite fit – until we discovered the growing field of decision intelligence. Better still, it described the work we do perfectly.
HyperFinity uses decision intelligence to help retailers unlock the power of their data, create better experiences and make profitable decisions; faster.
In the latest episode of our podcast, CCO Peter Denby chatted to Dr Lorien Pratt, the co-inventor of decision intelligence.
The bigger picture.
The amount of data we’re creating is mind blowing – one stat estimated we produce 2.5 quintillion bytes of data every single day. Businesses are creating their fair share of that data and it’s getting harder to cut through the noise to make informed commercial decisions.
Effective decisions are critical for business success. Despite this, a survey by McKinsey found only 20% of respondents are confident their business excels at decision making.
In the machine learning and artificial intelligence age, there must be a better way.
Turning actions into outcomes.
Decision intelligence (DI) is the discipline of turning actions into outcomes.
Dr Lorien Pratt founded decision engineering in 2010, alongside CTO Mark Zangari. It was renamed decision intelligence in 2012.
Lorien’s a machine learning and decision intelligence consultant, developer, entrepreneur, writer and evangelist. She’s also Chief Scientist and Co-Founder of Quantellia, a software and consulting company offering machine learning and decision intelligence consulting and solutions.
In her previous life Dr Pratt is known as the inventor of transfer learning: an important capability used in most machine learning systems today.
The biggest decisions use the least data.
Modern day workers are inundated with data and information. But it’s not so easy to look at dashboards, piece the data together in our brains, then make good decisions. Despite the rise of machine learning and algorithms, many commercial decisions are made using minimal data, without formal processes in place.
However, computers can take our decision making one step further. Not by simply displaying a dashboard, but by wiring up data to business outcomes and the actions available. Machine learning and AI can take giant data sets, then transparently present the key pieces of data needed for critical, expensive commercial decisions.
Decision intelligence plugs the gap – it’s visual and intuitive, giving senior decision makers greater confidence to make vital decisions.
DI is ubiquitous.
From farming and telecom to healthcare and retail – decision intelligence is ubiquitous.
If a dashboard or spreadsheet makes sense for decision making, ask yourself why. What’s the purpose of that dashboard or spreadsheet? What action will it drive? What business outcome will it lead to? Decision intelligence is more than likely relevant. It means you’re taking that data one step further – to get from actions to outcomes.
What DI is and isn’t.
In any new tech category, people’s interpretation can differ – and the same goes for DI.
Decision intelligence is an integration approach. It takes machine learning models, data and business decision simulations, then connects them in such a way that it drives value. It’s the last mile between data and action.
In the retail world, decision intelligence takes data from a dashboard and answers key commercial decisions. What price should we charge? What new market should we enter? Which SKUs do we need to sell through?
On the flip side, decision intelligence is not prediction. But it does use prediction models to forecast elements you can control – for example, projecting the weather forecast’s likely effect on market demand.
BI vs DI.
Decision intelligence is practical and forward-thinking. Business intelligence (BI) is data from the past or present, so it’s often backward looking, whereas DI is more instructional. For example, if you’ve created some insight, decision intelligence helps you decide what to do with it. What are the decisions you should make?
The one exception to the rule is businesses are increasingly incorporating predictive models into their dashboards, such as weather forecasts or market demand predictions.
It all starts with the humble post-it note.
The benefit of decision intelligence is twofold: (1) organisations save time with data management, and (2) better decisions and better outcomes. Ultimately, DI increases retailers’ bottom lines. So how can organisations get started?
- Leave the data out of the room and start with the specific, measurable outcomes you’d like to achieve. For example, ‘grow revenues by 15% in 24 months.’
- Then, using post-it notes, work collaboratively and think about the unintended consequences – the actions you might not anticipate. It’s easy to get tunnel vision in complex environments: for example, by only seeing two actions, like changing the price or running an advertising campaign. However, there could be other actions you haven’t considered – like trialling a social media influencer, for example.
- Surface all possible actions, then try to draw connections. How does one action lead to another outcome? For example, will running a marketing campaign drive more customers within the context of demand and pricing?
- Once you know the outcomes, actions and decisions you’d like to achieve, you can figure out what data is needed to inform them.
Back to the future.
There’s been a real proliferation of decision intelligence companies and people starting to use DI terminology over recent years. Both Gartner and Alibaba have named decision intelligence a top technology trend in the last two years alone.
The future of DI is bright. Dr Lorien Pratt predicts businesses will eventually create immersive, interactive environments for experimenting with different choices and observing the flow of causal impact. Computers will run various decisions in the background and present those that will be best.
She also anticipates monthly commercial meetings will not purely be backwards facing but will also look at forward facing numbers for planning.
Is your retail business decision intelligence ready? Contact us to find out how DI can improve ROI and customer experiences.
Further reading.
Resource one: ‘Link’ is Dr Lorien Pratt’s first book covering decision intelligence – www.linkthebook.com
Resource two: Dr Lorien Pratt’s second book, ‘The DI Handbook’, will be published this year – www.dihandbook.com
Resource three: ‘Getting Started with Decision Intelligence’ is an eight-week course focusing on decision intelligence by Dr Lorien Pratt – www.gettingstartedwithdi.com