The Digital Twin *Is* The Business Strategy

The era of setting strategy by force of argument and presentation skills is over. Show me the data. In real time.

In my role at Memia I am privileged to work with many organisations developing strategy and plans which look out into the future and inform key operational, investment and divestment decisions now. I really enjoy this work: particularly rewarding are those breakthrough moments when the whole team coalesces around an agreed direction and the way forward becomes crystal clear after a long period of impasse.

Data-based vs. “gut feel” strategy

But too often I observe that businesses operate at low levels of data maturity - only a small proportion use deeply tested data models to underpin operational and strategic decisions - and hardly any use real-time data to continuously improve the models. Instead, significant decisions are made based on a bar graph, a convincing Powerpoint presentation and “gut feel”. 

People have been creating lo-fi predictive models of businesses for centuries - traditional financial budgeting and forecasting techniques are the most ubiquitous - but the one-dimensionality of financial budgeting lacks the precision and predictive reliability needed in today’s complex, changing business environment.

Techniques for working out budgets range from the most basic “last year plus 5%” (still prevalent in 2020!) to more granular models which look deeper at the business metrics and indicators which actually drive financial performance. (Manufacturing output, inventory, marketing spend, customer satisfaction, competitor activity, regulatory changes...). These financial models are often very linear calculations, built up from complex linked spreadsheets or crunched inside monolithic ERPs - and not easily adaptable to rapid or subtle changes in external environment / internal operations. Furthermore, they only take account of a finite number of - mainly financial - data sources and sampling frequencies. At best they’re a rough approximation of what’s really going on.

(I have worked substantially with SaaS businesses where standard industry practice is that the business translates closely to a real time model made up of a diverse array of metrics: demand funnel cohorts, marketing spend, dynamic pricing tiers and discounts, customer churn… all used to calculate key indicators like CAC and LTV - so I’m personally surprised that many businesses still get by without deeply understanding their business at a data level.)

Some of the largest and most successful technology giants including Google, Amazon, Alibaba, Tencent and Facebook, have business models which are fundamentally built upon generating value from the vast quantities of data flowing through their platforms. Furthermore they continuously create “virtuous circles” of improved insights and predictive accuracy from (i) instrumenting new sensors and (ii) collecting more and more data. They are effectively creating a digital twin of their own business and operating environment - and are then able make better operational and strategic decisions based on this. Arguably Apple’s fundamental competitive differentiator is the richly instrumented data model which enables it to operate the world’s best supply chain.

A “digital twin” for your business

“A digital twin is a digital replica of a living or non-living physical entity. By bridging the physical and the virtual world, data is transmitted seamlessly allowing the virtual entity to exist simultaneously with the physical entity."

El Saddik, A. (2018)

Digital twins are not new. The concept has existed for nearly 3 decades and already has many live applications in industry - accurately modelling physical systems in real time: wind turbines, manufacturing, asset monitoring and even the built environment.

In the same way, a digital twin of a living business promises an order-of-magnitude step up from traditional financial and planning processes: creating an accurate real-time digital simulation which updates and changes as its real-world counterpart changes. Furthermore, the digital twin continuously updates itself by learning from multiple data sources, not just financials - representing the business in real-time as accurately as possible. 

Modern business “digital twins” promise to deliver three qualitatively different capabilities which differentiate them from traditional business information tools:

  • Near real-time accuracy - by instrumenting the model with real time data streams, management is able to get information and decision support with far greater accuracy and timeliness

  • Positive data feedback loops - the digital twin fundamentally strives to provide as accurate a picture of the real world business as possible. To do this it needs ongoing investment to receive internal and external operational environment data which may affect the business’ performance. 

  • Optimisation scenario modeling - once the digital twin is operational and delivering an accurate representation of the real-world business, its true power becomes apparent. Machine learning algorithms can be used on the twin to model scenarios and optimise the business for certain outcomes - these will usually be profitability, but increasingly other dimensions such as social and environmental outcomes.

“Digital twins are not developed in a vacuum. Both the business concept and model must be tested against an economic architecture – revenue, profits, return on investment (ROI), cost optimization – and a way to measure progress as the products/services are rolling out.” Gartner, Prepare for the Impact of Digital Twins

Constructing a Business Digital Twin

Constructing a digital twin of a business is a long term, iterative process. Each iteration cycle can be broken down into four distinct stages:

  • Identify metrics and indicators

  • Identify data sources

  • Construct model

  • Operate and optimise

Identifying the metrics is the opportunity to really put the organisation’s balanced scorecard front and centre - businesses now and in future need to be optimised not just for financial performance but also for social and environmental considerations. Metrics need to be designed to measure environmental impacts such as carbon emissions as much as the P&L.

When identifying data which affects your business it’s important that these are not internal - external data sources such as weather, exchange rates, interest rates, social media coverage and traffic patterns can all be used as inputs to the model.

After a few iterations the digital twin model should begin to approximate the operations of the various parts of the business with varying degrees of precision: production, inventory, sales, distribution and marketing are all good targets to focus on. Over time, the digital twin model becomes the primary tool for predicting demand and optimising operations across the business. (Bearing in mind that no algorithmic or data-based predictive model is ever infallible!)

The digital twin *is* the business strategy 

Internalizing these data-based, continuous decision making capabilities going forward - at a strategic level, not just operations - is a fundamental competitive advantage. In fact, from here on, the digital twin *is* the business strategy.

Horizon 1 strategy generally becomes just an optimisation exercise for the model of the existing operational model: run enough scenarios through the algorithm to optimise for outcomes based upon current business parameters.

But for Horizon 2 and 3 strategy, the model needs to be continually expanded. Traditional business strategy techniques such as Porter’s 5 Forces or PESTLE analysis can be readily incorporated into the Digital Twin as extra dimensions to the model. Strategic hypotheses need to be modelled out and tested with data simulations before actually committing to investment decisions. Other dimensions can be added as well to the digital twin as well - for example detailed competitor behaviour and environmental impact modelling.

In short, business strategy moves from an occasional, mostly manual spreadsheets-and-powerpoint exercise to a continuous detailed modelling and simulation capability built and executed on a company’s digital twin platform. The business digital twin becomes the master predictive tool for strategic investment and divestment decisions as well, leveraging the underlying model as the key competitive asset rather than conventional gut feel techniques that are in use today.

The era of setting strategy by force of argument or presentation skills is over. Show me the data. In real time.