A company’s enterprise automation journey often sprouts from a single project. For Florian Mihai, head of marketing technology infrastructure at HP, it started with integrating software-as-a-service tools in their MarTech stack. As Mihai’s team began automating marketing operations tasks, they recognized a broader opportunity. Today, their marketing processes are some of the most advanced you will find anywhere.
At the beginning, their questions were straightforward and aimed at identifying, for example, how to connect two apps together or reduce data entry. But over time, their questions evolved and they sought to understand other elements such as how every step in the process is connected end-to-end, how machine learning (ML) can make decisions for them and if the ML model could autonomously drive revenue.
Notice the delta between these groups of questions. The first set is simple and straightforward while the second set is more complex and innovative. Companies can only begin asking the second set of questions after the first are answered. That progression is a natural flow in technology we call enterprise automation maturity and it is worth a closer look.
Technology does not confer maturity
Technology does not drive automation maturity, people do. Maturity describes the progress an organization has made with technology. Just like owning the keys to a shiny new car does not indicate a mature driver, although the average 16-year-old may think it does, buying the latest technology does not make an enterprise more mature in their strategy.
A helpful way to measure this progress is with a theoretical maturity model. They are popular and can be a helpful benchmark, like a guide or GPS. They are most valuable in noisy markets awash with vendor marketing and analyst opinions, but not all models are created equal. If you Google “automation maturity model” you will find limitless options from vendors.
Martin Fowler notes: “The true outcome of a maturity model assessment is not what level you are but the list of things you need to work on to improve. Any maturity model, like any model, is a simplification: wrong but hopefully useful.”
Each model is different, but unfortunately some automation models are not useful at all.
In the automation space, many maturity models fall short. Criteria to sort through models include whether the model shows:
- People, process and technology coming together as the organization matures
- A technology-agnostic approach
- A sufficiently long-term view of the future
- No acceptance of shadow IT
Let us take a closer look at what a valuable model looks like.
The enterprise automation maturity model
After years of studying the outcomes that businesses achieved in automation, a picture emerged of six different stages that I call the enterprise automation maturity model. It is the most comprehensive view I have found of the technologies and strategies employed today. The stages are:|
- Tasks: Replacing repetitive work, typically done by humans, with bots
- Business functions: Automating a singular function in the business, such as integration
- Apps: Automating across a wide collection of apps
- Business processes: Automating an entire process
- Business decisions: Bringing artificial intelligence (AI) and ML in to automate decisions and reduce human-in-the-loop exception handling
- Cognitive: New business models or revenue driven by AI
In our experience, companies will fall into one of the six stages, but the majority are still in stages one to three. You can visualize this as an adoption curve, and that curve shows where competitive differentiation can be found. While most languish in the early stages, the top performers are way ahead and there is often a direct correlation with how much market share a company captures.
Over the next 10 years, the winners in the market will be those that push to steps four, five and six. Their investments in automation will directly lead to top performance in areas such as customer experience, employee experience and supplier ecosystem. In light of this, let us take a closer look at what is specifically involved in the latter three stages of this model.
Maturity stage four: Automating business processes
Automating end-to-end business processes that span multiple business functions, units, teams, systems and apps is no small feat. Achieving stage four of this maturity model means the entire C-suite is bought into the strategy and sustains an automation culture. A company at this stage may have between 200 and500 processes automated and the percentage left unautomated is low.
Typical examples are procure-to-pay, hire-to-retire, invoice processing and accounts payable automation. Fragmentation of our applications, people and data should be almost non-existent at this point as the major cross-organization processes are brought into unified flows.
Maturity stage five: Automating business decisions
Once organizations reach stage five, they can begin harnessing AI and ML to drive business decisions. Drawing on AI, ML, rules engines and natural language processing, time-sensitive and mission-critical decisions can be made by the machine to improve automated processes.
At this point of maturity the business comes to you, and companies begin managing the business by exceptions rather than tasks. Examples include threshold-driven inventory ordering, fraud detection and notification and customer offers augmented with behavior intelligence. To achieve this level of scale, an enterprise will have between 500 and 1,000 processes automated.
Level 6: Cognitive automation
The highest level of maturity in enterprise automation is driven by the core concept of autonomous business and it focuses on self-learnt and prescriptive decisions. A good example is automated traders that perform and adapt to high-frequency trading scenarios, although this only takes place at an application level now, rather than cognitively across the enterprise.
Still, just imagine all business functions performing like automated traders—while still aspirational at this juncture; this is what some enterprises are currently working toward. In this case, automation will be able to drive revenue directly and uncover greenfield opportunities for growth.
Automation is now more important than ever. As the economy shifts into new territory, companies are going to be looking at automating more tasks and processes. This is not just for the stereotypical reasons one might expect such as cost savings and efficiency, but for others too. IDC reports that companies are seeking revenue, sustainability, productivity and profit.
[inlinead-1]