The 7 things successful automation teams do to scale

By 2030, automation will drive 75 to 375 million people to reskill and even
change occupation – do you agree it’s time to upskill your workforce?

By 2025 97% of business will have started using Automation

By 2030, automation will drive 75 to 375 million people to reskill and even change occupation – do you agree it’s time to upskill your workforce on how to adapt and adopt these new technologies?

Over the last 5+ years we’ve communicated regularly with our international Automation and AI network, we’ve also observed, interviewed, worked with, worked for, shaped, trained, learnt from and researched teams globally in Europe, US, India, Hong Kong, China and the UK

We’ve extensively researched the automation industry and seen it evolve from when Robotic Process Automation (RPA) was brand new in the mainstream in 2017, to how it has become intelligent automation. Our research has been focused on two metrics that have remained unchanged all this time, globally! We asked:

Why do 50% of all RPA and AI projects fail?

What were they doing right, what was everyone else (even big companies and large automation teams) doing wrong? We started to see patterns, repeating cause-and-effect situations and coming from a Lean-Six Sigma world, the instinct was to understand the root-cause.

Why do less than 5% of businesses succeed at scaling automation?

We discovered that each automation team or business had their unique (and sometimes not-so-unique) issues and problems, but they all tended to stem from an array of similar challenges that they had all come up against (see the Top 7 challenges RPA teams will face). After reviews and tracing we identified that most of these challenges could be traced back to a handful of mistakes that had been made and in turn we were finally able to boil these down to just THREE root causes (see the 5 key mistakes of automation and AI).

These 3 reasons seemed to be the cause of all the issues and problems that was causing all these Automation teams to not get value and fail at scaling their automation!


By understanding these root causes and mistakes, we were able to notice the patterns of what successful teams were doing and time and time again saw SEVEN themes in successful teams, and saw a lack of these 7 things in struggling teams.

Here are the 7 things that successful automation teams do to roll out automation at speed and at scale which we put into the acroymn: A. E. I. O. Y. O. U. and the AEIO YOU methodology which we have explained in our book Business @ the Speed of Bots (the AEIO YOU method)

The AEIO YOU methodology

Based on the 7 things successful automation teams do, AEIO YOU is a unique blend of Lean Thinking, Change Management, Intelligent Automation best practices to maximise ROI.

This 7-stage process (and its 36 repeatable steps) can be used to adopt and roll out new cutting-edge technologies across an entire business, to improve Automation project success (current industry success rate is 50%). This methodology ‘changes how we change’ with technology so that the entire company is involved and comes along the digital transformation journey, vs having change ‘enforced’ on them


this is the Awareness and alignment stage, where a company becomes aware of the new technology from a senior level and learn about how this capability fits into the overall strategy. However, an important step sometimes missed in organisations is to also make staff teams aware of the technology, how it will help them personally as well as has the business will be strengthened.

This is also the stage which aligns the different key business areas to agree on how to use this technology, how this new capability will be governed (in a centre of excellence) and where this CoE will sit (e.g. in operations, IT or perhaps in the business change department.


Once a Automation roadmap and operating model has been established for the RPA CoE team, the staff should be educated in the new ways of working. This is your core RPA/Automation team, Automation Champions and staff who will be helping with the automation. They should also be educated in their part in the CoE, what is expected of them, how to use tools and techniques

Coupled with this, is the provision of standardised tools from the CoE to empower core CoE staff and stakeholders to be part of the digital transformation. This will ensure that your teams produce successful results in their automation projects repeatedly


This is the application of tools and training to identify and accurately assess opportunities. This is also where your teams follow strict guidelines to design (ideate) the right solutions


This stage in the implicating of the optimised solution, be that RPA, intelligent automation or a lean intelligent automation


This is a way to measure return on investment or yield and prove the benefits were realised as predicted


This stage focused on the support side. As your virtual workforce grows you will need to re-organise and develop your support team with role profiles and procedures which were needs when your bot factory was small, so that you can better oversee your bots.


This final stage is a reflection of the first and second stages, where you become discover of new technology, upskill your team to match the new technology, and upgrade your automated processes which AI

Tony writes for the Digital Transformation network, educating businesses in using new cutting-edge technologies.

To learn more about using Process automation and other technologies, check out the AEIO YOU® Portal, a one-stop centre of excellence enablement platform for getting value from RPA and IA – speak to your local AEIO YOU® partner to learn more


5 Levels of CoE maturity: How mature is yours?

There are several levels of maturity that I’ve seen. Which one are you?

Level 0 (Explore stage):

You’ve just started looking into RPA and automation, you’ve not deployed anything. You’ve seen a few demos but have not yet decided which vendor(s) to use.

Level 1 (Experiment stage):

Your team have self-built automation robots using a free license to understand how the technology works and/or you’ve engaged with preferred vendors who may have given you some licenses for free. No real structured approach yet for assessing or implementing opportunities.

Alternatively, if you have outsourced your CoE completely then your maturity would also be Level 1 as you have not yet developed any in-house capabilities.

Level 2 (Pilot stage):

You’ve hired a hybrid RPA developer who also does the business analysis work to find opportunities to automate, and you’ve deployed a POC (proof of concept). Some RPA documentation and assessment tools exist, and you have a logical approach to assessing opportunities.

Level 3 (Team formation stage):

You have a few expert team members doing some of the roles that make up a CoE team but not a complete structure; a solutions architect, developers, support engineers, RPA analysts, PMs and perhaps a Lean Analyst. You have documented processes, a framework and a toolkit to generate consistent automation solutions.

Level 4 (RPA factory):

A polished team of experienced professionals who have delivered automation in several environments, as well as automation champions throughout the organization as advocates. A steady pipeline and a refined framework to churn out automated solutions at a consistent quality and speed. A complete toolkit with controls and reporting tools.

Level 5 (CoE as a Service):

Experienced team able to provide their services to internal and external clients. Clear terms of service, SLAs and governance processes.

If you have a large organization, you have probably set up several satellite RPA teams that report to you.

There are 2 main types of CoEs. Centralized or Federated, however, the maturity of your CoE team and the geographical location of operations teams may determine which structure is right for you.


This is where you generally should start. Have one team that manages all RPA opportunities happening throughout the organization. This is good for companies where the CoE sits together and all staff are based in the same location, or teams and business units can be easily reached.


This is where you have resources embedded in different business units however they still feed into and draw from a central team for training, tools, knowledge management and governance

Whatever you do, ensure that you don’t have different teams running RPA projects in different ways, learning different lessons and using different RPA vendors – this will only lead in disaster

You can learn more about implementing RPA (Robotic Process Automation) and intelligent automation at The Lean Intelligent Automation consultant

You can read about Lean IA’s trademarked AEIO YOU method in their new book Business @ the Speed of Bots, Succeeding In The New Age Of Digital Transformation


5 Steps To Build Your Enterprise Automation Road Map (EAR)

1. Zoom Out:

a. Identify the departments most suitable for Robotic Process Automation (RPA) using an RPA Checklist

b. Solution layer: review and vet capabilities and vendors that can meet individual team needs (RPA, AI, Machine Learning, intelligent character recognition, process mining, business process management etc)

c. Build a Department or Team Complexity Map to show ease of implementation vs benefits (financial and non-financial)

2. Zoom In:

a. identify all suitable processes and add to your Automation Catalogue

b. Solutions layer: review and vet capabilities and vendors that can meet individual process needs (RPA, AI, outsourcing, lean process re-engineering, etc.)

i. ability to create libraries of re-usable code

ii. ability to scrape web pages, Citrix applications, PDFs

iii. Ability to integrate with plug-ins and APIs

c. Build a Process Complexity Map for your first target department

3. Categorize and Prioritise

a. Prioritize teams and processes into waves using your high and low-level complexity maps

Using a Complexity Matrix you can capture the difficulty of each automation initiative from the developer’s perspective. By quantifying the complexity and intangible benefits such as scalability and error rates, you can plot all processes on one map.

However, doing all quick wins upfront may not have significant impact on the business, so it is ultimately the business who should decide, guided by the organization’s strategy and using this map as a decision aid which backed by data4. Bake in continuous improvement

a. Keep coming back to implemented automation in order to improve or enhance them

5. Plan how to self-fund your CoE

a. Use savings generated from implemented processes to pay for the continued running on your CoE. That means get profitable fast, be that ROI from clients or wooden dollars internally

You can learn more about implementing RPA (Robotic Process Automation) and intelligent automation at The Lean Intelligent Automation consultant

You can read about Lean IA’s trademarked AEIO YOU method in their new book Business @ the Speed of Bots, Succeeding In The New Age Of Digital Transformation


Why AI means the end of ‘pure’ RPA

Interview with Process Excellence Network

PEX Network:

What is your working definition of ‘intelligent automation’?

Antony Walker:

RPA is dumb computer software doing manual tasks. Intelligent automation is combining things like RPA or BPM tools with some sort of artificial intelligence, so that it doesn’t just move data through a process, but it can interpret and understand data. Unstructured data is where RPA falls short, whereas AI can fill this gap. And 80 to 90 per cent of processes have unstructured data

PEX Network:

What problems can intelligent automation be brought in to fix?

Antony Walker:

RPA, AI and intelligent automation is seen in the media as something that could replace people as many businesses have been focusing on FTE savings, however it should be used to augment what people do; taking the tedious boring tasks so that people can focus on more intellectual, creative parts of their life. And it allows businesses and their staff to discover new revenue streams that were once unviable to reach, so that their business can in fact expand.

PEX Network:

What sort of organizations are looking into RPA and intelligent automation, and why are they doing it now?

Antony Walker:

Early adopters of this technology have been finance, insurance, utilities, healthcare and law firms. The manufacturing industry have used physical automation for decades and are also looking at automation on the software side as well. The biggest benefit are for industries that have a lot of paperwork and repetitive tasks, so you can imagine areas like legal, tax and accounting departments having a big appetite for this. Processes that are repetitive and logic-based can be automated, so this is why RPA has benefited them greatly.

Over the next year, there’s going to be a lot more focus on AI. In 18 months to two years, I don’t see there being any pure RPA initiatives, everything is going to have that intelligence built in. This is because IA is really starting to mature now.

PEX Network:

That is quite a statement. What do you think is driving all this?

Antony Walker:

There have been a lot of advancements in AI. There are a lot of use cases and quick wins for teams to implement AI and benefit from it straight away. And as mentioned previously only 10 to 20 per cent of processes use structured data. Many business leaders and analysts I’ve spoken to have become frustrated in their previous pure-RPA projects that weren’t providing the ROIs they had expected. That’s why lean IA hits this problem twice. Lean means you use fewer ‘robots’ to get more savings by optimising your processes first, and IA means you can target more processes and move away from just pure-RPA.

PEX Network:

How will these changes affect the big legacy banks and financial institutions?

Antony Walker:

Two big stories that have been in the finance technology world for a while now are robo-advisors and cryptocurrency. Independent financial advisors will start to feel squeezed out. They are being challenged by these AI capabilities advising people on how to best invest their money, so IFAs are having to adapt and focus more heavily on the human, customer relationship side. Then we have cryptocurrency which threatens to decentralize the monetary system.

One thing I have seen in the utilities industry are start-up companies in the energy sector that have popped up with very low overheads due to having minimal infrastructure, and are able to greatly undercut and move faster than the big corporations. This is definitely something legacy banks may fear happening to them.

Any industry that has a lot of paperwork, manual processes, large workforces and unnecessarily complicated or dated processes in their back office operations will feel the pinch as they start haemorrhaging money and customers. In asset management the baby boomers still prefer using paper forms, I’ve been working on a project where we used intelligent cognitive-recognition software to scan information from a paper form and then put this directly into the database, which saved staff having to read the document and manually typing the information into the database manually a thousand times. Do note that we are now at a time where computers can read handwriting more accurately than humans.

These are just a few of the things behind the scenes that AI and intelligent automation are having an impact on.

PEX Network:

If this technology really is maturing and becoming more ubiquitous, what’s going to be the big difference that we’ll notice?

Antony Walker:

How we interact with the banks and insurance will be a big one. Natural Language Processing is going to advance a lot more, and is where I am personally starting to focus. Website chat bots will become a lot smoother and smarter; they will understand what we’re after, and understand what our profile looks like; they have our data, so they will be able to really customize what their recommendations are to our specific needs.

In insurance situations like a car accident when we use our phones to make a claim, they will be able to read and respond to our moods, understand the images in photos we send them, know where we are and remember our history to tailor their advice

In summary, customization, speed and accuracy of service through more convenient avenues are how we’ll notice these technology advancements over the next few years.

Read the book, Business at the speed of bots