How to de-risk your fibre and 5G rollout programs

How to deliver better program outcomes using data analytics and AI.

In April 2020, Mike Hosie, Partner at Mentor, spoke to Total Telecom on how telco’s can take advantage of the latest AI in the rollout of their fibre and 5G programs

Given the vast amount of data that a telecoms operator collects every day, harnessing AI is becoming a growing feature of a modern, digital telco.

“Our industry is increasingly using digital transformation to help make better use of data in addressing customers’ needs and companies have been using these technologies to deliver new revenues for years now – so why not use them in program delivery?” asks Hosie.

Watch the full interview below!


Hello everyone, Harry Baldock here, editor at Total Telecom.

The pace of evolution across the telecoms industry is rapidly accelerating and operators have no choice but to transform if they’re to remain relevant to both consumer and enterprise customers as well as achieving steady growth to remain current effective and dynamic in a market characterized by constant change.

Telecoms companies need to deliver their strategic programs without delay the first time around and to do this they are going to need to get the most out of people processes and tools in order to safeguard their competitive advantage.

In November last year I spoke with Ian Waters, Partner at Mentor, a team of program management experts working in the UK telecoms industry for over 30 years, about a new program assurance tool that aims to de-risk large and complex business critical programs. 

Today I’m talking with Mike Hosie, Partner at Mentor, about how companies today are seeing the benefits of using the latest technology in AI and data science, in the delivery of their 5G and fibre rollout programs. Mike thank you very much for joining me today.

Mike Hosie: Thank you.

Harry Baldock: How can telcos use data analytics and artificial intelligence to aid in the deployment of their fibre and 5G programs?

Mike Hosie: These business-critical programs have lots of moving parts and lots of contributions. Finding ways to make the execution process more effective and more efficient, more certain, is a primary goal of good program management.

More often than not, it does require a mix of people, process and technology and establishing a robust methodology from the start means that program managers will be able to produce much more reliable outcomes. 

Now artificial intelligence is increasingly finding its way into project management tools, handling everything from scheduling to analysing working patterns of a team and offering suggestions for improvement. These tools augmented by AI are producing obvious benefits for project and program managers today. For example, they automate repetitive low value tasks, they use historical data to perform calculations that improve predictions and accuracy of results, they perform modelling and analysis tasks to help manage changes in scope time and costs, and they can improve the speed of decision making through the application of rules-based monitoring alerting and reporting. 

Whilst they are primarily task oriented and process driven, they do help to free up time for program managers to spend time on problem solving and other higher value management contributions.

Our industry is increasingly using digital transformation to help make better use of data in addressing customers needs and companies have been leveraging these technologies to deliver new revenues for years now, so why not use them in program delivery?

The proportion of projects that are managed using AI is expected to jump from around 20 to 40% in the next three years alone. At Mentor, we believe it is the right time to start thinking about how organizations can really leverage AI technologies to ensure project and program success. 

Lots of program data sits with employees working directly in the program, so let’s use that data to deliver better program outcomes. We’ve taken the use of AI a step further by incorporating it into our methodology for program management, and are now helping clients deliver their programs much more successfully. 

Harry Baldock: With so many program management methodologies already out there why is a different approach needed?

Even with the increasing number of methodologies available, professional certifications, tools and the freely available checklists, we can still see a number of unacceptably high number of programs failing to deliver what they were set up to achieve. Over 70 percent of programs still fail to deliver. Now that’s not to say that the methodologies and the tools aren’t valuable they are – but sometimes things are missing. 

It’s clear that things are missing and as I said, the majority of program management methods are task oriented and process driven, that provide road maps and checklists that help to get the program completed. Traditional project management primarily focuses on tasks scheduling and technical planning, and AI will automate a lot of this, helping shift the program manager’s focus towards a better understanding and management of their program.

At Mentor, we are now using AI to go beyond process and tools, using it to apply knowledge expertise and experience. We built the Mentor Execution Index [MEI] to combine AI with our knowledge and experience and it’s the practical application of these things that differentiates the MEI from the more commonly used program methodologies and tools.

Let me let me try and put that into context for you, we are typically asked to look at business critical programs that have got into difficulty, or are failing. We have over a hundred of these under our belts and historically our starting point, which is fairly typical, would be to carry out interviews. We’d analyse program documents and then come to a view on the state of a program and what needs to be done to get it back on track.

Using the old adage that it’s hard to improve what isn’t measured, we wanted to go further and properly quantify the state of any complex program. We wanted an objective assessment that would avoid any unproductive debates on judgment calls and then provide a clear mandate for action that can be implemented quickly. 

My personal view is that program execution is a neglected management discipline. Program management is dynamic – it’s as much about people, what they do, how they’re organized, how they interact, and how they behave, as it is to do with technologies and markets. These things can be measured using modern survey and data analysis techniques.

The MEI takes our experience and combines it with the latest AI, allowing us to quickly assess a program’s health. It provides a level of insight that leads to interventions and improvements that are simply not available with the more traditional methods of program assurance. It moves us away from viewing a program’s health based on an initial diagnosis much more towards an x-ray that confirms the diagnosis and then helps to determine the right treatment to secure a program’s overall health.

Data science has been around for quite a long time and anyone using Excel to look at data and get insights to allow better predictions or decisions, they’re already practicing data science. In the past, like most organizations, that’s exactly what we would have done. We would have used Excel to analyse survey results using simple or weighted averages to score a program, and then doing regression analysis to try and identify any patterns. Now AI and modern statistical modelling allow us to do something much more powerful.

AI is using machines to simulate and enhance human intelligence. We use specifically designed algorithms and other advanced analytical techniques such as machine learning – a branch of AI – where computers use statistical and mathematical models to work out on its own and to learn through parameters, weightings, and predictions. The machine is learning and spotting patterns and relationships far better than we as humans can. When looking at the same data we use deep learning which is the latest highly complex iteration of machine learning, which works using billions of parameters – literally billions of parameters – much more complex processing required but it does produce much better predictions.

To summarize, the MEI is not a program management tool, it provides an end-to-end assessment of a complex program with input from the program team, including its customers and suppliers. It looks at what is driving failure. It’s not just a bit of technology that automates tasks and it’s designed importantly to take account of human behaviour, driven out of a robust framework and the expertise that has been tried and tested in over 100 successful rescues. We believe it’s a unique approach to achieving successful program management.

Harry Baldock: What type of results can a company expect to see using the Mentor Execution Index?

How many programs have you seen that are months late in delivery or millions of pounds over budget? It’s a common story. We believe the MEI is a cost-effective and quick way for businesses to de-risk their programs and make sure they stay on track to deliver. 

We’ve been using the Mentor Execution Index for some months now with several clients, and they’re already seeing significant benefits. A properly structured and well-run program will help to make sure that the benefits being sought are ultimately delivered. 

The areas where I would expect to see results include, helping to manage risk is really important in any business-critical program. The MEI will flag issues that are developing, it will provide lead measures rather than the old traditional lag measures which report on what has already happened, when it’s often too late to alter course.

We can also benchmark between programs or benchmark programs periodically at different points in their life cycle. So, we can track progress and monitor the impact of changes or interventions that were put in place to fix problems discovered in previous surveys. 

More tangibly perhaps in terms of efficiency, basically delivering more for less or indeed an absolute increase in delivery. For example, better capital delivery performance, an increase in capex delivery per head, and both of these lead to increases in profit and better or more improved returns on investment.

Finally at a lower perhaps more detailed level, the benefits I would expect to see for the fibre and 5G rollouts that are happening on our streets today, would be increased speed of deployment, faster rollouts, improved quality of deployment, more project gates being passed as they were planned, more effective organizations, for example better program management, better PMOs. Better program organizations will ultimately lead to fewer delivery dates being missed.

And then there’s a development of people, increasing their capabilities, encouraging greater engagement, greater motivation to succeed. I mean who wouldn’t want to work in a thriving and successful program environment. 

And then finally for me, increased customer satisfaction from all of the above. As I say for me program management is actually all about meeting commitments.

Harry Baldock: Mike it’s been really great to speak to you about this really important topic. Thank you very much for your time.

AI and Program Management

The proportion of projects managed using AI is expected to jump from 20% to 40% in the next three years.

“Artificial intelligence is increasingly finding its way into project management tools, handling everything from scheduling to analysing working patterns for a team and offering suggestions for improvement,” said Hosie, thus allowing program managers to instead focus on high impact contributions to the program, such as problem-solving.

The Mentor Execution Index seeks to blend the human experience of those contributing to a program with the precise insights gained by AI and deep learning. As part of our HealthCheck service, the tool creates an objective view of the program, removing commonplace yet unproductive debates over judgment calls and subjective decisions.

The Mentor Execution Index creates a clear mandate for future action and delivers something akin to an x-ray that confirms the diagnosis and then with Mentor’s program management expertise recommends the necessary steps to improve that program’s health.

Program failure is incredibly costly – the Mentor Execution Index’s strength is that it can deliver AI and human-driven insights during the program’s life cycle, helping companies to avoid costly delays.

With so much of the telecoms industry-focussed right now on their 5G and fibre rollouts, programs reaching completion smoothly has never been more important.

“The benefits I would expect to see for the fibre and 5G rollouts that are happening on our streets today would be an increased speed of deployment, improved quality of deployment, more project gates being passed as they were planned, and better program management that will ultimately lead to fewer delivery dates being missed”.

Watch the full interview with Mike at the link above.

Further information:

Learn more about The Mentor Execution Index:

Try The Mentor Execution Index for yourself: 

What do AI and smart data analytics have to do with program execution? watch now


This video first appeared on Total Telecom