Joan, “lean operation” in telcos typically comes in two principal guises. You either automate your operations as much as possible or you outsource, though I admit that my “wisdom” mostly is based on my experience with European and Middle East operators.
There are different examples for both these cases.
Automation
When you look across the operation life cycle from planning to implementation to actual operation, service providers usually start automating the more repetitive tasks such as configuration and fault/incident management where traditionally most of the resources tend to go.
This ideally requires a sound technology stack i.e. your OSS and underlying network / IT components need to be up to date. In many operators this is not the case, and they are still hanging on to 20 years old mainframe systems in some cases. Planning and implementation aspects are somewhat harder to get “lean” although there are example of “access node factories” in Deutsche Telekom through which most of the installation and configuration is done centrally before shipping kit out to the field.
For a full automation you would have cloudified components and systems available in your network. Telia has started rolling out such a system (4G/5G) earlier this year. Once this is in place you can use cloud technologies in a similar way that Facebook or Google use them when they operate their huge server farms. Automating your network using “cloud technologies” is expected to cut down your operation staff by a factor of 5 to 10! You set it up once, you build up a huge repository of test cases that run automatically and continually, your orchestrator software e.g. ETSI MANO runs all the scripts to configure your network components, reacts to incidents and tries to “self heal”.
You also introduce software updates piece meal in small increments and roll back seamlessly if you run into problems. So, no more lengthy big upgrade projects that eat up a lot of your operation resources. Well, that’s the goal anyway…
Unfortunately, most vendors are not there yet (they have virtualised appliances that do not conform to a microservice architecture required for true cloudification and vendor independence). Also, most (network) operation departments in service providers are pretty clueless as all of the network functions now run as software on commercial IT.
But it doesn’t mean you shouldn’t be moving towards this goal.
Outsourcing
This is a route many (mainly small) network operators have taken over the past decade. There are operators, for example in the Netherlands (I won’t name names here), that are pretty much 100% dependent on their supplier running the show. They don’t have any meaningful (technology) expertise left in their ranks and simply pay the supplier’s headcount at a premium. But they are as lean as it gets.
The same picture holds across most of the Middle East.
I am personally not really in favour of these outsourcing approaches. Operators will not be masters of their own destiny anymore, and all their “innovation initiatives” are by and large empty shells scratching at the surface.
If I was an operator, I would want to be able to introduce innovations quickly and securely myself and not wait until my supplier can write me a piece of software.
This situation, by the way is one reason why operators are lamenting about being “disrupted” because other companies eat their lunch by launching and selling innovative solutions far more quickly than these telco dinosaurs can.