The evolution of (IT) managed services

9 min readJul 2, 2021

A managed service should focus on value being created by increasing the adoption of shareable resources that are not consumed in use, but instead gain value when they are used.

It’s all about making the entire ecosystem work. In the classic setup it is difficult to get anything done because of friction by processes slowing the system down and poor quality resulting in unhappy users and way too high support costs.

In many organizations services that would benefit from being managed collaboratively in a scope[1,2] economy are actually being managed through a services provider, using processes primarily suited to prevent overconsumption.

Resources that require collaboration and continuous improvement, are hidden behind multiple approval and ticketing processes.

We believe that by identifying and shifting these resources out of these scale based governance, organizations can unleash the value, reduce the toil involved in accessing them and create a shared responsibility for continuous improvement.

Introduction

“Managed services is the practice of outsourcing the responsibility for maintaining, and anticipating need for, a range of processes and functions in order to improve operations and cut expenses.”[3]
- Wikipedia

The above definition leaves plenty of room for interpretation. Both the classic model of managed services as well as the modern model can fit the description. At the core, the success of the managed service should be measured by the success of the teams using the service. Looking at existing managed services from a user experience point of view immediately draws attention to some issues with the classical model.

Revisiting the definition, we look at the element of outsourcing the responsibility. We can look at outsourcing from a team point of view or from an organizational point of view. In this document we will focus on the organizational point of view first and flow into the team point of view, the internal platform, later. Successful organizations will adopt both. They will also adopt both in a modern setup.

Classic managed services

Classic managed services utilize cumbersome ticketing processes to support charging and they suffocate innovation through over-standardization. Outsourcing the responsibility for delivering a managed service to another organization often (still) results in this classic approach, where the main focus is the cost savings perspective while in the meantime providing its users with an inadequate as well as unsatisfactory experience. In the field, this is also referred to as a “take my mess for less”-deal.

Most outsourcing is done for cost reasons, optimization of budgets, reduction of headcount (as system integrators have many people shared across customers). The one thing where most outsourcing deals fail is evolution. Classic managed services are all about preserving what you take over from the customer with the normal improvements, but little to no evolution.

Most agreements do have what is called “innovation budget” and some text around platform evolution, but the reality is that it’s all about (only) improvements. Adding a big new service, development of such service in a complicated subsystem approach, would be a separate deal, which really doesn’t incentivize evolution. So outsourcing becomes more about preserving the status quo.

From the managed services party operational excellence is expected and parties agree on the desired service level through a strict SLA. Clauses for not meeting the SLA are embedded to prevent the delivering party from deviating from the agreement. As mentioned earlier, because of the setup of the agreement the incentive for the managed service provider is doing the same with less. “How can we deliver the bare minimum service, securing the service level, with minimum costs and effort?” So in fact, because of this model, the party delivering the service is incentivized to actually do less with less, including 0 (zero) innovation.

Because of this race-to-the-bottom setup, and strict SLAs with no responsibility for value delivery to the users of the service, two different economies emerge; the economy of scale and the economy of differentiation. These immediately start colliding. Classic managed services are set up as economies of scale, driving operational excellence, and protecting scarce resources from overconsumption instead of enabling a scope economy, and therefore value is wasted. Scope economies create efficiency by offering variety, meaning the support of the teams in the differentiation economy by providing them with relevant, valuable, reusable capabilities. Enablement of a scope economy accelerates the value creation of the economy of differentiation meaning the teams using the service who are working to accelerate the delivery of value for their customers.

Within the (traditional) economy of scale approach the mentioned protection is secured through a ticketing system and complex flows. While automation of workflows would actually support service providers to do less with less, often very little automation is implemented. This leads to many levels of manual intervention and approval steps, resulting in wait times and a degraded experience. This form of managed service does not operate with the user of the platform in mind. As we learned earlier, the success at the core of the platform should be measured as the success of the teams using the service.

Even worse; The platform slows down the speed of the teams operating on the platform. The production process slows down and innovation slows down.

Losing the connection between the users and the delivery of the service is dangerous and the service is at risk of becoming irrelevant very fast. Procedures like a request for change exist but often these are complex and time consuming, adding to the dissatisfaction of the users. All of the above occurs because of an over-focus on operational excellence and a lack of understanding of, and willingness to invest in the users of the service.

Modern managed services

Modern managed services are value-driven and operate using a relational agreement with the consuming party based on trust and transparency. They are designed with frictionless user experience in mind supporting the flow of the teams, allowing them to increase cadence and speed. A modern service enables teams to deliver work autonomously with self-service capabilities, in order to reduce extraneous cognitive load on the teams. These services are positioned to support the innovation excellence happening within the customer organization, meaning supporting the aforementioned economy of differentiation. Modern services, like their internal (to the organization) relatives, are to be operated as digital platforms.

“A digital platform is a foundation of self-service API’s, services, knowledge and support which are arranged as a compelling product.”[4]
- Evan Bottcher, 2018

It’s critical that these digital platforms are led, architected, managed, developed and operated as products. New platform features flow through the same pipeline models as any other code or software release would. Managed service providers are expected to develop a balanced set of capabilities in the areas of leadership, architecture, product management, development and operations to effectively deliver an appealing product and adopt a way of working that supports continuous improvement. Through short and closed feedback loops the managed service team can execute continuous discovery as well as continuous delivery to maximize the experience of the teams using their service.

The amount of self-service functionality available on the platform is a key measure of platform maturity. The platform (the managed service) must provide services that do not require the teams using the managed service to raise support tickets. This will improve user experience by removing friction, allowing speed as well as the cadence to increase, and support flow.

“Platforms must be designed with the user in mind [..] focus on UX and DevEx.”[5]
- Matthew Skelton, 2021

Supporting the teams’ -using the platform- success is a core responsibility for the managed service provider. The managed service platform must support its users to achieve their goals by understanding needs and simplifying tasks.

“The platform should not increase the cognitive load on teams using the platform.”[5]
- Matthew Skelton, 2021

Combining the above we see a heavy focus on user experience. Frictionless, speed and flow are key. This means for users of the service to be able to focus on what is important: The creation of business value for the organisation by creating value for the end-customer. A managed service should be incredibly simple to use, any barrier for using the service should be removed, this will create the much needed “pull effect” of the platform service, making it appealing for teams to use.

Modern platform management

Site Reliability Engineering (SRE) is a discipline that incorporates aspects of software engineering and applies them to infrastructure and operations problems. Balancing market agility meaning allowing the teams to go fast, and reliability meaning securing the stability and availability of the services is fundamental. Key with SRE when talking about how a service is led, architected, managed, developed and operated are the service level indicators, service level objectives and the resulting service level agreement accompanied by the error budget and relevant metrics for measuring business performance.

Modern platforms measure business performance using the following metrics:

Product — These are the Accelerate[6] metrics related to balancing market agility and reliability.
Team/user satisfaction — Net Promoter Score, understanding what works
Adoption and engagement — % teams onboarded, platform services used
Reliability — Service level objectives, latency, # of incidents, etc.

Service level indicators (SLI) are a defined measurement of an aspect of a service. For example availability, latency and throughput when thinking about user-facing systems.

The service level objectives (SLO) are a target or a range of values as measured by the SLI. The design of the SLOs are a team effort and thus a collaboration of the managed service provider, the teams using the service as well as the business for whom they develop their products.

The service level agreement (SLA) is a known construct. Being an explicit or implicit contract between the managed service provider and the organization or specifically the teams, with consequences of meeting or missing agreed objectives.

Finally there is the construct of the error budget. With the error budget one can determine the acceptable level of unreliability. Simply said; when there is budget left one can focus on increasing the number of features in the platform. When budget is exhausted or overspend features releases are frozen, post mortem items are prioritized and focus should be on improving monitoring and observability to enable a more proactive response and restore reliability.

Your in-house supported platforms

The aforementioned components are not only relevant for working with managed service providers. They are equally important for your internal organization.

Team Topologies visualisation of an internal platform utilising a managed services platform

In the figure above the mentioned internal platform builds on top of the services provided by the managed service provider. This (internal) platform is the smallest set of API’s, documentation, and tools needed to accelerate teams. It extends the value delivered by the managed service by adding organization specific value added services for example a policy-driven automation pipeline providing trust through the enforced sequence of required activities/steps involved in the development and deployment process. And for example transitioning services supporting teams to get onboarded to the platform with as little friction as possible.

The platform enablement team in the example operates in a facilitating mode for the stream-aligned/product teams enabling them to effectively use all the capabilities of the platform. Enabling teams proactively seek to understand the needs of the product teams using the service. They stay ahead of the curve regarding approaches, tools and practices in their area of expertise. Besides this they have a key learning function, promoting learning across teams. Enabling teams support flow of product teams, removing friction as much as possible and allowing for the increase of speed as well as cadence by supporting them to continuously improve their way of working.

Ownership of these internal platforms is, as the name already states, internally. They can of course be staffed through combinations of associates and partner-staff. These internal platforms adhere to all the concepts of the modern managed service as mentioned before.

Conclusion

“Managed services is the practice of outsourcing the responsibility for maintaining, and anticipating need for, a range of processes and functions in order to improve operations and cut expenses.” Revisiting the definition we see that, while still important, a shift has happened from only focussing on cutting expenses to an emphasis on anticipating need and improving operations.

Traditional outsourcing deal economics are not made for the new world. Reconsidering the paradigm is mandatory and a shift is happening. Main focus of any party delivering a service should be the success of the teams using the service. When working with other organizations this means finding ways to create a win-win situation. A setup where both parties gain from adding more relevance.

A great way to set up such an agreement is a formal relational contract[7]. This type of contract is used to govern highly complex relationships that demand collaboration and flexibility. Building on top of a foundation of trust, joint vision and goals, and shared understanding of the guiding principles the terms & conditions are defined and governance mechanisms established.

Adopting modern platforms means taking a holistic approach when looking at the organization and your value streams. What do you develop yourself and what do you outsource? How do you optimize variation to maximize differentiation? And when applying outsourcing, how do you make sure you and your service provider(s) collaborate on your journey for joint-success?