Pushing back the pendulum

Having a look back over time shows how the cost risk pendulum has swung back and forth between clients and service provider in technology services, and new commercial thinking is now needed to meet the challenges of 21st century technology delivery

If you look back over the vast vista of time at how technology services have been delivered you can see the evolution of risk between clients and suppliers with the pendulum of doom swinging back and forth between clients and suppliers. In the very early years, most organisations did everything themselves, but then eventually realized they didn’t need to do that. And so did suppliers who set themselves up to take advantage of the growing market, and the unwary clients…

1. Evolution of IT cost risk over the years

From before “noughties”, there was something of a “wild frontier” as clients had little experience of outsourcing technology services and suppliers made their fortune selling their wares. The deals back then were often not well constructed and frequently client unfriendly (even “gouging”). Often all of IT was thrown to the suppliers leaving a very thin retained organization, with most of the “brain” outsourced as well as arms and legs. So clients were at a significant disadvantage, and suppliers could be like a fox in the chicken coop, writing business cases for their own new services to be signed off by client managers who had no bandwidth to challenge them! Being virgin territory, services were also poorly defined and relatively unstructured, accompanied by opaque or just plain bad commercial models (e.g., pure T&M or lop-sided ARC/RRC models) that dumped the risk firmly in the clients’ lap – the commercial nadir…

In the 2010s, the pendulum started to swing back in client favour, as they started taking back control of strategy and architecture, business case development and key aspects of the “intelligent client” model (think of the previous model as the “hostage client”!). Other developments in service and commercial structures also helped as lessons were learnt from the first generation experience. For example, the service “tower” model became more well established, also with the beginning of Service Integration/SIAM disciplines to manage multi-vendor setups. The pricing models improved with the introduction of transparent PxQ “utility” pricing, aligned and integrated with well defined performance management and incentives. This was probably the (first) zenith of the art of technology service outsourcing.

As we move through the 2020s, you can see things falling apart from that peak of perfection as client “digital” demands and the technology landscape change with their technology teams looking for new ways of working to meet that. Probably the most significant trend is client organisations bringing the management of technology services and indeed some execution and delivery back in-house. There are no doubt many reasons for this drive, including disappointment with and inflexibility of previous outsourced arrangements and a perception that direct control is needed to increase agility. To be honest, the problem doesn’t always sit with the service providers, and reorganizing doesn’t often solve systemic problems, but there you go.

Apart from the swinging “in/out” door, there are other drivers. In particular, SaaS and IaaS are eroding “traditional” infrastructure and application management services reducing the scope for outside services (you can read more about that here). This erosion significantly thins out the service management layer required on top of the “Cloud” services compared to old-style hosting, which impacts the implicit business case. Automation also changes the landscape, increasing the reach and operational leverage of the people in the driving seat; a fully automated DevOps/SRE type mode needs no IT Ops people (well, that’s the philosophy). Demand Management (including FinOps) are now key skills with cloud bloat replacing VM bloat of yesteryear. Demand management is logically a client side function, in terms of the link to the business, although you can outsource / delegate the supply side matching.

If we look forward to the future, we can expect to see the incursion of GenAI further eroding the traditional opportunities for outsourcing, and further confusing the clear and simple lines.

2. What does the future hold?

The actual effect of genAI in terms of cost risk is probably one for the crystal ball just now. GenAI and its siblings might improve the profile as automation replaces yet more labour costs, but flakey implementation of expensive systems with unclear benefits and increasing supplier lock-in can easily drive the pendulum the wrong way. For suppliers, it drives a further shift of revenue from people to technology; the more fleet of foot will probably win either way

Coming back to the here and now, as client organisations swing back to a more in-sourced model, you can see repercussions with, for example, sub-optimal (re)sourcing when using third-parties. Often, resourcing generally reverts to unstructured staff augmentation and body-shopping to fill specific gaps in the in-house teams, forgoing the benefits of a more coordinated approach to third parties. Whilst previous models may have had carefully crafted off-shore resourcing to benefit from labour cost arbitrage, the in-house models have the hidden costs of expensive on-shore day rate contractors back-filling vacancies.

2. Buying “Bodies” with exacting specifications is hard

The “bodies” are often requested in an ad-hoc manner and to an exacting specification that significantly constrains what can be provided, artificially limiting supply and so driving up the cost and risk. For example, the role they to be filled in the patchwork structure will have a very tight skill specification, must work in a specific location and can only be quite senior/experienced (the logic being “we’re not paying to train junior supplier people on our projects”). In this model, the client organization carries all the risk on cost, productivity & quality.

So the unintended consequence of the well-meaning strategic changes in service resourcing model is the loss of some of the good stuff that went before. As well as the fragmented supplier deployment and inefficient resourcing, another major impact is the resultant loss of clear service structure and definition and unclear/limited performance management. So the cost risk pendulum is swinging significantly against clients who are now again bearing the risk.

4. Reintroducing third party service structure

The way out of the quandary is to reintroduce some sort of structure with a clearer and somewhat wider scope that can be resourced more flexibly as a service responsibility (even a small one), including the innate ability to refresh and improve itself. One of the challenges is that supplier have quite understandably aimed to limit their risk by walling themselves inside tightly negotiated scope with exceptions for anything that goes wrong everything outside that.

As a concept, it is one that lawyers and commercial managers like and that they can wrap themselves up warmly when they go to bed at night. However, it is somewhat dinosaur thinking, and is unsustainable in the modern world where business-technology performance is better and more meaningfully measured by end-to-end end experience service levels (e.g., with much touted XLAs – “experience level agreements”) or even more holistic business outcome metrics that are actually linked to the performance of the client business. Employee bonuses are commonly linked to such “One Team” metrics, however, supplier agreements are rarely so.

Some of that lack of performance-benefit linkage is historically because previous attempts to connect end-to-end have often failed dismally, be it due to poor definition, poor data, or somehow the grand ideas of partnership formed with a handshake on the golf course just didn’t pan out in the cold light of implementation (golf courses are really not good venues for making major strategic decisions!).

Some of the challenge however is also down to old-fashioned thinking about how the boundaries of responsibility are defined and how the benefits of supporting a winning business can be shared (or, of course, losses taken). You probably can’t blame the lawyers as they still think using Latin in contracts is smart, are mired in centuries of historic case law and outdated legislation and just don’t think along commercial service lines!

5. Service structure design decision

Without attempting to solve the entire conundrum of aligning client and supplier incentives and risk-reward sharing, we can push back against the reversion to bad (re)sourcing behaviours, by imposing some structure on the service requirements. That could be by moving back to the comfort blanket of older “tower” models. However, perhaps instead moving forward to a micro-sourcing model with small service components plugged together to fit the client need, and framed with more innovative commercial and performance management structures with a “One Team” twist. There will be nay-sayers who declare “but it’s not market standard”, however the appropriate response to that is “stasis is death” and “average is for losers”, so I say: think up, think harder and imagine the commercial possibilities!!!

Leave a Comment

Your email address will not be published. Required fields are marked *