Manufacturing Technology Services

This is a planned Forrester category. No Wave or Landscape has published under this name. Forrester covers the underlying market through The Forrester Tech Tide: Smart Manufacturing, its Future Of Manufacturing research, and annual smart manufacturing predictions, which is what this article draws on.

The constraint on manufacturing technology is not the technology.

Forrester's 2026 outlook describes a workforce that will not be trained overnight, or even by the end of 2026, and frames the resulting choice starkly: risk the cost and negative headlines of waiting until a domestic workforce is ready, or risk the cost and negative headlines of importing experienced short-term workers from overseas to get the factories working.

Read that as the market rationale for a services category. Manufacturers are building capacity in places where the people who can run modern plants do not currently live, on timelines that do not accommodate training them. That gap is filled by somebody, and increasingly it is filled by service providers.

What the category would cover

A manufacturing technology services provider sits between industrial operations and enterprise technology, and the scope is unusually wide because the estate is unusually heterogeneous.

Forrester's Tech Tide on smart manufacturing identifies twenty technology categories underpinning the field, which is a fair indication of how much has to be integrated. In practice a services engagement spans some combination of manufacturing execution systems, industrial control and SCADA environments, industrial internet of things instrumentation, quality and traceability systems, maintenance and asset performance, supply network planning, digital twins, and the connection of all of it to enterprise resource planning.

Around that sit the disciplines that make it work: OT security, data engineering across plants that instrument differently, and the change management required to get people who have run a line one way for fifteen years to run it another.

Forrester's Future Of Manufacturing research names four trends driving modernisation, and they remain the honest description of what a provider has to be good at.

Integrating digital technologies with physical products. Adapting to a changing world order with local, near, and far manufacturing capacity. Electrifying, decarbonising, and manufacturing sustainably. And balancing the automation triangle to get the best from hardware, software, and people.

The automation triangle

That last phrase is the most useful framing in Forrester's manufacturing research, and it explains why so many factory technology programmes disappoint.

Hardware, software, and people are three levers, and the temptation is always to pull one. Buy robots. Deploy a platform. Retrain the workforce. Each in isolation produces a predictable failure.

Hardware without software produces expensive equipment generating data nobody uses. Software without hardware produces dashboards describing a process that has not changed. Either without people produces a system the line operators work around, because the people who actually run the plant were not part of designing how it would run.

The balance point differs by plant, by product, and by labour market. A high-volume, low-variation line in a high-wage economy justifies heavy automation. A high-mix, low-volume operation with skilled workers frequently does not, and automating it produces rigidity where flexibility was the advantage.

A services provider's real value is judgement about where that balance sits, which is consulting work rather than implementation work, and it is the part that distinguishes a partner from a systems integrator.

The divide that makes this specialist work

The structural feature of manufacturing technology, and the reason generalist services firms struggle here, is the gap between operational technology and information technology.

They are genuinely different disciplines with incompatible instincts.

IT optimises for confidentiality, integrity, and availability, in roughly that order, and patches on a schedule. OT optimises for availability and safety above everything, and does not patch a controller that is running because the plant does not stop.

IT equipment has a three to five year life. OT equipment has a twenty to thirty year life, which means a modern plant contains controllers older than the engineers maintaining them, running software from vendors that no longer exist.

IT accepts downtime windows. OT downtime is measured in lost production, and in continuous process industries a restart can take days.

And the failure modes differ in kind. A compromised IT system leaks data. A compromised OT system can damage equipment or injure someone.

A provider that treats a factory as an unusual data centre will make recommendations that operations will refuse, correctly. A provider that understands the divide arrives with different assumptions about change windows, testing, and what can be touched at all.

That is why manufacturing technology services resists absorption into general IT services, and why a dedicated category is a reasonable thing for Forrester to be planning.

Forrester's scepticism is worth borrowing

Two of Forrester's predictions are notable for pushing against the prevailing enthusiasm, and both are useful in a vendor conversation.

On humanoid robots, Forrester's assessment is that only a few will get paid to do useful work in 2026, and that despite a flurry of press releases, videos, and investment announcements, there is little evidence of more humanoids doing a day's work.

That is a firm statement in a field currently attracting enormous capital and attention. A services provider building a manufacturing roadmap around general-purpose humanoid robotics is describing a future rather than a plan, and the distinction matters when the roadmap is the thing you are buying.

The second is more prosaic and more immediately consequential: the surprising difficulty of connecting factories to electrical power.

Grid connection timelines have become a genuine constraint on industrial expansion in several European markets, with queues measured in years rather than months. An organisation planning new capacity, or planning to electrify existing processes as part of a decarbonisation commitment, may find that the technology decisions are downstream of a utility connection date nobody controls.

That belongs in a services engagement's assumptions rather than as a surprise in year two, and a provider that raises it unprompted is one that has done this before.

Geography is now a technology problem

Forrester's framing of supply network strategies combining onshoring, offshoring, reshoring, nearshoring, and friend-shoring describes something that used to be a procurement decision and has become a technology one.

If production moves closer to consumption, an organisation ends up operating more plants, each smaller, each in a different regulatory and labour environment. That multiplies the technology estate rather than relocating it, and it makes standardisation across sites both more valuable and more difficult.

Forrester also notes Chinese manufacturers moving into competitors' backyards, with car plants announced in Hungary and Turkey and joint ventures elsewhere, no longer operating as low-cost contractors to western manufacturers but building competitive products close to European customers.

For a European manufacturer that changes the competitive calculus in a way technology services intersect with directly. A new plant built now with current technology has a structural advantage over a plant modernised incrementally over thirty years, and closing that gap is precisely what a modernisation engagement is for.

The regulatory dimension is European-specific and heavy. Sustainability reporting obligations, product-level requirements including digital product passports, and supply chain due diligence rules all require data that most manufacturing estates do not currently produce in a usable form. That is a data engineering problem before it is a compliance problem, and it is one of the more defensible reasons to engage a provider.

What a scored evaluation would need to measure

If Forrester publishes a Wave in this category, three choices will reveal what it thinks the market is.

Whether OT depth is a criterion or an assumption. A criteria set that weights industrial control experience, safety system familiarity, and plant-floor change management heavily describes a specialist market. One built around cloud, data, and integration describes a general services market with a manufacturing flavour, and the two produce very different vendor lists.

How it handles the physical. Manufacturing services frequently involve equipment, commissioning, and on-site work, which most technology services evaluations do not contemplate. A provider that can specify a line, integrate the controls, and connect it to the enterprise stack is doing something few software services firms can.

And whether workforce capability is scored. Given Forrester's own finding that the constraint is people rather than technology, a provider's ability to supply, train, and transfer capability may be the most consequential differentiator, and it is one that capability matrices tend to underweight.

Where this leaves a buyer now

Without a scored evaluation, the useful preparation is internal rather than comparative.

Establish honestly where you sit on the automation triangle, and whether the plant's constraint is equipment, systems, or people. Programmes fail when a provider is engaged to solve one and the actual constraint is another.

Map what your estate can currently report, because the sustainability, traceability, and product-level data obligations arriving in Europe require information that most plants generate as paper, spreadsheets, or not at all.

Treat Forrester's own framing as the filter for vendor claims. Durable business transformation against expensive experiments is the line its 2026 research draws, and a proposal weighted toward humanoids, general-purpose robotics, or autonomous plants is on the experimental side of it, whatever the demonstration looked like.

And ask about power and permitting early, because on current evidence in several European markets those timelines will determine the programme schedule regardless of how good the technology plan is.

Forrester does not endorse any provider named here, and nothing in this article should be read as a recommendation to buy.