The McKinsey-ALICE Alliance: What Owners, Contractors and Founders Should Take From It

Reported by:  Owen Drury,as published on Bricks & Bytes

McKinsey has formalised a commercial alliance with ALICE Technologies, the generative scheduling platform founded by René Morkos.

The pair have already deployed across 35+ capital projects in infrastructure, data centres, energy, mining and manufacturing, claiming up to 20% schedule reductions on average and a 40% cut on one global data centre programme. The signal matters more than the deal itself: a Big Three consultancy is publicly underwriting AI-driven scheduling as a serious lever for fixing capital project delivery.

For five years, McKinsey and ALICE have been quietly running joint engagements on capital projects most of us never hear about. Data centre megaprograms. Energy infrastructure. Mining. The kind of work where a slipped schedule isn’t a Monday morning headache, it’s a nine-figure financial event.

This week, that quiet collaboration became an official alliance. And while press releases between consultancies and tech vendors are usually background noise, this one is worth pausing on. McKinsey isn’t endorsing a productivity tool. They’re staking a position on how capital projects should be planned, full stop.

For anyone building, buying, or investing in construction tech, the implications are bigger than they look.

In Short:

Generative scheduling treats labour, equipment, materials, space and sequence as variables a model can stress-test, rather than fixed inputs a planner has to commit to upfront. McKinsey is now formally selling that approach as part of a broader operating model overhaul. The bet: that consulting plus AI-native software, sold together, beats either one alone.

What McKinsey and ALICE actually announced

A five-year quiet collaboration becomes a formal alliance

The headline is straightforward: McKinsey’s Capital Excellence practice and ALICE Technologies have formalised a commercial alliance focused on generative scheduling for large capital projects. According to McKinsey’s announcement, the two firms have spent more than five years working together on client engagements before putting a label on it.

The numbers they’re publishing are the ones to watch. Up to 20% schedule reductions across 35-plus deployments. One unnamed global data centre operator achieved roughly a 40% reduction against the baseline construction programme by simplifying schedule logic and rethinking sequencing across 13 identified inefficiencies. That’s the kind of outcome that usually only shows up in vendor case studies. Seeing it co-signed by McKinsey carries different weight.

The technical setup is recognisable to anyone who’s followed ALICE. Their platform ingests BIM models and Primavera P6 schedules, then simulates millions of sequencing and resource-loading combinations, treating labour, equipment, materials, spatial constraints and sequence as adjustable variables. As AEC Magazine notes, this is meaningfully different from conventional critical-path scheduling, which forces planners to construct scenarios manually one at a time. Generative scheduling generates and ranks alternatives instead.

Why generative scheduling is suddenly a strategic conversation

The productivity gap got too expensive to ignore

Capital project delivery has been stuck for decades. Demand keeps rising, capital deployed keeps growing, but schedule and cost overruns remain the industry’s defining feature. McKinsey’s own framing in the announcement calls out the paradox directly: rising investment hasn’t translated into better outcomes.

The pressure on data centres alone is reshaping the conversation. Hyperscale operators need capacity yesterday, and a multi-month schedule overrun on a $500M programme isn’t a margin issue, it’s a competitive problem. The same logic applies to energy infrastructure, where projects are tied to grid commitments, and to manufacturing capex, where every week of delay is a week of lost output.

That economic backdrop is why generative scheduling is having its moment. As we covered in our deep dive on the fragmenting scheduling market, the sector has finally stopped looking for one tool to replace Primavera and started accepting that different scheduling problems need different tools. Generative scheduling sits at the megaproject end of that spectrum, where complexity makes manual scenario planning impractical.

"When embedded within the right operating model and supported by strong project controls, it can help organizations make faster, more informed decisions."

Erikhans Kok, senior partner and leader of McKinsey’s Capital Excellence Practice

The consulting plus software model is not new, but the framing is sharper

This is a deliberate move on the operating model, not a software resale deal

McKinsey is unusually direct in the announcement about what they’re not selling. The published commentary cautions that lasting improvement requires capability building and integration with existing planning processes, not a software rollout. That distinction matters. A tool dropped into a broken planning culture produces broken planning faster.

This is the same lesson David Rockhill made on the podcast when he discussed McKinsey’s playbook for scaling construction tech. His advice to clients was to start with what data they had, run the change, and let usage drive data quality upward. The ALICE alliance is essentially that thesis productised. The tech is real. The transformation work is what closes the gap between a 5% improvement and a 40% one.

For ConTech founders, this should land as a wake-up call. The big consultancies are no longer just vendor-neutral advisors pointing clients at a Gartner Magic Quadrant. They’re picking partners, formalising alliances, and bundling the change management with the software. If you’re building in a category where a McKinsey or BCG could choose a single technology partner, that decision is now part of your competitive landscape.

What this means for contractors, owners and ConTech founders

Three different read-throughs from one announcement

Owners of large capital programmes are the most obvious beneficiary. If you’re a hyperscaler, an EPC client, or a major industrial operator with a portfolio of complex builds, the question is no longer whether generative scheduling works on paper. It’s whether your internal planning team and your contractors can absorb it without breaking the rest of the workflow. That’s a procurement and change management problem, not a technology bet.

Contractors face a sharper read. If your major clients start asking for ALICE-modelled schedules as a condition of bidding, your scheduling team capability becomes a competitive lever. The contractors who can engage credibly with parametric, scenario-driven schedules will win work. The ones still defending a single CPM schedule as the only legitimate plan will struggle.

For ConTech founders, the read is more complicated. ALICE has just been validated by the most credible operating model authority in the industry. That moves the goalposts for any startup operating in adjacent scheduling, optimisation, or capital project planning categories. The mid-market opportunity that we’ve highlighted previously is still wide open. But the high end of the market just consolidated around a single combined offer.

How generative scheduling compares to the rest of the stack

A quick reference for where ALICE fits among the alternatives

The honest caveats that don’t make the press release

Replicating these outcomes is harder than the headline numbers suggest

The 40% data centre case is striking, but it’s also a single data point. McKinsey’s own framing acknowledges that results vary by project type, data quality, model fidelity, and the maturity of the owner’s planning function. A 40% reduction at a sophisticated hyperscaler with clean BIM data and a full McKinsey engagement is a different result from what a regional contractor with messy P6 files might achieve.

There’s also the question of attribution. Some of the gains in any consulting-plus-software engagement come from the consulting itself: better governance, sharper accountability, cleaner planning rituals. The technology is the visible artefact, but the operating model changes are usually doing a lot of the work.

None of which makes the alliance less significant. It just means the contractors and owners reading the announcement should resist the temptation to treat 20% as a number they can quote in their next board paper. The honest answer is that the upside is real, the methodology is credible, and the actual return depends on how much organisational rewiring you’re prepared to do alongside the software deployment.

Frequently asked questions

What is generative scheduling, in plain language?

Generative scheduling treats your project schedule as a model with adjustable inputs (labour, equipment, materials, space, sequence) and simulates millions of possible execution paths to find the most efficient and resilient ones. Instead of building one schedule and hoping it holds, you generate and rank alternatives. ALICE Technologies’ platform ingests BIM data and Primavera P6 schedules to run those simulations.

How is this different from existing AI scheduling tools like nPlan or Nodes and Links?

nPlan and Nodes and Links typically layer AI on top of existing schedules to automate reporting, forecast risk, or run agentic workflows. ALICE generates the schedule scenarios themselves from a parametric execution model. They’re complementary categories rather than direct competitors, though there’s overlap as each platform expands its scope.

Are the 20% and 40% schedule reduction numbers realistic for a typical project?

The 20% figure is McKinsey’s claim across 35-plus deployments, weighted toward complex capital programmes. The 40% case is a single global data centre operator. Both are upper-end results from sophisticated owners with full operating model engagements. Replicating those numbers depends heavily on data quality, schedule baseline accuracy, and the client’s appetite for organisational change.

Does this make ALICE the dominant player in scheduling?

It strengthens ALICE’s position at the high end of the market, particularly for mega-projects in data centres, energy and infrastructure. The wider scheduling market is still fragmenting across collaborative planning, look-ahead tools, AI-augmented reporting, and CPM compliance, with different vendors leading in each category. The mid-market between $20M and $100M projects remains largely unsolved.

What should a contractor or owner do about this in the next 12 months?

If you’re running capital projects above $100M, get a baseline read on your current schedule discipline before you bring in any tool. Audit your BIM and P6 data quality. Identify the planning rituals that are working and the ones that exist for compliance reasons only. The technology is the easier part. The hard work is making sure your team is set up to act on the outputs of any generative model you adopt

You can read the full article here.