AINEWS
McKinsey's 5–20% revenue uplift claim points at a quote quality lever in configured to order and data center work, not a growth miracle.
Three days. That is roughly how long a senior sales engineer at a mid-sized industrial equipment maker used to spend assembling a configured-to-order quote for a data-center cooling skid: emails to the panel supplier, a call with the switchgear vendor, margin gymnastics against an internal cost model, and a final pass to defend the discount. A mid-tier rep working from the same catalog and the same price book, but with the right AI tooling wired into the quote configurator and the prior-proposal archive, can now do the same work in an afternoon. The senior engineer is not replaced. The win-rate stops being a function of which rep happens to pick up the phone.
This is the workflow McKinsey's industrial-sales research is actually pointing at, and it is the one being squeezed hardest right now. Data-center build-outs are pulling industrial suppliers into integrated, tailored packages that look more like engineered projects than commodity orders. Geopolitics has redrawn the cost basis for inputs, and input prices stay volatile enough that a quote issued on Monday can be wrong by Friday. Industrial customers themselves are more sophisticated, asking for solutions that bundle hardware, software, and service in ways the relationship-driven sales model was never designed to handle at scale.
McKinsey's research suggests that industrial manufacturing leaders that overhaul pricing, growth strategy, and sales productivity in a coordinated way can unlock 5 to 20 percent revenue uplift and 5 to 10 percent EBITDA improvement within two years, with substantial gains showing up in the first 12 months (McKinsey). That is a wide range, and it reads better as a research signal about the ceiling and the floor than as a forecast for any single company. The useful part of the framework is what the work looks like, not the top-line number.
McKinsey sorts industrial operators into five archetypes by how they actually sell: turnkey solutions, engineered-to-order, configured-to-order manufacturers, standard-product businesses, and aftermarket and service providers (McKinsey). Turnkey is the slow lane, with 6 to 24 month sales cycles driven by direct project sales that are heavily specification- and relationship-driven. Configured-to-order sits in the middle: high SKU counts, configurable BOMs, and quotes that have to absorb volatile input costs without losing margin. Standard-product and aftermarket sit on the other end, where the bottleneck is coverage and renewal cadence rather than quote complexity. The mechanism McKinsey describes lands hardest in the middle of that map.
The shift is not about AI replacing the senior engineer. It is about codifying the institutional knowledge that used to live in that engineer's head and folding it into the quoting and pricing workflow. Quoting gets faster because the right prior proposal, the right cost benchmark, and the right margin guardrails are surfaced in the same tool the rep is already using. Coverage expands because the same rep can credibly quote configurations they have not personally configured before. Pricing discipline tightens because the discount logic is consistent across reps and regions. In a configured-to-order operation, the result is a quote that is faster, more accurate, and less dependent on a handful of irreplaceable senior engineers.
What AI does not replace matters as much as what it does. Project-level relationships still close the deal in a turnkey bid. Commissioning and on-site coordination still belong to people. Complex subcontractor handoffs do not survive being handed to a model. The work AI handles best is the part that has been hardest to standardize: assembling the quote, surfacing the right prior, and pushing margin discipline into the day-to-day. In return, it asks sales engineers to spend less time pulling data and more time on the conversations only they can have.
For a mid-sized industrial sales organization, the first 12 months of this kind of overhaul are mostly a quoting-floor project, not a strategy project. The team that picks one archetype, instruments the quote-to-cash workflow, and routes the senior engineers' institutional knowledge into the tooling will see the McKinsey range as a planning anchor. The teams that try to lift the number without changing the workflow will find the same ceiling their relationship-driven model already knows about.