A Structural Firm Rolled Out BIM 360 Three Times Before It Stuck — What Changed

What "BIM in Structural Engineering" Looks Like When It's Working

BIM in structural engineering means authoring the structural model in a tool like Autodesk Revit or Tekla Structures and hosting it in a common data environment (a shared cloud workspace, such as Autodesk's BIM 360, now Autodesk Construction Cloud) so the structural, architectural, and contractor models coordinate continuously. Clashes get caught early, changes propagate, and engineers spend less time reconciling drawings by hand.

A common data environment (CDE) is the shared cloud workspace where every discipline's model lives, so coordination happens continuously instead of in a fire drill the week before issue. A BIM execution plan (BEP) is the document that says how a project team will actually use BIM— file structure, naming, who owns what, when models get exchanged.

One naming point matters for the rollout conversation. Autodesk has moved customers off BIM 360 and onto Autodesk Construction Cloud (ACC), folding BIM 360's separate Docs, Build, and Coordinate modules into one place.⁷ When this article says "BIM 360," read it as "the cloud platform— BIM 360 then, ACC now."

Here is the part most firms underuse. Even where BIM is adopted, it tends to get used narrowly. One peer-reviewed study put 3D coordination and faster model creation at roughly a 60% application rate, with BIM for structural analysis at about 27% and energy analysis around 25%.³ Most firms own far more capability than they run.

For a sole practitioner, skipping full BIM can be a perfectly rational call— the overhead may not pay back at that scale. For a firm with a BIM manager, real project pressure, and contractors who expect a coordinated model, it is table stakes. Which is exactly why a stalled rollout hurts.

So if the tools are mature and the workflow is well understood, why did the first two rollouts fail? Partly the software. Mostly not.

What Was Actually Broken— and What Wasn't

Two things are true at once. BIM 360 had real product friction— slow sync, sync failures after long waits, modules that ran in their own silos, a learning curve steep enough that anyone outside the BIM coordinators needed near-constant hand-holding⁸— and Autodesk's move to the more unified ACC is a quiet admission of it.⁷ And: a rollout that fails three times does not fail because of sync latency. It fails because no one with authority owned the change.⁴

The friction was real, and it is worth naming honestly so nobody who lived through it feels gaslit. Practitioners on Autodesk's own forums reported slow syncs, syncs that failed after a long wait, and lost work when files and x-refs were not kept extremely clean.¹⁰ Non-specialists— the project engineers who are most of the firm— needed ongoing oversight just to use the platform reliably.⁸ And BIM 360's Docs, Build, and Coordinate modules each ran in their own space, which created the data silos ACC was built to consolidate.⁷ So yes: sometimes the tool was part of the problem.

Now the harder truth. Industry analysis converges on the same finding— BIM implementation failure rates stay high even as adoption rises, and the surface reasons look technical while the actual problem is how the change is led.⁴ The most-cited barrier in survey after survey is a skills gap, not a software defect.⁴ The classic anti-pattern shows up everywhere: buy the tool before defining the process, hand the execution plan to one person, then expect a return inside a single project.⁵⁶ When that plan belongs to one BIM manager, the rest of the team forgets it exists.⁵

It helps to separate the two layers:

What was genuinely the software's:

• Sync speed and sync failures under load
• File and x-ref fragility that cost people work
• Siloed modules that fragmented project data
• A learning curve steep enough to need constant support

What was the firm's:

• No principal accountable for the outcome
• An execution plan one person wrote and nobody else used
• Generic training disconnected from real project work
• Expecting payback inside one project, under deadline pressure that punished anyone slow

Structural firms carry an extra weight here. Construction has historically spent under 1% of revenue on IT— less than a third of what automotive and aerospace spend¹— so the muscle for running a tooling change is underbuilt to begin with. Add a project-by-project culture and deadlines that make the fastest path the old path, and the resistance you hit is rational under the incentives.⁶ The fix is organizational. Building an AI culture inside the firm runs on the same logic: the firm has to decide to work differently, or the tool just sits there.

On attempt two, the firm bought faster hardware. It still didn't stick— because the hardware was never the bottleneck. The platform was never the bottleneck. The people, the process, and the absence of an owner were.

Which brings us to the third attempt— the one that stuck. Five things were different, and none of them was a piece of software.

The Five Things That Changed on the Attempt That Stuck

A BIM rollout tends to stick when five things change: a principal owns it instead of delegating it, it starts on one pilot project with a named owner instead of going firm-wide, the team redesigns the workflow instead of just installing the tool, training happens inside live project work instead of in a generic seminar, and leadership gives it a full year and one honest metric before deciding whether it worked. None of those is glamorous. Each one costs something.

1. A principal owns it. The BIM manager still runs execution. But a firm leader is visibly accountable for whether the firm actually changes how it works. On the third attempt, the managing principal sat in the pilot project's weekly coordination call for the first two months. That was much of the difference, and it cost her two hours a week plus the political capital to say this was happening. The price: leadership time— and the fact that a plan owned by one person gets forgotten, while a decision owned by a principal has to be lived with.⁵⁴

1. It starts on one pilot project with a named owner. Pick a real project, put someone accountable for that project's result, and work the new process out there— not firm-wide. Large engineering firms that treated a BIM 360 rollout as an enterprise flip found project setup across thousands of users genuinely hard; the ones who got it treated it as a multi-year change program rather than an install.⁹ The price: you tell the rest of the firm "not yet," and the pilot project carries extra overhead while the kinks get worked out. The hidden costs of a technology project are mostly this— the slowdown you absorb on purpose.

1. The team redesigns the workflow. Decide how models get authored, shared, coordinated, and issued in the new environment. Rewrite the steps. Don't bolt the platform onto the old steps and hope. This is the step most firms skip, and skipping it is why the tool ends up as expensive cloud storage.⁶⁴ The price: real design work up front, before anyone sees a benefit.

1. Training happens inside live project work. People learn the new workflow on a real deadline, with real stakes, with someone to ask— instead of a Tuesday lunch-and-learn they've forgotten by Thursday. Non-specialists need that ongoing support to actually adopt the platform.⁸ The price: the pilot project runs slower while people climb the curve, and leadership has to absorb that without punishing it.

1. Leadership gives it a year and one honest metric. Pick one real measure— coordination issues caught before issue, rework hours, time to incorporate an architectural change— and let it run across more than one project before you call it. Standardized practices and regular reviews are what carry adoption past the first project.⁶⁹ The price: patience, and the discipline not to declare victory the moment everyone has logins. Worth settling how you'll measure the rollout on day one, not month nine.

None of this is a secret framework. It is the change-management playbook the literature and the practice converge on. The reason it works is not novelty. It is that someone with authority decided to pay the costs— the slowdown, the "not yet," the design work, the year of patience— instead of pretending a rollout is free.

If you are about to attempt this again— second time, third time— you can tell early which version you are running.

How to Tell, Early, Which Rollout You're Running

You can tell within the first month. If the rollout is "the BIM manager, plus a training calendar, plus logins," it is the version that already failed— kindly, that one is going to stall again. If a principal is visibly accountable, there is a real pilot project rather than a firm-wide flip, the workflow was redesigned rather than just the tool installed, and someone is measuring one real thing, it is the version that sticks.

Four questions, this week:

1. Can you name the principal who is accountable for the outcome— the principal, not the BIM manager?
2. Is there a contained pilot project, or did the firm flip everyone at once?
3. Was the workflow redesigned for the new environment, or did the platform get layered on top of the old steps?
4. Is anyone measuring one real thing the rollout is being judged on?

If the answer to the metric question is no, the rollout doesn't have an owner— it has a volunteer. If the answers are mostly the failed pattern, the honest move is the uncomfortable one: don't roll it out again yet. Relaunching into the same conditions— no principal, no pilot, no year— just adds organizational scar tissue and makes the next attempt harder. If leadership can't commit a pilot, a principal, and a year, a fourth attempt fails the same way the first three did.⁴⁵

Watch for the rollout that looks successful and isn't: everyone has logins, the files are in the cloud, the dashboard is green— and the firm hasn't actually changed how it works. That is the failure mode that wastes the most time, because it takes the longest to admit.

The same discipline applies to any tool you're weighing. Deciding when to invest in new technology is mostly deciding whether you'll fund the rollout properly, not whether the tool is good.

The reason this matters beyond BIM: the next tool in line will test exactly the same muscle.

The Same Muscle— Why This Is Really About What Comes Next

Adoption is a capability a firm builds once and reuses. The structural firm that finally made BIM 360 stick didn't just get a coordination platform working— it learned how to land a tooling change without it dying under deadline. That is the muscle the next thing will need: AI assistants arriving inside Revit, automated clash and code checking, digital twins (a live data model of the built asset). Rolled out the old way, those will stall the same way the first two BIM 360 attempts did.

The money is already moving. Venture and private-equity funding into AEC technology ran about $50 billion globally from 2020 to 2022— roughly 85% more than the prior three years.¹ More tools are coming, faster than firms can absorb them, which makes the absorbing the constraint.

One honest caveat: an AI rollout is not the same as a BIM rollout. AI can change what the work is, not just which tool you use to do it. The transferable part is narrow— the discipline of owning a change, piloting it, redesigning the workflow, and measuring it. Build that once on BIM 360 and you have a head start on everything after.

If holding that line— the pilot, the principal, the year, the metric— is hard to do from inside the firm, that's ordinary; you can't read the label from inside the bottle. An outside partner can map the workflow, run the pilot, name the metric, and give leadership the structure to hold it, whether the tool is BIM 360, ACC, or an AI workflow. If that sounds like the help your firm needs going into attempt number three, that is what Dan Cumberland Labs does— build the adoption muscle with you, then leave it with you. It is the same discipline behind our AI implementation work with founder-led firms.

A few questions that come up every time a firm is staring down attempt number three:

FAQ

Why do BIM rollouts fail?

Repeated failures are organizational, not technical. No executive ownership, an execution plan one person wrote, generic training, and deadline pressure that pushes people back to the old workflow— that's the pattern.⁴⁵⁶ The software has rough edges, but a rollout that fails three times fails for reasons no platform change fixes.

Should we pilot BIM 360 on one project or roll it out firm-wide?

Pilot it on one project with a named owner. Firm-wide rollouts have nowhere contained to work out the new process and no one accountable for the result; the firms that succeeded treated it as a multi-year change program rather than a one-time install.⁴⁹

Is BIM 360 still a product?

Autodesk has moved customers to Autodesk Construction Cloud (ACC). BIM 360 is the legacy product, and ACC consolidated its previously separate Docs, Build, and Coordinate modules.⁷ The platform change fixed some real friction— it did not fix rollouts that fail for organizational reasons.⁸

How widely adopted is BIM?

Industry surveys put BIM adoption around 70%, a level it has held since roughly 2018, with almost one-fifth still saying they plan to adopt it.² So about a third of firms still haven't adopted it and the curve has flattened— struggling with adoption is common, not a sign you're uniquely behind. (That survey is UK-centric; treat the figure as directional for a US audience.)

Who should lead a BIM rollout— the BIM manager, IT, or leadership?

A principal or partner should own it, with the BIM manager running execution. Plans owned solely by a BIM manager tend to be forgotten by everyone else.⁵⁴

Does adopting BIM actually pay off?

Aggregate construction-industry research puts well-executed digital transformation at roughly 14–15% productivity gain and 4–6% cost reduction— meaningful, and slow.¹ Most BIM users report a positive return; the widely circulated exact percentages trace to a small set of older industry surveys, so treat them with attribution and a date rather than as fresh fact.

In structural engineering, BIM works when the firm decides to work differently and a leader holds that line for a year. The tool that comes next— AI in Revit, automated checking, digital twins— will work the same way, or fail the same way. The hard part was never the software.

References

1. McKinsey & Company, "Delivering on construction productivity is no longer optional" (2024) — https://www.mckinsey.com/capabilities/operations/our-insights/delivering-on-construction-productivity-is-no-longer-optional
2. NBS / Adrian Malleson (RIBA Journal), "UK National BIM Survey: the design community's adoption of BIM" (2023) — https://www.ribaj.com/intelligence/uk-national-bim-survey-design-community-adoption-bim-adrian-malleson
3. B. Habte & Y. Negash, "Application of BIM for Structural Engineering: A Case Study Using Revit and Customized Structural Panels," Journal of Information Technology in Construction (ITcon), Vol. 26 (2021) — https://www.itcon.org/papers/2021_53-ITcon-Habte.pdf
4. Elecosoft, "Why BIM implementation can fail and how to fix it" (2024) — https://elecosoft.com/news/overcoming-bim-failure/
5. BIM Corner, "Why most BIM Execution Plans fail?" (2023) — https://bimcorner.com/why-most-bim-execution-plans-fail/
6. Novatr, "6 BIM Implementation Challenges Faced by the AEC Industry" (2024) — https://www.novatr.com/blog/challenges-of-implementing-bim
7. VIATechnik, "Autodesk Construction Cloud vs. BIM 360" (2023) — https://www.viatechnik.com/insights/blog/acc-vs-bim360/
8. BIMPRO, "BIM 360 vs Autodesk Construction Cloud (ACC)" (2023) — https://www.bimprous.com/bim-360-vs-autodesk-construction-cloud/
9. Autodesk University, "Scaling Up—Enterprise Launch: A Case Study of BIM 360 to Autodesk Construction Cloud" (2022) — https://www.autodesk.com/autodesk-university/class/Scaling-Enterprise-Launch-Case-Study-BIM-360-ACC-2022
10. Autodesk Community, "BIM 360 — Advantages/Disadvantages" (practitioner discussion, 2020) — https://forums.autodesk.com/t5/bim-360-support-forum/bim-360-advantages-disadvantages/td-p/9681892