McGough had a Salesforce org its teams had stopped reaching for. Rebuilding it meant starting with the operational problems rather than the platform — and then putting enterprise AI in the hands of the field, not just the executive floor.
| Industry | Commercial construction |
|---|---|
| Company | 60+ years, multi-region |
| Products | Salesforce CRM, OpenAI / ChatGPT |
| Services | Re-implementation, process design, AI enablement |
| Engagement | Ongoing partnership |
Sixty years of relationships stopped being enough
McGough has built its business on people, reputation and repeat relationships. That still wins work. What leadership recognised is that it no longer wins it on its own — not against tightening margins, not against competitors moving faster, and not in a year when AI started changing what a bid team can do in an afternoon.
There was already Salesforce in the building. It had been configured around an idealised process rather than the one the business actually ran, so people worked around it, and the data thinned out until leadership could not trust what it showed. Re-investing meant finding a partner willing to ask why before prescribing anything, and willing to build with the culture rather than against it.
Four places the business could not see itself
- Disconnected workflows. Projects, regions and teams each ran their own version of the process, with no shared standard to reconcile them.
- No reliable pipeline view. Leadership could not assess where the business stood without a significant manual effort every time they asked.
- A field team with reason to be sceptical. Anything that read as surveillance, or as a prelude to replacement, was going to be resisted. The tool had to give something back before it asked for anything.
- AI without a framework. The tools arrived faster than any policy for them, creating enthusiasm and confusion at the same time.
Rebuild around the way people already work
What was hard
The configuration was not the difficult part. Convincing a 60-year-old organisation to try a system that had already disappointed it once was. The first version of anything we showed the field got measured against that memory, so early releases had to be small, visibly useful, and shipped where people could see them working before we asked for behaviour to change.
AI carried the opposite risk. Interest was high enough that the danger was a hundred private experiments and no shared approach. Framing the rollout as a company-wide initiative, with field teams invited to name their own use cases, mattered more to the outcome than any part of the deployment itself.
Adoption is growing because the system is useful
Salesforce usage is climbing, and it is climbing for the right reason — people are opening it because it helps, not because they were told to. That is the result the previous build never reached, and it is the one everything else depends on.
AI is no longer a leadership initiative. It is an organisation-wide conversation, with field teams actively bringing use cases forward. Leadership now reads pipeline and performance from the system rather than assembling a view by hand, which is a real shift from intuition-led to evidence-informed decisions. The cultural change — the hardest thing to move in an organisation this old — is likely to outlast every configuration decision in the project.
“We need partners who understand construction workflows, ask why before prescribing solutions, and align technology to business outcomes.”
What the AI rollout sets up next
- The enterprise ChatGPT deployment is a live proof of how AI accelerates knowledge work at scale, rather than a pilot waiting on a business case.
- Agentforce and field-specific AI use cases are actively scoped for the next phase, on a platform now carrying trustworthy data.
- Because the field named its own use cases, the next wave starts with demand rather than a change-management campaign.
Clouds & products
Salesforce CRM
OpenAI / ChatGPT
Agentforce (scoped)
