The estimate is not wrong
about the parts it includes

When a team estimates building a data platform, the estimate is usually accurate for the work it describes. The problem is the work it does not describe, which is most of the work.

What the estimate covers

Connecting the sources, landing the data, writing the transformations, standing up the warehouse, building the first dashboards. This is real work and teams estimate it reasonably well, because it is visible and has been done before by somebody they can ask.

It is also the part that finishes. The demo happens, it works, and the plan looks vindicated.

What it does not cover

Schema drift, which is continuous. Source systems change without telling you, and someone has to notice, diagnose and fix each time, usually after a number has already moved.

Backfills, which are always harder than expected because the logic that was right for new data has to be made right for old data with different shapes.

On-call, which is not a task but a permanent tax on the team's attention. A pipeline that fails at two in the morning is somebody's night, every time, forever.

And the rebuild, because the first architecture was chosen while learning and one of its early assumptions will turn out to be load-bearing and wrong.

Why this is not a criticism of the team

None of these are visible from the start. You cannot estimate the cost of a source system changing in a way it has not changed yet, and no reasonable planning process would ask you to.

This is why build estimates are systematically optimistic across the whole industry rather than at any particular company.

How to decide honestly

The question is not whether your team could build it. Most competent teams could. The question is whether the eighteen months and the permanent operational load buy you something a customer would notice.

If your product's differentiation is in the data layer itself, build it and accept the cost knowingly. If the data layer is infrastructure under a product that is differentiated elsewhere, the honest answer is usually that it is not where your quarters should go.