Do not average conflicting estimates automatically. Reconcile the assumptions that are driving the gap.

Two methods are useful because they expose which input is unstable.

Compare the implied usage rate, penetration, or price in each approach and identify which assumption is least credible.

Choose the better-supported estimate, explain why, and use the second method as a sanity check. You can also use this CaseBasix guide for a structured review.

Testing a resume bullet on someone outside your industry is such a good filter. Jargon becomes obvious fast.