Example: Should we build our new reporting tool in-house or buy an off-the-shelf SaaS solution?

Buy an established SaaS solution for the initial need, but run a comparative three-year TCO calculation and check the vendor's exit terms before signing — the spread in the panel's self-assessed confidence (35–82) is too wide for an unconditional recommendation.

Moderate-to-strong agreement that buying an established SaaS solution is the most cost-effective first choice when requirements are standard and time is short. Agreement is weaker on how much integration work existing systems will actually require.

Participant B objects that a SaaS solution locks the organisation into a vendor whose pricing can change once migration costs are already sunk — evidence that would settle it: the vendor's historical pricing-change rate and the contract's exit terms. Participant D notes that in-house builds historically underestimate post-launch operating and maintenance cost — evidence that would settle it: an honest three-year TCO calculation, not just development cost.

The panel assumes the reporting tool's requirements are largely standard and do not require deep integration with proprietary internal systems, and that the organisation has enough negotiating leverage to secure reasonable exit terms in a SaaS contract.

If requirements turn out to be more specialised than assumed, or if the vendor raises prices sharply after migration, an early SaaS purchase could end up costing more over time than an in-house build — and switching vendors after lock-in may require another costly migration.

Medium. The spread in the panel's self-assessed confidence (35–82) is too wide for high confidence; the evidence that would settle it is missing (a concrete requirements spec and a comparative TCO calculation).

Participant A alone favours building in-house now, arguing that reporting logic is part of the organisation's core competence and should not be outsourced — a well-reasoned position not voted down by the majority.

Draft a concrete requirements spec, request exit terms and historical pricing trends from 2–3 SaaS vendors, and run a comparative three-year TCO for both options before deciding.

If your requirements are highly specialised or regulatorily sensitive (e.g. sector-specific data residency), if you already have the in-house expertise to build and run the tool more cheaply than market price, or if your negotiating position with vendors is weak.

How sensitive are the requirements to future change, and can a SaaS solution's configurability keep up? What would a full data migration cost if the vendor is switched later? Is there already internal expertise that would make a build cheaper than it looks on paper?