Evaluating an audit software vendor's roadmap claims

A firm evaluating a vendor's roadmap claims should never buy based on a feature described as upcoming without separately checking the vendor's track record of actually shipping what it previously said was coming — the specific test is to ask for the last two or three features the vendor promised in a prior sales cycle or release note, with dates, and verify they actually shipped on roughly that timeline, because a vendor's past shipping behaviour is a far better predictor of whether a currently promised feature will arrive than anything said in the current sales conversation.

Why a roadmap claim is the easiest thing to oversell

A sales conversation has every incentive to describe a missing feature as close, because 'it's coming soon' keeps a deal that might otherwise be lost over a feature gap. This isn't necessarily dishonest — a product team genuinely believes a feature is close more often than it turns out to be — but the incentive runs the same direction regardless of intent, which is exactly why it's worth checking against evidence rather than taking on trust.

The specific questions that separate shipped from planned

A firm's own track-record check

1. 01Ask the vendor for release notes or a changelog going back at least a yearA vendor with nothing to show here either doesn't track its own shipping history or doesn't want to share it — both are informative.
2. 02Pick two or three features from a year or more ago that were described as upcoming at the timeUse the vendor's own historical sales material or release notes if available, or ask a reference client what they were told a year ago.
3. 03Check whether those features actually shipped, and roughly on the timeline originally statedA pattern of features shipping six months to a year later than first promised is common and not necessarily disqualifying, but it should recalibrate how the firm weighs the current promise.

Never make the purchase decision contingent on an unshipped feature alone

• If a feature the firm genuinely needs is only promised, not shipped, evaluate the vendor as if that feature doesn't exist yet, and decide whether the rest of the product is worth adopting on its own merits.
• If the firm does proceed partly on the strength of a roadmap promise, get a specific commitment in writing with a date and a consequence if it's missed, not just a verbal assurance repeated in the contract negotiation.
• Treat a roadmap claim about an AI feature with extra scrutiny — see the companion article on what AI-powered claims actually mean, since roadmap promises about AI features are especially prone to overstatement.