The difference in one sentence
A prospective adjustment changes the price only for work done from the adjustment date forward. A retroactive adjustment changes the price for work already done, going back to when the new index value technically took effect, even if it was published later.
Why this matters more than it sounds
Publication lag means the index value for a given month is not known until later. If a contract adjusts prospectively on the 1st using the prior month's index, there is always a real cost change happening during the lag that never gets priced in. Retroactive adjustments close that gap, at the cost of reopening and correcting invoices that already went out.
When prospective works fine
If the publication lag is short relative to the adjustment frequency, or the index is stable enough that a one-period lag rarely matters, prospective adjustment is simpler for both sides to administer and rarely creates a meaningful gap.
When retroactive is worth the complexity
If the publication lag is long, or the index moves fast enough that a one-period delay is a real cost, retroactive adjustment protects whichever side would otherwise absorb that gap every single period. This is common in fast-moving input categories, where a systematic one-period lag compounds into a real, ongoing loss for one side.
The decision that actually matters
Check who benefits from the lag under a prospective structure, by default. If rising prices mean the buyer always gets a period of the old, lower price before the adjustment catches up, prospective adjustment is quietly biased in the buyer's favour — and the supplier should know that before agreeing to it.