Cost per record is the wrong metric
Buyers compare suppliers on cost per record because it is the number on the invoice. But the record is rarely the dominant cost in an outbound program — the agent minute is. Once you fold in the labour spent dialing numbers that no longer connect and contacts who no longer remember opting in, an aged file at a fraction of the price routinely produces a higher cost per acquisition than premium fresh inventory.
Run the calculation properly: cost per acquisition equals total spend, records plus labour, divided by conversions. Do it separately for each freshness tier over a fixed test period. The ranking that emerges is frequently the opposite of the cost-per-record ranking, and it is the only one that should drive purchasing.
What actually decays
Three things erode simultaneously and at different speeds. Contactability goes first and fastest: numbers change, mailboxes fill, people stop answering unknown callers. Recall goes next — the consumer's memory of the entry, which is what makes your opening line land as a follow-up rather than an intrusion. Intent goes last but most decisively: the need that motivated the entry has often been met elsewhere, frequently by a competitor who called on day one.
Aggregate freshness figures hide this. 'Under thirty days' can describe a file that is mostly twenty-eight days old. Ask for the age distribution, not the ceiling, and price accordingly.
Match the tier to your operation
Real-time and sub-24-hour inventory is only worth its premium if you can work it immediately. A team that batches leads overnight has already destroyed most of what it paid extra for. Be honest about your actual response latency and buy the tier your operation can service — under three days for teams working a daily queue, aged inventory only for low-cost email nurture where the economics tolerate it.
If you want the premium tiers to pay, fix routing before you buy: instant delivery into the dialer or CRM, automatic assignment, and a first-touch target measured in minutes. Freshness is a joint property of the data and the operation. Buying one without building the other is just a more expensive way to get the same result.