Spent $4,800 on an email content retainer, 38 percent open rate, no calls

38 percent on a four year old list is high enough that I'd double check the platform is filtering bot opens from security scanners. Some corporate mail systems open every link before delivery. Doesn't change your conclusion, the calls are zero either way, but it might mean the 38 was never real and you were paying for a number that didn't exist.

The meetup signup sheet portion is a different list from the listing inquiries and I'd guess they behave nothing alike. Anyone who wrote their email on a clipboard at an event has near zero intent. Segmenting those out might have shown you the inquiry half was actually performing.

No property in it is the whole thing. I get four of these newsletters and I read all of them and I've never contacted any of the senders, because there's nothing to contact them about.

Your diagnosis is close but the mechanism is worth stating precisely, because it changes what you'd buy next time. You bought content production and measured it with an audience metric. Opens measure subject lines and sender reputation. Calls measure whether the send contained something a reader could act on. Those two are only loosely connected, so a vendor optimizing the first can hit benchmark for years and never move the second.

Your $4,800 only makes sense if the list had current intent. Listing inquiries from four years ago mostly bought something already, and a person who owns a house is not a buyer. Before you spend again, pull the inquiry dates and see what share of that 2,340 is inside 18 months. If it's 300 names, your real cost was $16 per active contact over the six months, about $2.67 a month each, which is a different business than the one you thought you were running.

@thatch is right that scanner opens inflate open rates on older lists, and most platforms now report a machine open filter if you look for it. Check whether yours does before you use that 38 as a baseline for anything.

On contract terms, tying vendor payment to replies or calls sounds clean and usually goes badly, because the vendor doesn't control your inventory or your phone answering. A workable middle is to keep the retainer but define the deliverable as sends that each contain a specific property and a single action, then track replies as your own metric. You keep the ability to judge the work. They keep the ability to do it.