DM automation pricing: the four meters explained

Almost every tool in this category prices on something other than the work it does. Once you know which meter you are on, comparing them stops being guesswork.

We build one of these tools, so treat the framing as interested. The meters themselves are just facts about how the category works.

The four meters

Active contacts

You pay for how many unique people the tool talked to in a billing period. Common in the creator DM tools.

What it means for you: your bill tracks your reach, not your workload. A reel that lands sends your contact count up whether or not those people were worth talking to. A large dormant audience is cheap; a viral week is expensive.

Seats

You pay per human with a login. Standard in shared team inboxes.

What it means for you: if you are one person, this is often the cheapest option available. It is also frequently the wrong product — seat-priced tools are usually built so several humans can collaborate on a queue, which is not the problem a solo founder has. Watch for seat minimums: some tools will not sell to fewer than five or ten seats, which quietly disqualifies you.

Messages, credits or conversations

You pay per unit of talking. Most “train it on your docs” chatbots work this way.

What it means for you: predictable and fair, but check what counts as one unit. A “conversation” that resets after 24 hours behaves very differently from one that lasts until resolved, and credits often expire monthly.

Outcomes

You pay per resolution. The support-agent vendors have moved here, typically around $0.50 to $1.50 per resolved conversation.

What it means for you: it sounds like the fairest model and it can be, but only if the vendor defines “resolved” rigorously — including what they don’t charge for, and whether you get credited when a customer reopens. The definition is the product. If a vendor cannot show you their exclusion list, the model is marketing.

The question that actually decides it

Not “which is cheapest” — the meters are too different to line up. Ask instead: what does my usage look like, and which meter punishes that shape?

• Big audience, low message volume? Contact pricing punishes you for people who never message. Per-action or per-message wins.
• Small audience, very chatty? Contact pricing is cheap. Per-message pricing is where you will feel it.
• Spiky — quiet, then viral? Anything with tiers punishes the spike, because you get bumped up a bracket for one good week. Usage-based with a cap is more forgiving.
• Steady and predictable? Almost anything works. Optimise for fit, not price.

Three things that quietly change the real number

Whether AI is included. In several tools the AI is a separate add-on, sometimes only available on higher tiers. The headline plan price is not what you will pay.

What happens when you go over. Some tools charge overage. Some upgrade you automatically. At least one stops replying when you hit an unpublished fair-use limit — which is an outage your customers experience and you may not notice for hours.

Whether failures are billed. Ask whether a send that failed still costs you. It should not.

What we charge, for completeness

BABAV is per-action: $0.02 per reply, with $25 of usage included every month (roughly 1,250 replies) before any overage, then a hard spending cap you set that the agent cannot exceed. Failed actions are not billed. Plans are priced by how many of the six core channels you connect.

That is the right model for spiky, conversation-heavy accounts and the wrong model if you have a huge audience that almost never messages you — in which case a contact-metered tool with a generous tier will likely beat us. We would rather you did that arithmetic before you buy than after.