Conversations Optimization

Conversations optimization tells Meta to deliver an ad set to people likely to start a message thread with your business instead of clicking through to a website. It belongs to click-to-message ads, where the call to action opens Messenger, Instagram Direct, or WhatsApp. In the Marketing API the matching optimization goal is named CONVERSATIONS, and Meta's developer page for click-to-Messenger ads lists the Engagement, Sales, and Traffic objectives as supported for those ads. In Ads Manager the same choice appears as a messaging destination in the ad set, with conversations as the performance goal. Check the objective too: Meta's developer documentation lists Engagement, Sales, and Traffic as the objectives that support click-to-Messenger ads.

Mechanically, the ad carries a welcome experience. Meta's developer documentation gives the default greeting as Hello! Can I get more info on this? and lets you replace it with your own welcome text, icebreakers (tappable starter questions), and autofill messages. Icebreaker titles are capped at 80 characters, and message text at 300. When someone taps the ad, a chat opens with that greeting prefilled, and Meta counts a result when they send a first message. The system learns from people who did this recently and looks for similar users, which favors people who habitually message businesses. You can edit the greeting in the ad itself, so test a version that mentions the offer against the default text and compare how many people actually write back.

This goal interacts with your account setup more than most. Messaging apps need to be connected to your Page or business accounts, and someone has to be present to answer. Because the result is the first message and nothing after it, bid strategy reports cost per conversation started, not cost per sale. Conversion tracking can be extended with the Conversions API for Business Messaging, which lets you send events back from chat, such as a lead qualified or a purchase, so Meta can see what happens after the chat. Without that, the model optimizes only for the opening message, and quality varies widely across audiences. Decide in advance what counts as a good conversation for your business, such as a message that includes a date, a size, or a budget, and send that event back when it happens.

The classic mistake is treating a high conversation count as success while the inbox goes unanswered. An unanswered thread still counts as a result, so cost per conversation can look excellent while leads go cold. Response time matters because people who tap an ad expect a quick reply. Another mistake is using it in a category where buyers expect forms or checkout, such as low-priced consumer goods, where chat adds friction. A third is writing a welcome message that is vague; icebreakers that name the real question, like asking about pricing or booking, filter out idle taps. Another mistake is sending every thread to one overworked person. If replies take a day, people move on, and the cost you paid for the click is wasted.

This week, count conversations in Ads Manager and match them to the threads your team actually answered. Measure median first-response time, the share of threads that got a reply within an hour, and how many became bookings or sales. If fewer than half get answered quickly, fix staffing or add automated first replies before raising the budget. Then rewrite the icebreakers around the three questions customers really ask, and compare cost per qualified conversation, not just cost per conversation, across two audiences before deciding where to put more money. Keep a short script of answers to the five most common questions, and review a sample of threads each week to see what the ad is attracting.