X Lead Gen Ads: the native form is back

#Quick answer

X Lead Gen Ads are promoted posts with a call-to-action button that opens a form inside X. The person taps, fills it in without leaving the app, and you collect the submission. X announced the global launch on 25 August 2026, nine years after Twitter retired the original lead generation card. The format works on image, video, carousel and Timeline Takeover ads, and Ads Manager now has a dedicated Leads objective plus a leads centre that exports to a CRM.

Last updated: October 2026

#The format, precisely

Strip the marketing language and a Lead Gen Ad is three pieces glued together.

There is a creative - an ordinary promoted post with an image, a video or a carousel. There is a button at the bottom of it, which X colours to match the dominant colour of the media so it reads as part of the ad rather than a bolted-on widget. And there is a form that opens in a sheet over the timeline when someone taps that button.

The form is the whole point. Every other performance format on X eventually hands the person off to your website, where a page has to load, a layout has to work on their phone, and they have to retype an email address they have typed a thousand times. Each of those steps loses people. The native form deletes all three.

X's pitch for it is blunt: "The result is more leads, higher lead quality, and lower cost per lead." Two of those three are mechanical consequences of removing the handoff. The third is the one worth arguing about, and we will come back to it.

#This existed before, and X killed it

Worth knowing, because the first version's death explains the second version's design.

Twitter shipped lead generation cards years ago and then shut them down. The announcement landed in December 2016: from 1 February 2017 you could no longer create or edit lead generation campaigns or the cards themselves, and from 1 March 2017 the cards stopped being viewable at all. The company's own line at the time was that it intended "to focus our efforts on building and improving other performance offerings." The replacement it was excited about was chatbots, welcome messages and quick replies.

That is the context to hold. The original format was not retired because people hated filling in forms. It was retired because the platform's ad product went somewhere else, and by the time X wanted performance advertisers back, the feature was nine years gone and the competition had kept iterating.

So the 2026 version is not a restoration. The pre-fill, the dedicated objective and the CRM path are all additions, and they are the parts that matter.

#What is new in the 2026 version

X names the predictable use cases: booking demos and appointments, growing an email list or waitlist, collecting quote requests, sweepstake entries, and filling webinars, events and classes. Monique Pintarelli, head of global advertising, framed the value as momentum: "Lead Gen Ads simply make it easier for them to take the next step without breaking that momentum when an ad has caught their attention."

That is the honest version of the benefit. It is not that X's audience suddenly wants to hear from you. It is that you are no longer asking them to change apps mid-thought.

#The API, which is where the real constraints live

X documented the Lead Generation API on 15 September 2026, and the documentation is more useful than any of the launch coverage, because it tells you exactly what a form is allowed to be.

A form has six parts. name (your internal label, 1 to 255 characters) and purpose (what the leads are used for, up to 1000 characters) are required. The fields array is required, keys must be unique, and at least one field has to be non-optional - you cannot ship a form that collects nothing. A privacy_policy block is required and must carry an https:// URL plus a company name. The introduction card shown before the fields and the final_page thank-you screen are both optional, though if you put a call-to-action on the final page, X requires both its url and title.

The documented field keys split into two groups:

• Contact: EMAIL, PHONE_NUMBER, STREET_ADDRESS, CITY, STATE, COUNTRY, ZIP_CODE
• Person and firm: FULL_NAME, FIRST_NAME, LAST_NAME, DATE_OF_BIRTH, COMPANY_NAME, JOB_TITLE, WORK_EMAIL, WORK_PHONE_NUMBER, GENDER, HANDLE

Attaching the form to creative is a card operation. You create a card and set a component's destination type to LEAD_FORM with the form's id, which means everything you already know about cards applies here - if your link cards render wrong, the lead form is riding the same machinery, and the X card validator page covers how to test that. LEAD_FORM is valid anywhere a destination is accepted, including individual buttons in a button group.

Then there are three constraints that will bite you later if you skip them now:

1. Exports are admin-only. The Leads API needs the Campaign user permission on the ads account, and lead exports are restricted further, to account administrators, because the exported files contain personal information. If your agency has campaign access but not admin, it can run the ads and cannot retrieve the leads.
2. Updates replace, they do not patch. A PUT on a lead form takes the same parameters as create and every one is optional, but the nested objects (introduction, final_page, privacy_policy) and the fields array are replaced wholesale when you send them. There is no partial update inside a nested object, and no documented way to unset one once it is set.
3. Delete means archive. Deleting a form soft-deletes it. Archived forms stop accepting submissions and cannot be updated at all, and update attempts come back as not-found. Pass with_deleted=true to see them again.

Exporting is an async job: create it, poll it, download the file. The window is [start_epoch, end_epoch) - start inclusive, end exclusive - which is the sort of detail that quietly double-counts or drops a day's leads if you assume otherwise when you schedule a nightly pull.

#What it costs, and where the cost actually moves

There is no separate fee for the format. Lead Gen Ads run in the same auction as every other promoted post, so what you spend is media cost, priced the way the rest of X ads cost is priced.

What changes is the conversion rate on the step you were already paying for. You were buying taps. Previously a tap bought you a page load and a form on your own site, and you lost people at both. Now the tap buys you a form that is already open and, for repeat submitters, already filled in. Cost per click does not have to improve for cost per lead to drop a lot.

This is the same structural move X made with X Lift, its URL-to-campaign ad builder: take a step out of the funnel and the funnel converts better. For context on where this format sits among the others, what an X ad is walks the full format lineup.

#The lead quality question

Here is the part the launch coverage repeats without examining. X says Lead Gen Ads deliver "higher lead quality." More leads and lower cost per lead follow directly from removing friction. Higher quality does not.

Friction is a filter. Someone who loads your landing page, reads it, and types their work email has demonstrated something. Someone who tapped a coloured button and submitted a form that was already filled in has demonstrated considerably less. Both land in your CRM looking identical.

The pre-fill makes this sharper rather than softer, because the second and third form a person submits cost them almost nothing. That is excellent for volume and neutral at best for intent.

Two things follow, and they are cheap to do:

• Make one field expensive. Not more fields - one field that requires a thought. A free-text "what are you trying to solve" beats three more pre-fillable contact fields for sorting the list, because the pre-fill cannot answer it for them.
• Measure past the form. Judge the campaign on booked calls or qualified opportunities, not on lead count. Set up the conversion side before you turn the budget on, so the number you optimise against is the one you actually care about.

None of that is an argument against the format. It is an argument against reading "lower cost per lead" as "better pipeline" when the two can move in opposite directions.

#Who should bother

This format earns its place when you have a single, specific, low-commitment ask and a process waiting on the other side. A webinar seat, a demo slot, a quote request, a waitlist. Those are exactly the cases where a landing page adds a step and no information.

It earns less when the next step is genuinely complex, or when your product needs explaining before anyone should be giving you contact details. A form is not a pitch. If the ad has to do the convincing in one creative, you are back to needing a page.

And if your actual constraint is that too few of the right people see you on X at all, lead capture is the wrong lever. That is a distribution problem, and the fix is consistent presence in the conversations your buyers already have. X-Autopilot runs that side from your own Mac, in your own browser session, which keeps a lower detection surface than cloud tools that log in from a shared IP pool. It is not a sanctioned-API product and we do not claim it is compliant with X's terms - browser automation is a grey area and X's automation rules are the document to read before you decide. Paid lead capture and organic presence solve different problems; running one does not excuse skipping the other.

#The bottom line

Lead Gen Ads are a competent rebuild of a format X should never have dropped. The native form removes two real sources of loss, the Leads objective means the auction is optimising for the thing you want, and the leads centre plus the Ads API give you a sane path into a CRM. For a demo booking or a webinar signup, it is now the shortest route X offers.

Go in with the permissions sorted, because exports are administrator-only and discovering that after the campaign is live is a bad afternoon. Treat the [start, end) export window literally. And hold "higher lead quality" at arm's length: you removed the step that used to do your qualifying, so the qualifying has to happen somewhere else now. Put one unfillable question on the form and judge the campaign on what happens after the submission, not on the submission count.