Platforms & Channels

X Ads

Also called Twitter Ads

By the AdFlint research team · Last reviewed November 2026

Advertising on X, the platform formerly called Twitter Ads, sold against real-time conversation, keywords, follower lookalikes, and trending topics.

The ad platform for X, still widely referred to by its former name Twitter Ads, built around timeline placements targeted by keywords, followers of chosen accounts, and live conversation. It fits real-time, news-adjacent, technology, and finance audiences. The practical issue is measurement and brand-safety scrutiny since the platform's changes, so test with small budgets and verify conversions independently.

Key takeaways

  • X Ads is the ad platform for X, formerly Twitter, and the old name Twitter Ads is still common.
  • Interest and follower look-alike targeting are additive, so each one widens the audience, while demographic filters narrow it.
  • Real-time targeting can help launches and events but also needs attention to where ads appear.
  • Verify conversions with your own analytics and start with small test budgets.

In practice.

X Ads is the advertising platform for X, the service formerly called Twitter. The older name, Twitter Ads, is still widely used, and X's help pages for targeting still carry material from the Twitter era. Advertisers promote posts and accounts in timelines and search, choosing audiences by keywords, followers of chosen accounts, conversation topics, interests, and custom lists. Campaigns are created in X Ads Manager, where you pick an objective, set a budget and bid, and build ad groups with targeting and creative. X Ads Manager lets you choose an objective first, so decide whether the goal is clicks, followers, video views, or conversions before building targeting.

Several targeting types work differently from other platforms. Follower look-alike targeting lets you enter account handles, and X finds people whose interests resemble those accounts' followers. X's help page suggests using around 30 handles and choosing accounts closely related to your business. Keyword targeting reaches people who searched for, posted, or engaged with posts containing chosen words, and you can exclude words. Conversation topics group content into categories and topics. Per X's help page, interests and follower look-alikes are additive, so each one widens the audience, while demographic settings such as location, age, gender, and language narrow it. Excluded keywords are always respected, even with Audience Expansion on, so add exclusions for words that would bring the wrong audience.

X Ads interacts with the real-time nature of the platform. A keyword campaign tied to an event, product launch, or news cycle can reach people while the conversation is live, but it can also place your ad beside content you did not expect. Audience Expansion builds a seed audience from your targeting and finds look-alikes, while always respecting excluded keywords, custom audiences, and do-not-reach lists. Budgets and bids work through an auction. Because interest and follower targeting are additive, stacking many of them produces a large, loose audience rather than a precise one, so start with one idea per ad group and add only what you can measure. Posts that read like a normal post, with a specific claim and a clear link, usually fit the timeline better than banner-style copy.

The practical issues are measurement and brand safety. Audience estimates and targeting options on X have changed over time, so check current numbers in your dashboard. A common mistake is treating reported conversions as final. Verify with your own analytics, such as UTM-tagged sessions and orders. A second mistake is targeting broad keywords like marketing and getting a vague audience. A third is running promoted posts that read like ads rather than posts. Plain, specific text with a clear link performs better than slogans, and a short test shows whether it works. Because keyword targeting follows live conversation, review it during events and trends, since a keyword can change meaning overnight.

This week, run a small test with one objective, one ad group per idea, and a modest daily budget. Use ten to thirty specific keywords or follower look-alike handles that show real buying or interest signals, and exclude obvious misreadings. Write two posts: one that makes a specific claim and one that asks a question. Tag links with UTMs, then compare clicks, on-site behavior, and orders in your analytics rather than in X's dashboard alone. Review where the ads appeared and read the replies. If the results do not beat your other channels at a small scale, do not scale up. Keep a record of which handles and keywords you tried and what each returned, so each test builds on the last one.

Worked example

A product launch test with keyword targeting

Suppose a developer-tools company spends $400 over a week on X Ads for a new command-line tool. It targets 25 keywords tied to the problem the tool solves, plus 20 follower look-alike handles of well-known developer accounts. The campaign earns 40,000 impressions, 400 link clicks, and a cost per click of $1.00. The click-through rate is 1 percent.

UTM-tagged sessions in the company's analytics show 310 visits, and 31 people install the tool, so cost per install is about $12.90. The dashboard claimed 38 installs, a gap of 7, so the team uses its own count. Because the 20 handles brought most clicks, they remove the weakest keywords and test a second post with a short demo clip. They scale only if install cost stays near $13 for another week.

X Ads compared with

The settings this gets confused with, and how to tell them apart.

Common questions.

Is X Ads the same as Twitter Ads?

Yes. X Ads is the current name for the advertising platform previously called Twitter Ads, and some help pages and tools still use the old name.

How does follower look-alike targeting work on X?

You enter account handles and X targets people whose interests resemble those accounts' followers. X's help page suggests using around 30 handles that are closely tied to your business.

Does stacking interest and follower targeting narrow or widen my audience?

They can widen it. X's help page describes interests and follower look-alikes as additive types, while demographic settings such as location and language narrow the audience.

How should I measure X Ads results?

Use UTM-tagged links and your own analytics or sales data alongside X's dashboard, then compare the two. Differences are common, and your own records are the safer basis for budget decisions.

Sources

You should not need to know this to advertise.

AdFlint handles the settings for you, inside the Google and Meta accounts you already own.

Next step: Run Google Search and Meta ads from one place

Try AdFlint free