# X Ads explained: costs, targeting and what results to expect

> By Martin Vos, Founder, X Ad Campaign. Published 2026-09-19. 11 min read. Explainer.
> https://www.xadcampaign.com/blog/x-ads-guide/

X Ads is the paid side of X, formerly Twitter. Everything about it comes down to four choices: what you pay for, who sees the ad, what they see, and how the result is measured.

**TL;DR.** X Ads sells attention through an auction: you choose an objective, an audience and a budget, and pay for the result that objective names. Targeting reaches people by place, interests, keywords and your own lists. Benchmarks put clicks and impressions on X below Meta’s prices, but results depend on fit and measurement. X Ad Campaign runs campaigns inside the client’s own account and starts with a measured first month.

## What are X Ads?

X Ads is the advertising system of X, formerly Twitter. It places paid posts, videos and a handful of special formats in the places people already look: the home timeline, search results, profiles and the replies under popular posts. X Ad Campaign runs these campaigns for businesses inside their own X Ads account, so the account, the data and the history stay with the business.

Every campaign is built from the same four decisions. The **objective** says what you are paying for: reach, video views, clicks to a website, app installs, sales. The **audience** says who can see the ad. The **ad** is what they see. The **measurement** says whether it worked. The platform's auction then decides, many times a second, which advertiser's ad a given person sees and what that impression costs.

Most of what goes wrong with X Ads goes wrong in one of those four places: an objective that rewards the wrong action, an audience that is too narrow to spend the budget or too broad to matter, creative that does not fit a fast timeline, or tracking that cannot tell a sale from a scroll. The sections below take each decision in turn, then turn to prices, results and the cases where X is the wrong place to spend.

## How does the X Ads auction decide who sees your ad?

Every time someone opens a timeline or a search on X, a set of ad slots is up for sale. The advertisers whose targeting includes that person compete for the slot, and the platform weighs each bid against how likely the ad is to get the response it is optimising for. Hootsuite describes it as a second-price auction, in which the highest bidder pays a cent more than the next bid[^1], not its own maximum.

Two practical consequences follow. First, you rarely pay your full bid: the price is set by competition for the same person at the same moment, which is why costs rise around big events and fall in quiet weeks. Second, a better ad wins more cheaply. An ad that people engage with is worth more to the platform per impression, so it can win slots at a lower price than a weaker ad with a higher bid.

That is why bidding is mostly left to the platform's automatic options in the first weeks of a campaign. Manual bids and cost caps are useful once you know what a result is worth to you; before that, they mostly stop a campaign from learning.

## Which objective should a campaign use?

The objective is the most important setting in the account, because it tells the auction what to look for and decides what you pay for. Hootsuite lists nine objectives[^2], grouped by what they buy:

| You want | Objective | What you pay for |
|---|---|---|
| people to see the brand or a message | Reach | impressions, per thousand[^3] |
| people to watch a video | Video views or pre-roll views | views[^3] |
| people to visit a website | Website traffic | impressions, optimised for clicks[^3] |
| people to install an app | App installs | clicks to the app store[^3] |
| replies, reposts and likes | Engagements | engagements, per thousand[^3] |
| purchases from a catalogue or website | Sales | impressions, per thousand[^3] |

The rule of thumb is to choose the objective closest to the result you will judge the campaign by. The same logic is why a followers campaign beats [buying followers](/blog/buying-x-followers/): it pays for real people, chosen by targeting. If the campaign will be judged on sales, an engagement objective will find people who like posts, not people who buy, and its low cost per engagement will hide that. If the campaign is about awareness ahead of a launch, a sales objective will spend slowly and look like it failed. The deeper treatment of each objective is in the [campaign objectives guide](/guides/campaign-objectives/).

## Who can you target with X Ads?

X targeting works in two layers: who people are, and what they are paying attention to. The first layer covers the familiar settings: country, gender and age range[^4], plus language and device. The second layer is where X differs from most platforms: keywords and conversation topics[^4] reach people by what they post, search for and read about, close to the moment it happens.

On top of both sit your own audiences. Custom audiences and retargeting[^4] let you upload customer lists, reach people who visited your website, or follow up with people who watched your videos or used your app. For most businesses, these are the audiences that convert best, because the people in them already know the brand.

A few habits make targeting work better:

- **Start broad enough to spend.** An audience so narrow that the daily budget cannot be spent never leaves the learning stage. Hootsuite recommends at least two dozen keywords[^4] per keyword campaign for the same reason.
- **Exclude existing customers** from acquisition campaigns, and include them in retention ones.
- **Separate audiences that behave differently** into their own ad groups, so one does not take the whole budget and hide the other.
- **Check the reachable audience before launch.** The ads manager estimates it for each setup, which answers the "is my market even on X?" question without spending anything.

The [targeting topic](/blog/topics/targeting/) collects the detailed posts on keywords, follower lookalikes and custom audiences, and [B2B marketing on X](/blog/b2b-marketing/) shows how account lists and conversation targeting combine for longer sales cycles.

## What ad formats does X offer?

Most X ads are ordinary posts that are paid to be seen: a line of text with an image, a video or a link card. Hootsuite lists promoted posts with text, images or video, carousels, cards, polls, pre-roll and live video, GIFs and dynamic product ads[^5] among the options.

The format follows the objective more than taste. A website traffic campaign usually runs on a post with a website card, because the whole card is clickable. A video views campaign needs a video that makes sense with the sound off in its first seconds. An app campaign uses an app card that opens the store. Sales campaigns for shops can use dynamic product ads that pick the product from a catalogue for each person.

What matters more than the format is how the ad reads on a fast timeline. The ads that work look like good posts: one idea, a clear first line, an image or video that shows the thing rather than a logo. The [ad formats guide](/guides/x-ad-formats/) lists the formats with their specifications, and the [creative guide](/guides/creative-that-works/) covers how to write and design for them.

## What do X Ads cost?

X does not sell ads at a list price. You set a budget per day or for the whole campaign, and the auction sets the price per result. So the useful question is what results typically cost, and the published benchmarks give a range rather than a number.

| Benchmark | What it says | Source |
|---|---|---|
| cost per thousand impressions | around two dollars, below Meta's[^6] | Hootsuite |
| cost per click | under a dollar, about half of Meta's[^6] | Hootsuite |
| cost per click | from half a dollar to two dollars, more around big events[^7] | WebFX |
| cost per click | under twenty cents across agency accounts[^8] | Enrich Labs |
| cost per acquisition | a little over twenty dollars as a median[^8] | Enrich Labs |

The spread is not an error. The benchmarks come from different samples, objectives and industries, and an engagement campaign's cost per click is not comparable with a sales campaign's. What they agree on is that attention on X is relatively inexpensive, especially compared with Meta, and that prices rise for competitive topics and live events. The detailed breakdown, including why the benchmarks disagree and how to set a first budget, is in [what X ads cost and why the benchmarks disagree](/blog/x-ads-cost/).

## What results should you expect from X Ads?

Small percentages, which is normal for social advertising. WebFX reports a typical click-through rate of one to two percent and an engagement rate of half a percent to one percent[^9]. Enrich Labs, working from a large pool of agency accounts, reports an engagement rate of a little over half a percent[^8].

Conversion rates vary much more, because they depend on the landing page and the offer more than on the platform. Enrich Labs reports a very low conversion rate[^8] across its accounts; WebFX quotes a range of one to three percent[^9] for its own. Neither number predicts yours: the only conversion rate worth planning with is the one your own tracking reports once the first weeks are in.

Who sees the ads shapes the results too. In the U.S., Pew Research Center finds that about a fifth of adults[^10] use X, with the highest share among adults under thirty and the lowest among those over sixty-five[^11]. An offer for younger, professional or news-following buyers starts with an advantage; an offer for older consumers starts with a thin audience. Whether that adds up to a good fit for a particular business is covered in [is X good for advertising?](/blog/is-x-good-for-advertising/).

## How do you measure X Ads properly?

By deciding what a result is before the campaign starts, and making sure it is counted once. For a shop that is a purchase, for a service business a qualified lead, for an app an install followed by a meaningful action. Engagements and clicks are signals along the way, not the result.

Measurement on X runs through the X pixel on the website or through a server connection (the conversions API), and through the app's measurement partner for app campaigns. In Europe, both depend on consent: a visitor who declines tracking is not counted, which makes reported conversions an undercount rather than an exact figure. That is fine as long as everyone reading the report knows it.

The practical test of a setup is simple: make a test purchase or submit a test lead, and check that exactly one conversion appears in the ads account, attributed to the right campaign. A campaign without that check is optimising on guesswork. The [measurement guide](/guides/measuring-results/) covers the pixel, the conversions API and attribution windows in detail.

## How do you start with X Ads?

With one campaign, set up as a test. No subscription is needed for this: the ads account is separate from [X Premium](/blog/x-premium/). Open an X Ads account, add a payment method, install measurement, and build a single campaign with one objective, one audience broad enough to spend the budget, and three or four versions of one message. Let it run for a few weeks before judging it, refresh the creative before engagement falls, and compare the cost per result with the channels you already use.

The full sequence, from account to first campaign, is in [how to advertise on X: a first campaign, step by step](/blog/how-to-advertise-on-x/). Businesses that would rather not build it themselves can have X Ad Campaign set up and run the first month in their own account; either way, the first month is a test, not a commitment. What to ask any agency before handing over a budget is in [how to choose a digital marketing agency](/blog/digital-marketing-agency/).

## When are X Ads the wrong choice?

When your buyers are not on X. If your customers are mainly older consumers, or a professional group that does not use the platform, no setting will create an audience that is not there. The estimated audience size in the ads manager shows this before anything is spent.

When the offer needs space. Products that sell through a visual catalogue, a detailed comparison or a consultation usually do better where people arrive already searching. X can support those, mostly through retargeting, but rarely carries them alone.

When the result cannot be measured. If sales happen offline or months later and cannot be tied back to the campaign, X Ads becomes a cost you cannot judge. Fix measurement first, or choose a channel where it already works.

When the budget is too small to learn. A budget split across many audiences and objectives produces a little data about everything and a clear answer about nothing. One focused test, or waiting until the budget allows it, is a better use of the money.

## Quick answers

### What are X Ads?

X Ads is the advertising system of X, formerly Twitter: promoted posts, videos and other formats shown in the timeline, search results and replies to people you choose by location, interests, keywords and your own customer lists. You pay through an auction for the result your objective names. X Ad Campaign runs these campaigns inside the client’s own X Ads account.

### How much do X Ads cost?

There is no fixed price. You set a daily or total budget and the auction sets the price per result. Published benchmarks put a click on X in the range of a few tens of cents to a couple of dollars, with prices per thousand impressions below those of Meta, but your own cost depends on the audience, the objective and how good the ad is.

### Is X advertising cheaper than other platforms?

Per impression and per click it often is, according to benchmarks that compare it with Meta. Cheaper attention is not the same as cheaper customers, though. The comparison that matters is the cost per sale, lead or install for your own offer, measured the same way on each platform, after a test long enough to settle.

### Do X Ads actually work for businesses?

They work when the buyers are on X and the offer makes sense in a few seconds on a busy timeline. They fail when the audience is elsewhere, when results cannot be measured, or when the budget is spread across too many audiences to learn anything. A measured first month shows which case applies before larger sums are committed.

## Sources

[^1]: Hootsuite describes X ads as a second-price auction in which the winner pays $0.01 more than the next highest bidder. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^2]: Hootsuite lists nine X ads objectives: Reach; Video Views, Pre-roll Views, App Installs, Website Traffic and Engagements; App Re-engagements, Sales and Keywords. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^3]: Per Hootsuite: Reach is billed per 1,000 impressions; Video Views and Pre-roll Views per view; App Installs per click to the app store; Website Traffic per impression; Engagements per 1,000 engagements; Sales per 1,000 impressions. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^4]: Hootsuite lists X targeting by country (include or exclude), gender, age range, keywords, conversation topics, custom audiences and retargeting of website visitors, video engagers and app users, and recommends 25 or more keywords per campaign. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^5]: Hootsuite lists X ad formats including promoted posts with text, images or video, carousels, cards, polls, pre-roll video, live video, GIFs and dynamic product ads. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^6]: Hootsuite reports its own benchmarks for X ads: a CPM of $2.09 against $2.53 on Meta, and a CPC of $0.74 against $1.41 on Meta. Hootsuite, "X (Twitter) ads: The practical 2025 guide for social marketers" (Michelle Martin, 19 September 2025), https://blog.hootsuite.com/twitter-ads/ (read 2026-09-30).
[^7]: WebFX puts the average CPC of X ads at $0.50 to $2, spiking to $3 to $5 for competitive topics or events, and the average cost per follow at $2 to $4, from its in-house performance data and third-party intelligence. WebFX, "X (Twitter) Marketing Benchmarks for a Winning 2027 Strategy" (14 November 2025), https://www.webfx.com/blog/social-media/x-twitter-marketing-benchmarks/ (read 2026-09-30).
[^8]: Enrich Labs reports, from data of more than 7,000 marketing agencies using AgencyAnalytics, a median cost per acquisition of $21.55, a cost per click of $0.18, a median cost per engagement of $0.13, an engagement rate of 0.59% and a conversion rate of 0.02% for X ads. Enrich Labs, "Twitter/X Benchmarks 2026", https://www.enrichlabs.ai/blog/twitter-x-benchmarks-2025 (read 2026-09-30).
[^9]: WebFX reports typical rates for X ads of a 1% to 2% click-through rate, a 0.5% to 1% engagement rate and a 1% to 3% ad conversion rate, from in-house performance data and third-party intelligence. WebFX, "X (Twitter) Marketing Benchmarks for a Winning 2027 Strategy" (14 November 2025), https://www.webfx.com/blog/social-media/x-twitter-marketing-benchmarks/ (read 2026-09-30).
[^10]: In the 2025 survey, 21% of U.S. adults say they ever use X (formerly Twitter). Pew Research Center, "Social Media Fact Sheet" (20 November 2025; survey of U.S. adults, 5 February to 18 June 2025), https://www.pewresearch.org/internet/fact-sheet/social-media/ (read 2026-09-30).
[^11]: Share of U.S. adults who ever use X by age: 33% of 18 to 29 year olds, 25% of 30 to 49, 16% of 50 to 64 and 10% of those 65 and older. Pew Research Center, "Social Media Fact Sheet" (20 November 2025; survey of U.S. adults, 5 February to 18 June 2025), https://www.pewresearch.org/internet/fact-sheet/social-media/ (read 2026-09-30).

---
*X Ad Campaign, https://www.xadcampaign.com/. Free to read, index, quote and cite with attribution and a link.*