How Does Twitter Make Money? Ads, Data and Subscriptions

Twitter, now called X, makes money in three ways: advertising shown in the timeline, paid subscriptions to X Premium, and charging developers to use its API. Advertising is the line a free public network is built around, and the other two add to it. This post maps each line to a mechanism, and says which ones a smaller network can copy and which ones need scale first.

We cite no revenue, profit or user figures. We did not open a primary page that states them, and a number repeated from a secondary source would be a guess. What we can describe are the mechanisms the company documents on its own pages. If you want to run a network with the same set of lines under your own name, the white-label Twitter clone models six of them.

The short answer

A free public microblog sells attention first. Members post and read for nothing, which builds an audience, and the network sells that audience's attention to advertisers. Then it sells a better experience to members who want one, which is the subscription line. Finally it sells access to the data and the platform to developers who want to build on top.

The same structure appears on other networks. Our guides on how TikTok makes money and how creator platforms make money compare the pattern across formats.

Advertising: the timeline is the ad surface

X's Home Mixer documentation lists the last stages of building a timeline. After posts are scored and filtered, non-post elements, namely advertisements, follow recommendations and prompts, are mixed in with the results. That one line tells you where ads live: in the same list of items that members read, produced by the same pipeline.

X's ad products and prices change often, so we do not describe them here. The general shape of ad products on a network like this is well understood, and the table below labels it as general practice rather than X's documentation.

Two features of the buying side are worth knowing. First, most are bought against a budget with a daily cap, so an advertiser can stop spending without a contract. Second, many are priced by auction, where several advertisers compete for the same slot and the price moves with demand. Both of those need many advertisers and many viewers to work well, which leads to the scale question below.

A worked example

Say a small network sells a promoted post at 5 per thousand impressions, and a brand buys 200,000 impressions. That is 1,000 of revenue for the campaign. Now say your members see 40,000 timeline pages a day and you show one promoted slot on one page in ten. That is 4,000 ad impressions a day, so it would take fifty days to deliver that single campaign. The numbers are invented and the price is arbitrary. The point is that ad income is volume times price, and a small network usually lacks the volume, which is why it starts elsewhere.

Subscriptions and verification

The second line is selling members something better than the free product. X's Premium page, which we verified as of October 2026, describes three tiers: Basic, Premium and Premium+.

• Premium adds a checkmark after an eligibility review, reduced ads (X says approximately 50 percent fewer in the For You and Following timelines), access to apply for Original Content Rewards and Creator Subscriptions, and ID verification.
• Premium+ is ad-free across most of X.
• Pricing is localized, and the page says it starts at 3 dollars a month or 32 dollars a year on the web.
• Premium Business and Premium Organizations exist for businesses, governments and nonprofits.

Look at what is for sale. Fewer ads is a product that only makes sense if ads exist, so the subscription and the ad line are linked: members who pay for fewer ads give up some ad income in return for a direct payment. That trade is sensible only when enough members pay more than the ad income they remove. It is a decision to model before you copy it. Our guide to X verification and paid tiers goes through how the ladder is built and why the badge needs a review step.

Creator programs

The same page describes Original Content Rewards, with payouts calculated from qualified impressions from Premium users, and Creator Subscriptions. These are not revenue lines so much as spending to keep creators posting, funded by the subscription line. The loop is simple: Premium members' engagement funds the rewards, and the rewards keep the creators whose posts bring members in. For the mechanics, see how X creator payouts work.

Data and developer access

The third line is access. X's API documentation describes a pay-per-usage model: developers buy credits and each request deducts from them, and the same resource requested twice in 24 hours is charged only once. It is a usage meter, not a fixed plan.

The logic is general. A public network holds posts, profiles and graph data that other businesses want to read, search or post to. Selling controlled access turns that into revenue, and it also turns it into a risk, because anything you sell can be scraped or misused if keys are loose. A smaller network can borrow the idea without the scale: issue scoped keys, rate-limit each one, hash the keys in storage and charge per tier. Our platform ships OAuth apps and per-key rate limits for this, with eleven route families under one API.

Why scale matters for ads

Ads are the line that needs the most audience, for four reasons.

1. Auction depth. An auction needs several buyers per slot to push the price up. With two advertisers there is no competition and prices stay low.
2. Measurement. Advertisers want reports on impressions, clicks and results. Producing them reliably needs tracking, de-duplication and fraud checks.
3. Brand safety. A brand does not want its ad beside abuse. You need a visible report queue and moderation record to show them. Our guide to ads and sponsorship on a creator platform goes into what sponsors ask.
4. Sales. Self-serve tools help, but large accounts expect a person. That is a cost before the first dollar arrives.

The result is a pattern: networks start with direct money from members and creators, then add sponsorships sold by hand, and only later self-serve campaigns. X's range of tiers, from reduced ads to ad-free, comes from a network that already has ads at scale and can afford to trade some of them for subscriptions.

What a smaller network can do instead

You do not have to copy the lines in X's order. A smaller network can run a simpler stack in this order.

1. Tips. A member pays a creator directly from a post or profile. Low friction, usually the first to move, and it signals to creators that this is a place they can earn.
2. One paid tier. Sell what members already use: more media, longer posts, a badge after review, early access. Keep one tier until you know which features matter.
3. Creator subscriptions. Fans pay creators for a paid following, and the operator keeps a share of a transaction it did not have to fund.
4. Direct sponsorships. One sponsor, one slot, sold by hand. No auction needed.
5. Typed campaigns. Promoted posts, banners and sponsored placements with budgets and daily caps, once you have attention to sell.
6. Scoped API access. Priced per tier, once other products start to depend on you.
7. White-label licensing. Deploy the same platform under a second brand for a partner or regional operator.

That is the order our own Twitter clone business model page follows. The six modeled lines are subscription tiers, ad campaigns, creator subscriptions, tips, API access and white-label licensing, and none needs a code change to switch on. The tiers in the shipped configuration are Free, Premium at 9.99 dollars and Pro at 29.99 dollars, and they are starting values you can change.

A note on app stores

If members buy subscriptions or tips inside an iOS app, Apple's App Review Guidelines treat unlocking features as in-app purchase. They also allow in-app purchase currencies to be used to tip digital content providers, and describe a route for creator content apps to monetize creators' content. Check the current text of section 3.1 before you design the checkout, and read our guide on Apple and Google in-app purchase rules for the decisions it raises.

A worked example of the stack

Say a niche network has 10,000 members. Say 2 percent buy a tier at 5 a month, which is 200 payers and 1,000 a month. Say tips add 300 a month, of which you keep 10 percent, which is 30. Say one sponsor pays 500 a month for a weekly sponsored post. That is about 1,530 a month before processor fees and tax. All of these figures are invented, and a real network will differ widely. What the example shows is the mix: on a young network most of the money comes from members and one sponsor, not from auctions.

Now compare an ad-first plan on the same network. If you sell promoted posts at the invented rate from earlier, 1,500 a month would need 300,000 impressions delivered, and you would also need an advertiser, measurement and a moderation record. The direct lines reach the same income with far less machinery.

How the lines fit together

The lines are not independent. Each one changes the others, and the interactions are where most mistakes happen.

Ads and tiers pull against each other

Every member who moves to an ad-reduced or ad-free tier takes inventory off the market. If your ad line is small, that costs little, and a subscription is worth more than the ads that member would have seen. If your ad line is your main income, a cheap ad-free tier can reduce revenue. The test is simple: compare the monthly price of the tier with the ad income a typical member generates. Say a typical member generates 0.50 a month in ad income and the ad-free tier costs 3. The tier is worth six times the ads it removes, so selling it is clearly good. Say instead that your heavy readers generate 4 a month. Then an ad-free tier at 3 loses you money on exactly the members most likely to buy it. The numbers are invented, but the check is real and takes ten minutes.

Verification and trust pull in the same direction

A badge that follows a review step makes advertisers and members more comfortable. A badge sold with no review invites impersonation, which scares advertisers. Because Premium on X adds the checkmark after an eligibility review, the review is part of the product. On your own network, treat verification as something you grant after a decision with a record behind it, whether it rides on a tier or is priced separately.

Creators bring the audience that every line needs

Ads need viewers, tiers need members who care, API access needs data worth reading, and all of them need posts. Posts come from creators and from ordinary members. Creator programs, whether rewards or subscriptions, are the cost of keeping that supply. If you copy them, set the budget from a line you already earn, not from a forecast.

Questions to ask before you copy a line

1. How many members must I have before this line pays for its own support?
2. Which line does it reduce when it grows?
3. What does it require from moderation, payments or legal review?
4. Can I switch it off without breaking the product?
5. What will members think when they notice it?

Reading a revenue mix without the numbers

You can learn a lot about how a network earns without a single revenue figure, by reading what it sells and where it places it. The table lists the product evidence we could open and what it implies. It is inference, not accounting.

If you need dated figures for a study or an investor deck, look for the company's own published statements or filings and cite the date on them. Do not copy a figure from a blog post, including this one. The structure above stays useful when the numbers move, and the numbers change every quarter.

Use the same method on your own network. Write down what you sell, where it sits in the product, who pays and what it displaces. If you cannot fill every column for a line, you are not ready to switch it on. For a longer treatment of how operators price and sequence these lines, see our guide to choosing a platform commission rate, which covers the share an operator keeps from creator income.

Takeaways for an operator

• Know which line you need first, and pick the one that needs the smallest audience.
• Treat the feed as an ad surface only after the feed is trusted.
• If you sell fewer ads as a perk, model what it costs before you launch it.
• Pay creators from a line that can afford it, not from hope.
• Sell API access with scoped keys, rate limits and a clear data policy.
• Write your terms before you take money, including refunds and what happens when a plan lapses.

Glossary

• Impression: one showing of an item to a viewer.
• Auction: a system where advertisers bid for the same slot.
• Daily cap: the most a campaign may spend in a day.
• Entitlement: a right a plan grants, checked on the server.
• Scoped key: an API key limited to specific actions.
• Brand safety: an advertiser's concern about where its ad appears.

What to decide next

List your revenue lines in the order you can switch them on, and next to each write the audience it needs. Start with the first line that your current audience supports. If you want those lines built into a working network under your own brand, look at the Twitter clone platform, and for the setup cost, the Twitter clone development cost page and the pricing page state the published price.