X Ads metrics are only useful when they match the objective of the campaign. For awareness, watch reach, frequency and cost per thousand impressions. For engagement and traffic, focus on link clicks, cost per link click and landing page behaviour. For conversion campaigns, the numbers that matter are conversions tracked by the X Pixel or Conversion API, cost per conversion and return on ad spend, checked against your own analytics.
Where do you find X Ads metrics?
All campaign data sits in X Ads Manager (ads.x.com). The main campaign view shows default columns, but you can switch between metric presets such as Engagement, Website conversions and Video, or build a custom set of columns. Data can be broken down by campaign, ad group and ad, and segmented by placement, device, location, age and gender where available.
For deeper reporting, export data as CSV from the Analytics or reporting section, or pull it through the X Ads API into a reporting tool. Set the attribution window before you compare periods; changing it between reports makes numbers incomparable.
Which metrics matter for each campaign objective?
X offers objectives including Reach, Video views, Pre-roll views, App installs, Website traffic, Engagements, Followers and Website conversions (sales). Each one optimises delivery for a different action, and judging a campaign on the wrong metric is the most common reporting error.
- Reach campaigns: impressions, reach, average frequency and CPM. Clicks are not the goal here.
- Video views: video views (X counts a view at 50 percent of the player visible for two seconds or more, or a click to expand), view rate, completion rates at 25, 50, 75 and 100 percent, and cost per view.
- Website traffic: link clicks, cost per link click, click-through rate on links, plus sessions and bounce behaviour in your analytics.
- Engagements: engagement rate and cost per engagement. Treat these as signals of creative interest, not business results.
- Website conversions: conversions by event type, cost per conversion, conversion value and ROAS.
- Followers: new followers and cost per follower, but also how many of those followers engage later.
Engagements versus link clicks
X reports “engagements” as any interaction with the ad: clicks on the profile, hashtag, media expansion, likes, reposts, replies and link clicks. A post can have a high engagement rate and almost no link clicks, because people expanded the image or tapped the profile.
If your goal is traffic, add the “Link clicks” column and calculate cost per link click yourself if needed. Then compare link clicks in Ads Manager with sessions from X in Google Analytics or another analytics tool. A large gap usually points to slow landing pages, accidental clicks on mobile or a broken UTM tag.
How should conversions be tracked on X?
Conversion metrics are only as reliable as the setup behind them. Install the X Pixel through Events Manager, or through Google Tag Manager, and define events such as PageView, AddToCart, Lead and Purchase. Pass the order value and currency with purchase events.
Add the Conversions API (server-side tracking) where possible. Browser tracking loses events to ad blockers, cookie consent choices and browser privacy features, and server events help fill those gaps. Use event deduplication so a purchase reported by both pixel and server counts once.
Attribution windows
X lets you choose post-engagement and post-view windows for conversions, for example 30-day post-engagement and 1-day post-view. View-through conversions are counted when someone saw the ad without clicking and converted later. They can be real influence, but they also inflate results. Report click-based and view-based conversions in separate columns so stakeholders can judge for themselves.
Metrics that help diagnose problems
Some metrics are not goals but explain why a campaign is under-performing:
- Frequency: if it climbs while click-through falls, the audience has seen the creative too many times.
- CPM trend: rising CPMs can mean more competition for your audience or a narrow targeting pool.
- Click-through rate by creative: shows which message or image earns attention.
- Video quartile drop-off: if most viewers leave before 25 percent, the opening seconds need work.
- Placement split: results in the Home timeline, Search results and profiles can differ; check before excluding any.
- Spend by device: a desktop-heavy B2B audience may behave differently from mobile.
What makes an X Ads report defensible?
A defensible report ties platform numbers to business numbers and states its assumptions. Finance and leadership teams tend to distrust reports that only show likes and impressions.
Build monthly reports around three layers. First, delivery: spend, impressions, reach and CPM. Second, response: link clicks, cost per link click, video completion and engagement rate. Third, outcome: conversions, cost per conversion and revenue as seen in your own analytics or CRM, alongside what X reports. Show the difference between the two outcome figures and explain it rather than hiding it.
For example, a B2B software company spending a fixed monthly amount on X might report 40 demo requests in Ads Manager and 28 in the CRM with an X source. The honest summary is that X drove between 28 and 40 demo requests, with the gap explained by view-through attribution and leads who later came back through search.
Testing to improve the numbers
Use X’s A/B testing feature, available when creating campaigns, to split audiences cleanly between two variants. Test one variable at a time, such as creative, audience or bid strategy, and run the test long enough to collect a meaningful number of results for each side.
Good test candidates on X include conversational text versus feature-led text, image versus short vertical video, keyword targeting versus follower look-alike targeting, and automatic bidding versus a target cost.
Benchmarks: use your own, not someone else’s
Published click-through and CPM averages for X rarely match a specific account, because they mix industries, countries and objectives. Build an internal benchmark instead. Take the last three to six months of your own campaigns, group them by objective, and record the median CPM, cost per link click and cost per conversion for each group.
New campaigns are then judged against that median. A traffic campaign running 30 percent above your own median cost per link click needs attention; one running below it is a candidate for more budget. Update the medians every quarter, since seasonality and auction pressure shift over time.
- Keep separate medians for prospecting and retargeting.
- Exclude test campaigns with very small spend, which distort the figures.
- Note major changes in tracking, since they break comparisons with earlier periods.
Where to start this week
Open Ads Manager, pick your highest-spend campaign and check that its columns match its objective. Add link clicks and conversions if they are missing. Then compare last month’s X-reported conversions with the same period in your analytics or CRM, and write down the gap. Fixing tracking or attribution settings before the next test will make every future number easier to trust.
Keep reading: X Ads (Twitter Ads) · Social Media Marketing