{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-09-26T10:24:06.532Z"},"content":[{"type":"blog","id":"a2465d70-8776-49b1-bc4d-8d0e84bcf9b5","slug":"datadog-rum-alternatives-2026","title":"8 Best Datadog RUM Alternatives in 2026 (And the 20x You Pay to Look)","url":"https://www.koji.so/blog/datadog-rum-alternatives-2026","summary":"Datadog Real User Monitoring uses a two-part meter: RUM Measure at 0.15 dollars per 1,000 sessions on full traffic and RUM Investigate at 3.00 dollars per 1,000 on filtered sessions, exactly 20x. Session Replay adds 2.50 dollars per 1,000. On-demand runs 1.44x to 1.50x above annual. Sessions expire after 15 minutes idle and cap at 4 hours. Retention filters require committing to a hypothesis before seeing the data. Eight alternatives compared.","content":"Datadog Real User Monitoring is priced unlike almost anything else in the digital-experience category, and the structure is easy to misread in a procurement spreadsheet. There is not one meter. There are two, and they differ by a factor of twenty.\n\nWe pulled Datadog's published price list and billing documentation, worked the arithmetic, and compared eight alternatives.\n\n## The short answer\n\nDatadog RUM splits into **Measure** and **Investigate**. Measure is listed at \"$0.15\" per 1,000 sessions per month on full traffic. Investigate is listed at \"$3\" per 1,000 sessions per month on filtered sessions. That is **exactly 20x** to move a session from counted to inspectable.\n\nSession Replay is a further add-on at \"$2.50\" per 1,000 sessions per month. So a session you can measure, investigate and watch costs about **5.65 dollars per 1,000 on annual pricing**, against 0.15 dollars for one you can only count. That is a **37.7x** spread across the same underlying traffic.\n\nNone of it tells you why the user did what they did. That is a different instrument, and it is why [Koji](https://www.koji.so) leads the list below.\n\n## How Datadog RUM pricing actually works\n\nHere is the published Digital Experience price list, annual against on-demand:\n\n| Product | Annual | On-demand | Basis |\n| --- | --- | --- | --- |\n| RUM Measure | 0.15 dollars | 0.22 dollars | per 1,000 sessions, full traffic |\n| RUM Investigate | 3.00 dollars | 4.50 dollars | per 1,000 sessions, filtered sessions |\n| Session Replay add-on | 2.50 dollars | 3.60 dollars | per 1,000 sessions |\n| Product Analytics | 0.80 dollars | 1.20 dollars | per 1,000 sessions, full traffic |\n| Error Tracking | 25 dollars | 36 dollars | first 50k errors per month |\n\nTwo structural facts fall straight out of that table.\n\n**First, the on-demand penalty is consistent.** Across all five lines, not committing annually costs between 1.44x and 1.50x. Datadog is unusually uniform here, and the premium is easy to model: assume roughly a 50 percent surcharge for flexibility.\n\n**Second, Investigate is 20.0x Measure.** Not 20-ish. Three dollars against fifteen cents, exactly twenty. That ratio is the single most important number in any Datadog RUM forecast.\n\nYou also need to know what Datadog counts. Its billing documentation defines a session as \"A user journey on your web or mobile application. A session usually includes multiple page views with their associated telemetry.\" And critically: \"A session expires after 15 minutes of inactivity, and its duration is limited to 4 hours.\"\n\nThat four-hour ceiling matters more than it looks. A session is a **billing unit, not a person**. A user who keeps a tab open across a working day is not one session; they are several. Any forecast built on \"monthly active users\" will understate the bill.\n\n## The finding: a retention filter is a hypothesis you commit to in advance\n\nThe mechanism behind the two meters is RUM without Limits. Datadog ingests everything and charges the low Measure rate on all of it, then lets you write **retention filters** that decide which sessions get promoted to Investigate at the high rate. Datadog's documentation confirms it provides \"performance metrics for all ingested sessions\" regardless of those retention decisions.\n\nAs cost control this is genuinely clever, and it deserves credit. You keep aggregate metrics on 100 percent of traffic while paying the expensive rate only on the slice you care about. Compared with vendors that simply stop recording when you hit a cap, it is a far more humane design.\n\nBut there is an epistemic cost that no pricing page mentions.\n\n**A retention filter is a hypothesis you have to write down before you have seen the data.** You keep sessions with errors, or sessions from a particular account tier, or sessions that hit a checkout route. Then, weeks later, you go looking for the cause of a problem you did not anticipate, and the sessions that would have explained it were never retained. You can only investigate what you already suspected was worth investigating.\n\nThat is a structurally different limitation from a quota blackout, and arguably a subtler one. When [Mouseflow stops tracking](https://www.koji.so/blog/mouseflow-alternatives-2026) at its refill date, you know you have a gap. When a retention filter silently declines to keep the session that mattered, the dashboard looks complete. Unknown unknowns are exactly the category a pre-committed filter cannot catch, and the resulting dataset is a non-random sample of your traffic, with all the problems described in our guide to [sampling bias](https://www.koji.so/docs/sampling-bias-research).\n\n### The two meters do not share a basis\n\nOne more detail worth checking against your own contract. Datadog's billing documentation says of replay that \"Datadog charges per one thousand (1,000) sessions ingested in the Datadog Real User Monitoring (RUM) service.\"\n\nRead that against Investigate, which the price list bases on \"filtered sessions\". As printed, the two meters are described on **different bases**: one on sessions ingested, the other on the filtered subset. If that holds for your account, then tightening retention filters reduces your Investigate line without reducing your Replay line by the same proportion, and the cost-control lever works on only one of the two expensive meters. This is worth confirming with Datadog directly before modelling savings, because the published wording and the intuitive reading come apart here.\n\n### A worked example\n\nTake one million sessions a month, on annual pricing.\n\n- **Measure only:** 1,000 units at 0.15 dollars = **150 dollars per month.** You get aggregate performance metrics on everything and can inspect nothing.\n- **Investigate 10 percent** (100,000 sessions): 100 units at 3.00 dollars = 300 dollars. Running total **450 dollars.**\n- **Add Session Replay on that same 10 percent:** 100 units at 2.50 dollars = 250 dollars. Running total **700 dollars per month.**\n- **Investigate 100 percent instead:** 1,000 units at 3.00 dollars = 3,000 dollars, for **3,150 dollars per month** before replay.\n\nThe jump from 150 to 3,150 dollars is the difference between counting your users and being able to look at them. On on-demand rates the same full-visibility stack runs about 1.47x higher again.\n\nFor context on where this sits in the market, Grafana's Observability Survey 2026, based on 1,363 responses collected between 1 October 2025 and 10 January 2026, found \"65% Cite cost as an important tool selection criteria-the most common response\", \"50.3% Plan to spend more on observability next year\", and \"38% Say complexity/overhead is their biggest concern-the most cited response\". Cost and complexity are the two things buyers say they care about most, and a two-meter model with a 20x spread is squarely both.\n\n## The 8 best Datadog RUM alternatives in 2026\n\n**1. Koji, best for the question RUM cannot answer.** Every tool below this line records behaviour. Koji records reasoning. It runs AI-moderated voice interviews with your real users, transcribes them, and applies automatic thematic analysis to surface the themes across every conversation, with a one-click report at the end. Where a retention filter forces you to guess in advance which sessions matter, an interview lets a customer volunteer the thing you never thought to filter for. Koji supports six structured question types (open_ended, scale, single_choice, multiple_choice, ranking, yes_no), so one study can produce both a quantified rating and the narrative explaining it. Pricing is EUR 29 per month for Insights or EUR 79 per month for Interviews, with EUR 1 per extra credit, and only conversations scoring 3 or above consume credits at all.\n\n**2. Microsoft Clarity, the free baseline.** Free heatmaps and session recordings with no session cap is an unusual offer in this category, and for many teams it removes the entire Measure-versus-Investigate dilemma. Fewer enterprise controls, but hard to argue with the price.\n\n**3. FullStory, for retroactive search.** Its model is closer to indexing everything so you can query it later, which directly addresses the pre-committed-filter problem. Note that FullStory publishes almost no pricing, as we documented in our [FullStory alternatives guide](https://www.koji.so/blog/fullstory-alternatives-2026).\n\n**4. Contentsquare, for enterprise digital experience analytics.** Deep zone-based and journey analysis aimed at large ecommerce and media teams. Its commercial model has its own edge, covered in our [Contentsquare alternatives guide](https://www.koji.so/blog/contentsquare-alternatives-2026).\n\n**5. LogRocket, for frontend debugging with session context.** Combines replay with console and network state, which suits engineering-led teams. Be careful with third-party pricing summaries here; the tier names in circulation are frequently wrong, as our [LogRocket alternatives guide](https://www.koji.so/blog/logrocket-alternatives-2026) sets out.\n\n**6. Sentry, if the real need is errors rather than sessions.** Many teams buy RUM when what they actually want is to know what threw. Sentry is cheaper and more focused for that, though its own ladder has a surprise, covered in our [Sentry alternatives guide](https://www.koji.so/blog/sentry-alternatives-2026).\n\n**7. Grafana Faro with OpenTelemetry, the open-source route.** Open standards remain a priority for the field: the same Grafana survey found \"77% Say open source/open standards are important to their observability strategy\". Self-hosting converts a per-session bill into infrastructure and engineering time.\n\n**8. Datadog Product Analytics as a partial substitute.** Listed at 0.80 dollars per 1,000 sessions annual, it sits between Measure and Investigate and may cover funnel and retention questions without paying the Investigate rate. Worth modelling before assuming you need the full stack.\n\n## How to choose a Datadog RUM alternative\n\n**Forecast on sessions, not users.** With a 15-minute inactivity timeout and a four-hour cap, session counts run well ahead of user counts. Get a real number from a month of traffic before signing anything.\n\n**Decide what fraction of traffic you genuinely need to inspect.** This single percentage drives the entire bill because of the 20x. Most teams need Investigate on far less traffic than they initially assume, and the Measure tier on everything else is cheap.\n\n**Commit annually if your volume is stable.** A consistent 1.44x to 1.50x on-demand premium is a large penalty for flexibility you may not use.\n\n**Check the consent position before you turn replay on.** France's CNIL ran a public consultation on session replay tools from 25 February 2026 to 22 April 2026. Its draft guidance treats replay as requiring prior consent, presented as a distinct purpose rather than folded into general analytics, with withdrawal as easy as giving it, and states it does not qualify for the exemptions that cover strictly necessary cookies. This is not only a compliance question. Visitors who decline consent are not a random subset of your users, so consented replay data self-selects in a way that interview research does not. Our guide to [survey response bias](https://www.koji.so/docs/survey-response-bias) covers the broader problem of who ends up in your dataset.\n\n**And be honest about which question you are funding.** RUM tells you what happened, at what speed, on which browser. It is very good at that. It does not tell you what the user believed, expected, or feared. If your roadmap debates keep stalling on motive rather than mechanics, the missing instrument is conversation, and our guides to [jobs-to-be-done interviews](https://www.koji.so/docs/jobs-to-be-done-interviews) and [thematic analysis](https://www.koji.so/docs/thematic-analysis-guide) are the place to start.\n\n## Frequently asked questions\n\n### How much does Datadog RUM actually cost per 1,000 sessions?\n\nIt depends entirely on what you want to do with the session. Datadog lists RUM Measure at \"$0.15\" per 1,000 sessions per month on full traffic and RUM Investigate at \"$3\" per 1,000 on filtered sessions, so measuring and investigating together comes to 3.15 dollars per 1,000 on annual pricing. Adding the Session Replay add-on at \"$2.50\" per 1,000 brings a fully inspectable session to about 5.65 dollars per 1,000. On-demand rates run roughly 1.44x to 1.50x higher across the board.\n\n### What is the difference between RUM Measure and RUM Investigate?\n\nMeasure applies to all ingested traffic and gives you aggregate performance metrics; Datadog states it provides \"performance metrics for all ingested sessions\" regardless of retention decisions. Investigate applies only to the subset you promote through retention filters, and it is what lets you open an individual session and examine it. The price difference is exactly 20x, which makes the percentage of traffic you retain the dominant variable in your bill.\n\n### How long is a Datadog RUM session?\n\nDatadog's billing documentation states that \"A session expires after 15 minutes of inactivity, and its duration is limited to 4 hours.\" A session is therefore a billing unit rather than a person. One user across a long working day can generate several billable sessions, so forecasts built on monthly active users will understate the invoice.\n\n### Is Datadog Session Replay billed on retained or ingested sessions?\n\nAs printed, the two meters describe different bases. Datadog's billing documentation says it \"charges per one thousand (1,000) sessions ingested in the Datadog Real User Monitoring (RUM) service\", while the price list bases Investigate on filtered sessions. If that distinction applies to your account, tightening retention filters would reduce the Investigate line without proportionally reducing Replay. Confirm the exact basis with Datadog before modelling savings from retention filters alone.\n\n### Do I need consent to run Datadog Session Replay in the EU?\n\nFrance's CNIL consulted publicly on session replay tools between 25 February 2026 and 22 April 2026, and its draft guidance treats replay as requiring prior consent presented as a distinct purpose, not bundled into general analytics, with withdrawal as easy as consent and no exemption of the kind that covers strictly necessary cookies. Beyond compliance, this has a research consequence: people who decline are not a random sample, so your replay data is systematically skewed toward those who consent.\n\n### What is the best Datadog RUM alternative for understanding user intent?\n\nNone of the behavioural tools solve it, because intent is not observable in a session recording. You can watch someone hesitate; you cannot see why. A platform like Koji that runs AI-moderated interviews and applies automatic thematic analysis answers the motive question directly, and it pairs well with RUM rather than replacing it: RUM finds the flow where users struggle, interviews explain what they were trying to do.\n\n## See the reason, not just the replay\n\nDatadog RUM will tell you precisely how many users hit your checkout, how fast it rendered, and which of them saw an error. Twenty times over, if you want to watch them.\n\nWhat it will never show you is the user who completed the flow perfectly, felt uneasy about it, and never came back.\n\nKoji runs AI-moderated voice interviews with real customers, applies automatic thematic analysis across every transcript, and delivers a one-click report with themes, quotes and frequencies. No moderator bias, no scheduling, no research background required. From question to insight in hours, not weeks.\n\nSign up and get 10 free credits, with only conversations scoring 3 or above consuming them. [Run your first Koji study](https://www.koji.so) and stop guessing at motive.","category":"Customer Research","lastModified":"2026-09-26T03:29:39.840378+00:00","metaTitle":"8 Best Datadog RUM Alternatives in 2026 (Pricing)","metaDescription":"Datadog RUM bills twice: 0.15 per 1,000 sessions to measure, 3.00 to investigate. The 20x explained, plus 8 alternatives.","keywords":["datadog rum alternatives","datadog rum pricing 2026","datadog session replay cost","real user monitoring tools","datadog rum vs fullstory","digital experience analytics"],"aiSummary":"Datadog Real User Monitoring uses a two-part meter: RUM Measure at 0.15 dollars per 1,000 sessions on full traffic and RUM Investigate at 3.00 dollars per 1,000 on filtered sessions, exactly 20x. Session Replay adds 2.50 dollars per 1,000. On-demand runs 1.44x to 1.50x above annual. Sessions expire after 15 minutes idle and cap at 4 hours. Retention filters require committing to a hypothesis before seeing the data. Eight alternatives compared.","aiKeywords":["datadog rum","real user monitoring","session replay pricing","observability costs","customer research"],"aiContentType":"comparison","faqItems":[{"answer":"It depends entirely on what you want to do with the session. Datadog lists RUM Measure at \"$0.15\" per 1,000 sessions per month on full traffic and RUM Investigate at \"$3\" per 1,000 on filtered sessions, so measuring and investigating together comes to 3.15 dollars per 1,000 on annual pricing. Adding the Session Replay add-on at \"$2.50\" per 1,000 brings a fully inspectable session to about 5.65 dollars per 1,000. On-demand rates run roughly 1.44x to 1.50x higher across the board.","question":"How much does Datadog RUM actually cost per 1,000 sessions?"},{"answer":"Measure applies to all ingested traffic and gives you aggregate performance metrics; Datadog states it provides \"performance metrics for all ingested sessions\" regardless of retention decisions. Investigate applies only to the subset you promote through retention filters, and it is what lets you open an individual session and examine it. The price difference is exactly 20x, which makes the percentage of traffic you retain the dominant variable in your bill.","question":"What is the difference between RUM Measure and RUM Investigate?"},{"answer":"Datadog's billing documentation states that \"A session expires after 15 minutes of inactivity, and its duration is limited to 4 hours.\" A session is therefore a billing unit rather than a person. One user across a long working day can generate several billable sessions, so forecasts built on monthly active users will understate the invoice.","question":"How long is a Datadog RUM session?"},{"answer":"As printed, the two meters describe different bases. Datadog's billing documentation says it \"charges per one thousand (1,000) sessions ingested in the Datadog Real User Monitoring (RUM) service\", while the price list bases Investigate on filtered sessions. If that distinction applies to your account, tightening retention filters would reduce the Investigate line without proportionally reducing Replay. Confirm the exact basis with Datadog before modelling savings from retention filters alone.","question":"Is Datadog Session Replay billed on retained or ingested sessions?"},{"answer":"France's CNIL consulted publicly on session replay tools between 25 February 2026 and 22 April 2026, and its draft guidance treats replay as requiring prior consent presented as a distinct purpose, not bundled into general analytics, with withdrawal as easy as consent and no exemption of the kind that covers strictly necessary cookies. Beyond compliance, this has a research consequence: people who decline are not a random sample, so your replay data is systematically skewed toward those who consent.","question":"Do I need consent to run Datadog Session Replay in the EU?"},{"answer":"None of the behavioural tools solve it, because intent is not observable in a session recording. You can watch someone hesitate; you cannot see why. A platform like Koji that runs AI-moderated interviews and applies automatic thematic analysis answers the motive question directly, and it pairs well with RUM rather than replacing it: RUM finds the flow where users struggle, interviews explain what they were trying to do.","question":"What is the best Datadog RUM alternative for understanding user intent?"}],"relatedTopics":["real user monitoring","session replay","observability pricing","customer research"]}],"pagination":{"total":1,"returned":1,"offset":0}}