Quick Answer
DX is an engineering intelligence platform built by researchers. It joins SDLC data from tools like GitHub and Jira with developer surveys, then rolls the result into the Developer Experience Index and the DX Core 4 framework for leaders running org-wide measurement programmes. DevClocked measures something narrower and closer to the keyboard: what each working session produced, which agent produced it, and what the tokens cost. If you need a survey programme and cross-company benchmarks, buy DX. If you need per-session, per-agent measurement of agent-assisted work, DevClocked.
At a glance
| DevClocked | DX | |
|---|---|---|
| Buyer | Developers and lean teams | Engineering executives, platform and DevEx teams |
| Core metric | Leverage Score (output per unit of effort) | DXI and DX Core 4 (speed, effectiveness, quality, impact) |
| Data source | Session telemetry from IDE, terminal and agents, plus a git baseline | Git and ticket metadata, plus developer surveys |
| Cadence | Continuous | System metrics plus survey cycles |
| Developer sentiment | Not measured | Yes, its signature strength |
| AI agent tracking | Per run, first-class | Adoption and impact at programme level |
| Token and cost tracking | Per session and per model | AI spend as a programme cost line |
| Benchmarking | Leverage benchmarking inside your workspace | Cross-company industry benchmarks |
| Rollout | Developer installs it today | Org programme with survey support |
| Pricing | Published on the site | Quote-based enterprise, verify current pricing |
What DX does well
DX has the strongest research pedigree in this category and it is not close. Abi Noda founded it after leading engineering velocity work at GitHub. Margaret-Anne Storey, co-author of the SPACE framework, is Chief Scientist, and Nicole Forsgren, lead author of Accelerate and the name most associated with DORA metrics, joined as a strategic advisor in 2022. The DX Core 4 folds DORA, SPACE and DevEx into one framework across four dimensions: speed, effectiveness, quality and impact.
The DXI is the part telemetry cannot copy. It is a survey instrument covering 14 dimensions of engineering effectiveness, including deep work, iteration speed and release processes. DX reports that each one-point gain maps to about 13 minutes saved per developer per week, and that it built the index on data from more than 40,000 developers across 800 organisations, with over four million benchmark samples for industry comparison. Asking developers what is slowing them down, then comparing the answer against hundreds of other companies, is a real capability. No git telemetry substitutes for it, and DevClocked does not try.
DX also handles what an enterprise rollout needs and a small tool ignores: survey deployment, executive and board-level reporting, single-tenant hosting with data residency, Jira connectors, and DX Data Cloud for custom reporting. Its AI work is credible too. DX measures utilisation, impact and cost, publishes a quarterly State of AI Impact in Engineering report, and argues that every speed or spend metric needs a quality counterweight before anyone calls an AI investment a success.
Where DevClocked is different
The gap is the measurement layer, not the ambition.
Sessions, not only merged artifacts. DX reads git and ticket metadata, so it sees the diff that landed. DevClocked records the session that produced it across IDE and terminal, including agent runs, exploration and work that never became a commit.
Agents as first-class subjects. DevClocked tracks Claude Code, Codex CLI and Cursor per run and attributes work between AI and human. DX measures AI adoption and self-reported time saved at programme level, a different resolution entirely.
Token cost next to shipped output. DevClocked records tokens and cost per session and per model, then sits that beside what shipped. DX treats AI spend as a line in a business case.
Continuous, not cycle-based. The DXI is a survey. It is accurate about how engineers feel and it moves at the speed of survey cycles. Session telemetry updates while you work.
Honest about hours. DevClocked uses git as an approximate baseline, never a clock. A lightweight editor extension and an editor-agnostic CLI tracker supply the real timing, and an algorithm calibrates the two over time. More in engineering team time tracking.
Feature by feature
| Feature | DevClocked | DX |
|---|---|---|
| Developer surveys and sentiment | No | Yes (signature strength) |
| Developer Experience Index | No | Yes (signature strength) |
| Cross-company industry benchmarks | No | Yes (strength) |
| Executive and board reporting | Team dashboards | Yes (strength) |
| DORA, SPACE, DX Core 4 frameworks | No | Yes (strength) |
| Broad SDLC integrations | Focused | Very broad (strength) |
| Single-tenant hosting, data residency | No | Yes (strength) |
| Per-session activity tracking | Yes | No |
| AI agent tracking per run | Yes | Programme-level adoption |
| Token and cost attribution | Yes | Programme-level spend |
| AI versus human attribution | Yes | Estimated and self-reported |
| Leverage Score | Yes | Core 4 dimensions instead |
| Published pricing | Yes | Quote-based |
Pricing
DX does not publish prices. It sells annual enterprise contracts priced per developer, and modules such as AI Code Insights are quoted separately from the base platform. Third-party contract-data sites report typical annual values in the tens of thousands of dollars. Treat those as directional and get a real quote.
DevClocked publishes its pricing, including a per-seat Business tier covering shared workspaces, org dashboards and leverage benchmarking, with a full-access trial before you pay. For a team of ten the gap between the two is not a rounding error. Verify current pricing on both sites.
Who should pick which
Pick DX if you run an org large enough to need a survey programme, you want to benchmark against other companies, and you need reporting that survives a board meeting. Nothing in DevClocked replaces that.
Pick DevClocked if you want to know what your agent-assisted sessions actually produced, at what token cost, without an annual contract or a rollout project. See teams for the shared-workspace side.
Run both if you already use DX for developer experience and need agent-level detail underneath it. They answer different questions.
Verdict
DX is the most research-serious product in this space. If you need to measure developer experience across a large org, benchmark it and present it to a board, buy it. DevClocked is not competing for that budget. What DevClocked does that DX does not is watch the session itself: which agent ran, how long the work took across IDE and terminal, what the tokens cost, and how that compares to what shipped. Surveys tell you how the work felt. Session data tells you what the work produced. Teams running heavy agent workflows need the second number, and far fewer of them need an enterprise contract to get it.
FAQ
Only for part of the job. DevClocked replaces DX for session-level and agent-level measurement of what got built. It does not replace the DXI, the survey programme, or the cross-company benchmarking. If those are why you bought DX, keep DX.
DX measures AI across utilisation, impact and cost, and publishes quarterly research on AI's effect on engineering. That runs on tool usage data, git metadata and developer self-reports rather than per-session agent telemetry, so it will not give you a token cost per session. DevClocked will. See how to measure AI adoption across an engineering team.
DX tells you whether developers report saving time and whether delivery and quality moved, benchmarked against other companies. DevClocked tells you which agent runs produced which shipped work and what those tokens cost. The second is narrower and harder to argue with. Related: benchmark AI leverage across a team.
Quote-based, sold as an annual per-developer enterprise contract, with some AI modules priced separately. There is no public price list, so ask their sales team rather than trusting a number you read in a comparison post, including this one.
DevClocked vs LinearB and DevClocked vs Swarmia cover the delivery-workflow side. For the wider category, see software engineering intelligence platforms and engineering productivity metrics.