Get my Datadog rebuild quote →
More Menu
Book a call →
UPSTREAM · DATADOG ALTERNATIVE
For engineering teams paying $3K–$30K+/mo on Datadog Pro / Enterprise

Datadog APM is Tempo.
Logs is Loki. Metrics is Prometheus.

Datadog charges per host + per million logs ingested + per synthetics test + per RUM session + per APM trace. The underlying capabilities — metrics, traces, logs, dashboards — are commodity primitives covered by the LGTM stack: Loki + Grafana + Tempo + Mimir, all open source, all maintained by Grafana Labs. We deploy + operate the LGTM stack on your AWS. Same dashboards, your data, no per-host pricing.

Get my rebuild quote · 24h turnaround →
Fixed price from $22,000·4–6 weeks to ship·
Typical SaaS · 200 hosts · 300GB logs/day

Datadog Pro · APM + Logs + Infra

200 hosts · 300GB logs/day · 50 synthetics · RUM
Datadog Y1 spend$0
Self-host LGTM (AWS · S3 + EC2)~$0
Allied BizTech deploy (one-time)$0
Managed ops (per year)$0
3-YEAR RECLAIM · INDICATIVE
$0
Build payback
~4 months
Y2+ saving
~$81K/yr
Stack
Yours

At 1,000 hosts + 5TB logs/day (typical 500-eng SaaS), Y1 reclaim crosses $300K.

The math · every Datadog footprint

Datadog has at least seven taxes.

Per-host APM + per-host Infrastructure + per-million logs + per-test synthetics + per-RUM session + per-CSM session + per-database query — each scales independently. Self-host scales with raw storage, which is dramatically cheaper.

Footprint
Datadog / yr
Self-host (AWS + Allied amortised)
Y1 reclaim after build
50 hosts · APM + Infra · 50GB logs/daysmall SaaS · ~50 engineers
~$24,000
~$26,000 Y1 / $4,000 Y2+
Marginal Y1 + $20K/yr after
200 hosts · APM + Logs + Infra · 300GB/dayscaling SaaS · sweet spot
~$96,000
~$31,000 Y1 / $15,000 Y2+
~$65,000 Y1
500 hosts · full suite · 1TB/daymid-stage SaaS
~$240,000
~$54,000 Y1 / $32,000 Y2+
~$186,000 Y1
1,000 hosts · 5TB logs/day · multi-regionlarge SaaS · enterprise tier
~$540,000+
~$110,000 Y1 / $88,000 Y2+
~$430,000 Y1
3,000+ hosts · global scalevery large SaaS · custom Datadog contract
~$1,800,000+
~$320,000 Y1 / $290,000 Y2+
~$1,480,000 Y1

Datadog list prices May 2026. Self-host numbers based on Grafana / Loki / Tempo / Mimir on AWS EC2 + S3 (long-term storage) + Allied $22K deploy amortised. Datadog is the most-overpaid SaaS in our experience: per-host pricing means the bill grows with your fleet, but the LGTM cost grows with bytes — much more gently.

What we deploy

Three layers. All open source.

The LGTM stack from Grafana Labs is the de facto open-source observability standard. Mimir for metrics (Prometheus-compatible). Loki for logs. Tempo for traces. Grafana for dashboards. All AGPL or Apache-licensed, all production-grade at scale, all on your AWS.

LAYER 1 · INGEST
OpenTelemetry collectors
  • OTel Collector daemonset (replaces dd-agent)
  • Sampling + filtering at ingest
  • Datadog-compatible API (drop-in dd-agent → otel migration)
  • Per-service routing to right backend
  • Auto-scaling on AWS ECS / EKS
→
LAYER 2 · STORAGE
LGTM backends
  • Mimir — Prometheus-compatible metrics, S3 long-term
  • Loki — log aggregation, S3 backed
  • Tempo — distributed traces, S3 backed
  • Pyroscope — continuous profiling
  • Configurable retention per tenant / project
→
LAYER 3 · UI + ALERTING
Grafana + Alertmanager
  • Grafana Enterprise UI (free for self-host)
  • Migrated dashboards from your Datadog config
  • Alertmanager → PagerDuty / OpsGenie / Slack
  • SSO via Okta / Azure AD / Google
  • SLO tracking with burn-rate alerts

All four LGTM components are open source — Loki and Tempo are AGPLv3, Mimir is AGPLv3, Grafana is AGPLv3. Self-host is fully unrestricted; you can fork the code if you want to.

What's preserved · what's different

Almost everything Datadog does, LGTM does too.

The LGTM stack is purpose-built as a Datadog/New Relic alternative. Coverage is comprehensive — metrics, logs, traces, dashboards, alerting, SLOs, profiling. The differences are operational (tenant config, retention, scale tuning) not feature-level.

CORE — AT PARITY WITH DATADOG
Application Performance Monitoring (APM)distributed traces, service maps, slow-query analysis
✓
Log management + searchstructured logs, full-text, log → trace linking
✓
Infrastructure monitoringCPU, memory, disk, network, container metrics
✓
Custom dashboardsdrag-drop, templating, variable substitution
✓
Alerting + on-call routingmulti-condition, threshold + anomaly, PagerDuty integration
✓
SLO tracking + burn-rate alertsnative in Grafana Enterprise
✓
Continuous profilingPyroscope · CPU + memory + goroutines
✓
Synthetics monitoringGrafana Synthetic Monitoring + k6 cloud
✓
PARTIAL — REBUILT BUT DIFFERENT
Real User Monitoring (RUM)Replaced with Grafana Faro (production-ready, less polished UI)
~
Cloud Security ManagementReplaced with Falco + Wazuh (different UX, comparable coverage)
~
Datadog NotebooksReplaced with Grafana Explore + saved queries
~
DROPPED — REPLACE PER-SERVICE
Datadog AI / Watchdog ML detectionReplaced per-service: Prometheus alert rules + custom anomaly detection if needed
✕
NEW — DATADOG DOESN'T DO
No per-host pricingAdd hosts free; storage scales with bytes, not host count
✓
Configurable retention per workloadHot logs 7 days, cold logs 365 days on S3 — your retention policy
✓
SQL access to all observability dataQuery Loki / Tempo / Mimir from any tool, no Datadog API rate limits
✓
Pricing

Fixed price. Sized by host + log volume.

One number, one ship date. Sized by host count, log volume, and required HA posture.

⭐ Datadog rebuild · up to 500 hosts · 1TB logs/day

From $22,000 — fixed

$22,000 · one-time
4–6 weeks · 1-week parallel run · 90-day warranty
  • LGTM stack deployment on your AWS (EKS / ECS)
  • OTel Collector with dd-agent compatibility for drop-in migration
  • Mimir + Loki + Tempo + Grafana sized for your volume
  • Pyroscope continuous profiling
  • SSO integration — Okta / Azure AD / Google Workspace
  • Migrated dashboards — converted from Datadog JSON exports
  • Alertmanager → PagerDuty / OpsGenie / Slack
  • Long-term storage on S3 with configurable retention
  • 1-week parallel run with Datadog before cutover
  • Runbook + handover doc
Get my Datadog rebuild quote →

24-hour quote turnaround · NDA on request

SCALE TIER · 1,000+ HOSTS
Multi-region clusters + advanced sharding: $32,000–$48,000.
+ RUM TIER
Add Grafana Faro for Real User Monitoring: +$8,000–$14,000.
+ SECURITY MONITORING TIER
Add Falco + Wazuh for cloud security observability: +$14,000–$22,000.
RUNNING IT FOR YOU
Allied BizTech-managed at $6,000/yr covers monitoring + monthly upgrades + SLA support.
The "Datadog alternatives" listicles miss the point

Why we don't recommend the obvious switches.

"Datadog alternative" results: New Relic, Dynatrace, Splunk, Honeycomb, Chronosphere. Same hosted-vendor model, different brand. Same per-host or per-event pricing.

PATH A · New Relic
"Per-user pricing instead."
New Relic's per-user pricing seems cleaner but ingestion overage caps the win. Same wrapper economics at the same scale; sometimes worse for log-heavy teams.
PATH B · Honeycomb
"Best traces UX."
Honeycomb's UX for traces is genuinely best-in-class. Per-event pricing scales aggressively; great for low-volume teams, expensive at SaaS scale. Still a hosted vendor.
PATH C · Chronosphere
"Observability for cloud-native."
Chronosphere is genuinely good and built on M3 (similar to Mimir). Hosted pricing is comparable to Datadog at scale. Self-hosted LGTM is the same architecture, your AWS.
PATH D · Grafana Cloud
"LGTM hosted by Grafana Labs."
Grafana Cloud is a great option for sub-200-host teams. Above 500 hosts the pricing approaches Datadog's; self-host is meaningfully cheaper at that point.
PATH E · Self-host LGTM yourself
"It's open source."
You can. Production-grade Mimir tuning + multi-tenant config + retention policy + on-call setup is 4–6 weeks of senior infra time. We do that work fixed-price.
PATH F · Negotiate Datadog down
"Just push back."
Datadog will discount Pro / Enterprise 15–25% on multi-year commits. Self-host margin is 75%+ at 200+ hosts. Negotiation captures less than 1/3 of self-host value.
Questions about a Datadog rebuild

"What about…"

Will my dd-agent integrations break?▾
No. The OTel Collector ships a Datadog receiver — it accepts traffic from existing dd-agents transparently. You don't need to re-instrument anything immediately. Migrate to native OTel SDKs gradually as you touch services.
How do we migrate dashboards?▾
Datadog dashboards export as JSON. We translate them to Grafana JSON via our migration tool, including templating + variables + alert rules. ~85% of dashboards convert automatically; the remaining 15% need manual re-creation (typically the most complex / least-used).
What about Datadog AI / Watchdog?▾
Watchdog's anomaly detection is useful but oversold. For 80% of cases, threshold + rate-of-change alerts in Grafana are sufficient. For the rest, we add custom anomaly detection (Prometheus + Python) — same outcomes, no per-detection fee.
How does log ingestion compare?▾
Loki ingests structured logs at line rate — your existing log shippers (Fluent Bit, Vector, Filebeat) work unchanged. Loki's index-free design is far cheaper at scale than Datadog's log management; trade-off is search is grep-style (label-based) rather than free-text-first. For most use cases this is fine; for forensics-heavy teams we add ClickHouse for free-text.
What about retention?▾
Configurable per workload. Typical setup: 7 days hot retention on EBS, 365 days cold retention on S3 (with automatic restore for query). Datadog's flat retention is much less flexible; this gives you per-team / per-compliance-requirement control.
Who owns everything?▾
You. Helm / Terraform config in your GitHub. LGTM runs in your AWS account. No Allied infrastructure in the loop. You can take over fully if you cancel managed-ops.
Other tools we rebuild on direct billing
Sentry alternative Mixpanel alternative Segment alternative Data ops stack combo B2B SaaS stack combo Mid-market full-stack → → Full Upstream catalog

Why this number works: AI-paired senior engineers raise the floor on quality + consistency · 250+ products shipped · direct vendor billing — you keep the wrapper margin. → The full method

UPSTREAM · DATADOG REBUILD

Stop the per-host pricing. Self-host LGTM.

From $22,000 fixed. 4–6 weeks to ship. Same dashboards. Your AWS. Your bytes, no per-host pricing. No deck-pitch — send your Datadog plan + host count + monthly log volume, we send a quote in 24 hours.

Get my Datadog rebuild quote → 30-min stack review · free

15+ years · ISO 9001:2015 · 250+ shipped · MSA · SOW · DPA ready · LGTM stack is open source.