Cisco AppDynamics Licensing: Observability Costs

Cisco AppDynamics prices full-stack observability by the CPU core, not the user — which means your bill is driven by how many cores your hosts expose, and over-provisioned infrastructure inflates it silently. This guide explains the per-core model, the three editions, the token and container add-ons, where consumption quietly runs away, and how to size and negotiate the commitment.

By Morten Andersen

The Per-CPU-Core Licensing Model

Cisco AppDynamics licensing is built on a single metric: the CPU core. A core is defined as the logical thread core or processor that the operating system reports on each unique physical or virtual host, and you pay an annual subscription per core for every host you instrument. The headline consequence is that AppDynamics cost is a function of your infrastructure footprint, not your headcount — a 200-developer team monitoring a lean estate can pay less than a 20-person team monitoring a sprawl of over-provisioned virtual machines.

This puts AppDynamics in a different category from the per-device and per-user shapes elsewhere in the Cisco portfolio. The discipline it demands is closer to cloud cost management than to seat counting: every core the agent sees is a core you pay for, so right-sizing the hosts is a licensing lever, not just an infrastructure one. That dynamic separates a controlled observability line from a runaway one, and it sits alongside the consumption discipline of the Cisco ThousandEyes licensing guide and the per-device logic of the Cisco Meraki licensing guide.

Editions: Infrastructure, Premium, Enterprise

AppDynamics is packaged into three editions, each licensed against the same per-core metric but bundling progressively more capability. Infrastructure Monitoring covers server and infrastructure visibility; Premium adds full application performance monitoring (APM) and business transaction tracing; Enterprise adds the analytics tier, including business performance monitoring and the deeper data services.

EditionWhat It CoversIndicative List (per vCPU/month, billed annually)
Infrastructure MonitoringServer and infrastructure metrics~$6
PremiumAPM, business transactions, code-level diagnostics~$33
EnterpriseAPM plus analytics and business performance monitoring~$50

List figures vary by source and configuration, but the spread is the point: Enterprise lists at roughly eight times Infrastructure Monitoring per core, so paying for the top edition across every host — including infrastructure that only needs server metrics — is one of the most common ways AppDynamics spend balloons. Tiering editions to what each tier of the estate actually needs, rather than buying Enterprise wall-to-wall, is the same right-sizing logic the Cisco DNA licensing tiers guide applies to network entitlement and the Cisco Smart Licensing compliance guide applies to entitlement tracking generally.

Add-Ons: RUM Tokens and Secure Application

Beyond the core editions, AppDynamics sells capability as separately metered add-ons — and these carry their own consumption traps. Real User Monitoring (RUM), covering browser and mobile end-user experience, is priced in tokens that represent traffic volume such as page views and sessions: a Peak RUM licence of 10 million tokens lists near $2,400, and high-traffic applications burn tokens far faster than buyers anticipate. Cisco Secure Application, which adds runtime application security, is metered in container units — one unit tied to APM Any Language Microservices monitors five containers — and additional event-data retention at 30, 60 or 90 days is a further chargeable add-on.

The pattern is consistent with the rest of the Cisco estate: the base licence is only part of the bill, and the add-ons are where forecasts drift. RUM tokens in particular behave like a consumption meter rather than a fixed seat, so they need to be sized against real traffic and watched through the term, the same way the security portfolio is managed in the Cisco security licensing guide.

The Core-Count Inflation Trap

Because the metric is the core, the single largest source of unexpected AppDynamics cost is core-count inflation — and it has two mechanisms. The first is over-provisioning: a virtual machine handed 16 vCPUs it does not need still consumes 16 licence units, so generous VM sizing translates directly into wasted spend. The second is more insidious — an agent fallback behaviour that catches even careful buyers.

If an AppDynamics agent receives no vCPU information after roughly 10 minutes, it falls back to a default core count — commonly 4 or 12 — even on a single-core host. A fleet of small hosts can therefore consume several times the licences their real cores justify, silently, until a true-up reconciles the count. Confirming agent reporting and right-sizing hosts before any reconciliation is the control.

The fix for both is the same: treat the licence position as a reflection of right-sized infrastructure, not of whatever the hypervisor happens to allocate. Reconciling reported cores against real, needed cores before a true-up surfaces both over-provisioned hosts and fallback over-consumption, and turns the count from a vendor-driven number into an evidence-driven one — the same evidence-led posture set out in the Cisco EA negotiation and pricing guide.

SaaS, Self-Hosted and the Splunk Transition

AppDynamics is available both as a Cisco-hosted SaaS service and as a self-hosted (on-premises) deployment, and both use the same per-core metric — the choice is about data residency and operational control, not a different licensing model. Following Cisco's acquisition of Splunk, the product has been folded into the Splunk Observability portfolio and rebranded Splunk AppDynamics. Cisco has been explicit that this is integration, not sunset: there is no forced migration, and existing AppDynamics entitlements continue.

For buyers, the transition matters less as a product risk than as a negotiation signal. As AppDynamics is integrated with the wider Splunk observability stack, the commercial conversation increasingly spans both — which is leverage if you are also buying Splunk, and a bundling risk if you are not watching the line items. Keep AppDynamics core counts and Splunk ingest priced and committed separately enough to see each, the way the broader move to subscription is handled in the Cisco subscription licensing transition guide.

AppDynamics Inside a Cisco EA

Where AppDynamics sits alongside other Cisco spend, the strongest commercial move is usually to buy it through the Enterprise Agreement rather than as a standalone subscription. Inside an EA, AppDynamics adds price predictability and flexibility, and the core commitment counts toward the committed value that drives the overall discount band — so a material observability spend can lift the discount on the rest of the estate rather than standing as an isolated, full-list line. It also aligns the renewal date with the wider agreement, removing a separate negotiation cycle.

The precondition is that the core count must be sized to real, right-sized infrastructure before it goes into the EA, or you simply commit your over-provisioning at scale. As with the collaboration and monitoring portfolios, the question is whether the AppDynamics commitment is large enough to move the EA band — and for enterprises running APM across a broad estate, it usually is. Structuring that inclusion is part of the broader strategy in the Cisco Enterprise Agreement and licensing guide, and it interacts with the network-side commitments covered in the Cisco SD-WAN licensing guide and the collaboration spend in the Cisco Webex licensing guide.

Sizing and Negotiating the Commitment

Optimising AppDynamics is a sizing exercise before it is a discount exercise. Right-size the hosts so the core count reflects real need, tier editions so infrastructure-only hosts are not paying Enterprise rates, size RUM tokens and Secure Application units against actual traffic and container counts, and confirm agent reporting so fallback defaults are not inflating the bill. Only then negotiate the price — with utilisation evidence in hand, volume discounts and custom pricing are available on large installations, and a buyer who shows up with a reconciled core position negotiates from fact rather than the vendor's growth assumptions.

Done this way, the per-core model becomes manageable rather than a blind spot: you pay for the observability your right-sized estate actually needs, folded into the wider Cisco relationship for the discount benefit. To size and negotiate your AppDynamics commitment before your next renewal, request a confidential briefing, or download our Cisco EA Playbook.

Common Questions

Cisco AppDynamics Licensing: FAQ

How is Cisco AppDynamics licensed?
AppDynamics is licensed per CPU core, billed annually. A CPU core is the logical thread core or processor the operating system reports on each unique physical or virtual host, so a host's core count — not its user count — drives cost. The product is packaged into three editions: Infrastructure Monitoring, Premium and Enterprise. Real User Monitoring and Cisco Secure Application are separate add-ons, RUM priced in tokens and Secure Application in container units. Both SaaS and self-hosted deployments use the same per-core metric.
What does Cisco AppDynamics cost per core?
List figures vary by source and edition, but indicative per-vCPU monthly rates billed annually are roughly $6 for Infrastructure Monitoring, around $33 for Premium and around $50 for Enterprise. Real User Monitoring is bought separately in tokens — for example a Peak RUM licence of 10 million tokens lists near $2,400. Because the metric is the core, the same application can cost very different amounts depending on how many vCPUs its hosts expose and whether they are right-sized.
Why does AppDynamics consume more licences than expected?
The licence count tracks the CPU cores the agent sees on each host, so over-provisioned virtual machines inflate consumption directly — a host given 16 vCPUs it does not need still draws 16 units. A second cause is the agent fallback: if an agent receives no vCPU information after about 10 minutes, it falls back to a default core count (commonly 4 or 12) even on a single-core host, silently over-consuming. Right-sizing hosts and confirming agent reporting before a true-up are the controls.
Should AppDynamics be bought inside a Cisco Enterprise Agreement?
For enterprises already running material Cisco spend, yes. Buying AppDynamics through a Cisco Enterprise Agreement adds price predictability and flexibility, and the commitment counts toward the committed value that drives the overall EA discount band rather than sitting as an isolated subscription. It also aligns the renewal date with the rest of the estate. The trade-off is that the core count must be sized to real, right-sized infrastructure first, so the EA commitment reflects need rather than over-provisioned hosts.

Size the Cores to Evidence, Not the Hypervisor

Our advisors right-size your AppDynamics core count, tier editions to need, control the RUM and Secure Application add-ons, and fold the commitment into the wider Cisco EA for the discount benefit.

Request a Confidential Briefing Explore the Cisco Hub

Cisco Licensing Intelligence

Monthly briefings on Cisco AppDynamics and observability licensing, per-core consumption economics, and EA negotiation tactics — from advisors who negotiate Cisco agreements on behalf of buyers.