Cisco ThousandEyes & AppDynamics Licensing

ThousandEyes Licensing Model

Cisco acquired ThousandEyes in 2020 for $1B. The platform provides internet and cloud network intelligence — monitoring application delivery paths across the public internet, ISP networks, cloud providers, and enterprise WAN. ThousandEyes is licensed using Units of Intelligent Systems (UIS), a consumption-based unit that bundles monitoring capacity across different test types.

What is a Unit of Intelligent System (UIS)?

A UIS represents a standardised block of monitoring capacity. Different types of ThousandEyes tests consume UIS at different rates depending on their complexity and data richness. Basic HTTP endpoint tests consume fewer UIS than BGP route monitoring or full path trace tests with browser sessions. Cisco publishes a UIS consumption table, but the mapping is not always intuitive and customers regularly underestimate consumption when scoping initial deployments.

ThousandEyes licences are sold in UIS bundles, typically in blocks of 100, 500, or 1,000 UIS. Enterprise agreements typically include a committed UIS quantity with True Forward mechanics applying if consumption exceeds the committed level at any point during the year.

ThousandEyes Agent Types

ThousandEyes operates through agents placed at strategic points across your network and cloud infrastructure. Understanding agent types is essential for scoping your UIS requirements accurately:

• Cloud Agents: Managed by ThousandEyes, deployed in over 200 locations globally. You do not manage these; you consume their capacity. Each test run against a cloud agent consumes UIS from your pool.
• Enterprise Agents: Software agents you deploy within your network — on physical appliances, virtual machines, or containers. Enterprise agents require deployment and management effort but consume UIS at lower rates for most test types.
• Endpoint Agents: Installed on end-user devices (laptops, desktops). Endpoint agents measure the end-user experience from the actual device perspective — critical for remote work visibility. Endpoint agent UIS consumption is typically lower than cloud or enterprise agents.

ThousandEyes Cost Drivers

ThousandEyes costs are driven by the number and type of tests you run, the agent types used, and the alert/dashboard consumption. Common scenarios that drive unexpected cost increases include:

• SaaS application monitoring expansion: IT teams frequently expand ThousandEyes monitoring to cover additional SaaS applications (Salesforce, ServiceNow, Workday) and cloud services. Each new application monitored from multiple vantage points multiplies UIS consumption rapidly.
• BGP monitoring deployment: BGP route monitoring is a UIS-intensive test type. Organisations deploying ThousandEyes for internet route visibility across multiple ISPs can consume significant UIS on BGP tests alone.
• Endpoint agent scaling: Deploying endpoint agents across all remote workers is a high-value use case for hybrid work visibility — but endpoint UIS consumption across 10,000 users monitored hourly adds up quickly.
• Alert proliferation: Every alert configured against a test consumes additional UIS for the alerting data path. Teams with aggressive alerting configurations can consume 15–25% more UIS than teams with equivalent test coverage but fewer alert rules.

Test Type	UIS Consumption (relative)	Common Use Case	Cost Control Lever
HTTP Server (Cloud Agent)	HIGH	SaaS availability monitoring	Shift to Enterprise Agents where possible
HTTP Server (Enterprise Agent)	LOW	Internal app monitoring	Default where network access permits
BGP Route Monitoring	HIGH	Internet route visibility	Limit to critical routes; reduce monitor frequency
Endpoint Agent Tests	MEDIUM	Remote worker experience	Define minimum agent population; rotate coverage
Path Visualisation	MEDIUM	WAN path troubleshooting	On-demand rather than continuous
Voice (RTP) Testing	MEDIUM	VoIP/Webex quality monitoring	Schedule off-peak; reduce frequency

AppDynamics APU Model

Cisco acquired AppDynamics in 2017 for $3.7B. AppDynamics is an Application Performance Management (APM) platform providing deep visibility into application behaviour, transaction tracing, infrastructure monitoring, and business performance analytics. In 2021, Cisco transitioned AppDynamics to the Agent Processing Unit (APU) model — a significant change from the previous per-agent named licence model.

What is an Agent Processing Unit (APU)?

An APU is a consumption unit that normalises the cost of monitoring across different AppDynamics agent types and infrastructure sizes. Rather than paying a flat fee per monitored application or per installed agent, APU consumption varies based on the agent type, the data volume it processes, and the infrastructure it monitors.

APUs are consumed in real time as your AppDynamics deployment processes monitoring data. The more agents you run, the more complex the applications they monitor, and the higher the transaction throughput — the faster your APU pool depletes. Cisco designed the APU model to capture more value from large-scale, high-transaction APM deployments.

APU Consumption by Agent Type

Agent Type	APU Consumption	Monitoring Scope	Overage Risk
Application (Java, .NET, Node.js)	Varies by transaction volume	Full APM + business transaction tracing	HIGH
Infrastructure (Machine Agent)	Moderate; fixed rate	Server health, JVM, OS metrics	MEDIUM
Database Agent	Lower; per DB instance	Query performance, wait states	MEDIUM
Browser (End User Monitoring)	Per page load / session	Real user experience	HIGH (traffic dependent)
Synthetic (Browser Robots)	Fixed per synthetic job	Synthetic UX monitoring	LOW
Analytics Agent	Varies by data ingest volume	Log analytics, BSOL dashboards	HIGH

AppDynamics Cost Drivers

The APU model creates cost dynamics that differ fundamentally from AppDynamics' previous per-agent pricing. The key shift is that your costs are now partially determined by your application workload — not just by how many agents you deploy.

High-Transaction Application Monitoring

Applications with high transaction throughput — e-commerce checkouts, high-volume API gateways, payment processing systems — consume APUs at significantly higher rates than lower-volume applications. If you scale a monitored application from 1M to 10M daily transactions (e.g., during a product launch or seasonal peak), your APU consumption can increase by 5–8x even with no additional agents deployed.

This creates a risk unique to APU licensing: your monitoring costs are partially correlated with your business success. Organisations that experience rapid growth or seasonal demand spikes face AppDynamics True Forward events driven not by IT decisions but by commercial performance.

Browser and End User Monitoring Scaling

End User Monitoring (EUM) through AppDynamics' Browser Agent or Mobile Agent generates APU consumption proportional to page loads and mobile sessions. For consumer-facing applications with high traffic volumes, EUM can become the largest single APU consumer in your deployment — sometimes exceeding the combined consumption of all application agents.

Cisco Full-Stack Observability (FSO) Platform Bundling

Cisco's strategic direction is to consolidate ThousandEyes, AppDynamics, and Cisco's broader observability portfolio (including network monitoring from DNA and Meraki) into the Cisco Full-Stack Observability (FSO) platform. FSO represents Cisco's answer to integrated observability spanning network, application, and end-user experience domains.

From a commercial perspective, FSO bundling creates both opportunities and risks. The opportunity: Cisco will typically offer FSO bundle discounts that reduce per-product pricing when ThousandEyes and AppDynamics are purchased together with networking licences. The risk: FSO packaging may lock you into consuming both products even if your current requirement is limited to one, and the bundled pricing can obscure the true cost of individual components.

FSO Bundle Considerations

• Evaluate whether you actually need both products: ThousandEyes and AppDynamics serve overlapping but distinct purposes. ThousandEyes excels at network-layer internet visibility; AppDynamics at application-layer transaction monitoring. If your primary need is one domain, the FSO bundle may include licences you will not use — and those unused licences do not reduce your True Forward baseline.
• Unbundle pricing for negotiation purposes: Even if you ultimately purchase an FSO bundle, negotiate the component pricing separately. This gives you visibility into the cost allocation between ThousandEyes and AppDynamics and allows you to benchmark against competitive alternatives for each product independently.
• Understand the FSO credit model: Some FSO packaging uses a shared credit pool that can be allocated between ThousandEyes UIS and AppDynamics APUs. If you have a credit pool, understand the exchange rates and whether unused credits from one product can offset overages in the other.

8 Negotiation Tactics for ThousandEyes and AppDynamics

Tactic 1: Model Consumption Before Committing

The most important preparation step is building a detailed consumption model before signing any commercial agreement. For ThousandEyes, map your planned tests, agent types, and monitoring frequency to expected UIS consumption. For AppDynamics, model APU consumption against current transaction volumes with growth assumptions. Request that Cisco validate your model before contract signing — if Cisco is unwilling to validate the consumption estimate, that itself is a red flag about the accuracy of their packaging.

Tactic 2: Negotiate APU True Forward Protections Explicitly

Given that APU consumption is partially driven by transaction volume (which may scale with business performance), negotiate specific True Forward protections for traffic-driven consumption increases. A reasonable position is that APU overages attributable to business traffic growth (rather than new agent deployments) should be treated separately from overages driven by new agent expansion — with a proportionality test applied before True Forward charges are assessed.

Tactic 3: Leverage Datadog and Dynatrace Alternatives

AppDynamics faces significant competitive pressure from Datadog, Dynatrace, New Relic, and Elastic APM. All of these alternatives have enterprise customer bases comparable to AppDynamics and offer feature parity or superiority in specific domains. Cisco knows that AppDynamics-to-Datadog migration is a credible competitive threat. Use this explicitly in renewal negotiations — a POC with Datadog is a legitimate negotiation lever, not just a bluff. For ThousandEyes, Catchpoint and Broadcom DX NetOps provide competitive alternatives with different pricing models.

Tactic 4: Demand Usage Analytics Access

Both ThousandEyes and AppDynamics provide usage analytics — data showing your actual consumption of UIS and APUs over time. Negotiate for contractual access to granular usage data (not just the summarised view in the Cisco portal) so you can run your own consumption models and identify optimisation opportunities. Cisco's default reporting may not surface the specific test-level or agent-level breakdown you need to manage costs effectively.

Tactic 5: Right-Size Your Agent Population Before True Forward Events

Conduct a quarterly AppDynamics agent audit to identify agents monitoring retired applications, decommissioned servers, or staging environments that do not require production-level monitoring. Unused or low-value agents that have been deployed and forgotten are a common source of APU over-consumption. Right-sizing your active agent population before the True Forward measurement date reduces both your current APU burn rate and your True Forward exposure.

Tactic 6: Negotiate EUM Separately from APM

AppDynamics End User Monitoring (browser and mobile) is a distinct use case from server-side APM. If your organisation has high public web traffic, negotiate EUM pricing with a traffic-based cap rather than an open-ended APU consumption model. Cisco will typically price EUM within the broader APU framework, but separating it allows you to benchmark EUM pricing against web analytics alternatives (Datadog RUM, Dynatrace, FullStory) that may provide equivalent value at lower cost.

Tactic 7: Use the Cisco EA as Bundling Leverage

If you already have a Cisco Enterprise Agreement for networking or security, use that relationship to negotiate preferential pricing on ThousandEyes and AppDynamics. Cisco places high value on expanding EA suite coverage — the sales team has incentives to add new suites to existing EAs. This creates a negotiating dynamic where your existing EA spend gives you leverage to request FSO products at preferential rates as part of a broader EA amendment, rather than as a standalone purchase.

Tactic 8: Push for Flex-Pool Structures

Where possible, negotiate for a shared UIS/APU flex pool rather than separate committed quantities for ThousandEyes and AppDynamics. A flex pool allows you to allocate capacity to whichever product is experiencing higher demand at any given time — reducing the risk of True Forward on one product while you have surplus capacity on the other. Cisco has offered flex-pool constructs in enterprise agreements where the customer is purchasing both products at meaningful scale.

Competitive Alternatives to Cisco Observability

Understanding the competitive landscape is essential for maintaining negotiating leverage with Cisco. The observability market is highly competitive, and Cisco is acutely aware that both ThousandEyes and AppDynamics face credible alternatives.

Cisco Product	Key Alternatives	Competitive Strength	Negotiation Leverage
ThousandEyes	Catchpoint, Broadcom DX NetOps, NETSCOUT	Catchpoint has stronger SaaS monitoring depth	MODERATE
AppDynamics APM	Datadog, Dynatrace, New Relic, Elastic APM	Datadog/Dynatrace highly credible; very active market	HIGH
AppDynamics EUM	Datadog RUM, Dynatrace, FullStory	Multiple strong alternatives at lower price points	HIGH
FSO Platform	Datadog full-stack, Dynatrace Platform, Grafana	Cisco FSO integration story is still maturing	HIGH

Frequently Asked Questions