Pricing data from 47 companies shows a consistent pattern where initial estimates miss by 3-12x. One VP of Engineering budgeted $12,000/month but received an actual invoice of $147,000. The multi-dimensional, usage-based model makes it nearly impossible to predict costs accurately. Teams regularly experience 'bill shock' when the invoice arrives.

The billing model is complex, and you have to be extremely careful with custom metrics and log indexing, or the cost will explode. Custom metrics are charged per metric per host, and enabling features like APM or custom instrumentation can quickly multiply your bill. Users report turning on 'all the things' and getting massive surprise bills at the end of the month.

The per-host pricing model ($15-$27/host for Infrastructure, $31-$40/host for APM) quickly adds up as your infrastructure scales. Adding containers, serverless functions, or microservices multiplies costs. Organizations that started small find Datadog becoming prohibitively expensive as they scale. Costs require constant attention and strategic planning to control.

Trustpilot reviews report dissatisfaction with Datadog's billing practices, accusing the company of employing questionable tactics such as hidden costs and not allowing users to remove their payment method or downgrade or end their subscription. Users advise being very careful about what is going to be charged and double-checking what is said even by their engineers.

Log Management has complex pricing: ingestion is one cost, indexing is $0.10/GB, and retention policies add additional charges. Users struggle to understand index logs vs non-index logs vs retention pricing. It's very flexible, which creates many confusing options when trying to control costs. Many teams accidentally index more logs than needed.

A significant issue with Datadog is vendor lock-in, driven by its proprietary agent. To use Datadog, you must embed their agent throughout your infrastructure. If you ever decide to switch platforms, you face a major migration project. The agent deployment and custom dashboards don't transfer to other platforms.

Trustpilot reviews mention the company harassing customers with unsolicited marketing emails and phone calls, with no intention to stop. Users who signed up for trials or evaluated the product report persistent sales outreach even after declining. This aggressive approach damages trust.

Datadog offers a wide range of features that can be overwhelming for new users. The UI is dense and configuration steps are overwhelming - especially when setting up multi-service tracing, infrastructure maps, or custom metrics dashboards. Creating custom dashboards and metrics isn't straightforward, with what should be simple configurations often turning into time-consuming projects.

Setting up dashboards, tuning alert rules, building monitors, navigating between products, and optimizing usage takes considerable time and expertise. Getting the full benefit of Datadog requires significant technical expertise. The setup process can be time-consuming and may require dedicated resources or experienced personnel.

Some areas are too vague, with no reply on issues like metric rate changes by time window, custom metrics insights, APM changes, and pricing structure. Users report that pricing details can be intricate and often require careful monitoring to avoid unexpected charges. The lack of clear documentation makes cost prediction nearly impossible.

Users report that Datadog's support is not very helpful, with one Reddit user stating 'Support is utter sh*t... they stonewall you with tickets and miss half the questions.' Capterra reviews mention experiences with support have been continuously negative, with the company going out of its way to avoid direct communication and offering irrelevant answers.

Users have reported difficulties in resolving deployment mistakes and billing issues, with frustration about the lack of support and inability to remove metrics from agents or adjust billing accordingly. Once you've turned on features that generate costs, getting them reversed or corrected is extremely difficult.

Datadog has become significantly slower over the last year, with performance potentially affected by ongoing feature work. Users report slow dashboard loading times, delayed metric queries, and lag in the interface. The biggest drawback is the relatively slow pace at which performance issues are addressed.

The incomplete and sometimes inconsistent implementation of OpenTelemetry compatibility is a significant issue, especially noticeable with features such as handling 'resources' in APM. Teams wanting to use open standards find Datadog's OTel support lacking compared to competitors. This creates vendor lock-in concerns for those wanting to maintain flexibility.