Michael Cizmar
President, Managing Director @ MC+A
Last fall, Elastic announced that they were no longer investing in App Search and directed customers to begin looking at using Elasticsearch natively. Earlier this year, they officially announced that it was deprecated with version 9. As such, the Enterprise Server components are no longer available. Elastic’s App Search, part of the Enterprise suite, has long been a go-to solution for developers seeking fast, scalable, and customizable search experiences. However, with Elastic shifting focus toward Enterprise Search built within native Elasticsearch capabilities, many teams are now evaluating migration paths away from App Search. If you’re considering a move, this blog and the associated blueprint outline the strategic, technical, and operational considerations to ensure a successful transition.
Understanding the motivation behind the deprecation
Search is hard
– Unknown, 2025
As organizations evolve their digital experiences, the tools powering search must keep pace. App Search gave users a ready-to-go basic platform. That comes at a cost. The most powerful, AI-first features are native to Elastic, and App Search generally has become a shell that often made it harder to really create compelling experiences.
The core features have been migrated out of the stack. The web crawler can run out of Elastic as well as the connectors. Improvements to the stack with Serverless going to GA mean that organizations with vast amounts of data can look to Elastic to host that. So, while Enterprise Search as a product is dead, the solution can still be stood up in a more native form that can scale to the needs of the use case.
Plan your migration - Assess your current App Search implementation
Before any migration, it’s critical to understand how search is currently delivering value to your business. As we’ve emphasized in previous MC+A articles and client engagements, this isn’t just a technical audit—it’s a strategic checkpoint.
A successful migration begins with clarity. You need to know:
- What your users expect from search.
- How search supports your business goals.
- Where current limitations are holding you back.
This typically comes from an assessment. This assessment helps you prioritize what to preserve, what to improve, and what to rethink as you move beyond Elastic App Search. If you have any trouble with this, MC+A does have an assessment package that you can purchase. But regardless of how you implement the assessment, it should first start with the key business considerations.
Some questions you should ask yourself are:
- Is your current search delivering relevant results—or just matching keywords?
- Are users finding what they need? (and how do you know?)
- Are there gaps in index coverage that impact the user’s experience?
- How tightly integrated is search into your workflows?
- Are your consumers end users? Do you envision agents needing to use your service?
Frame the Migration Strategically
With this assessment completed, you can plan your migration to Elastic native or the product that is best suited for your use case. We would like to point out that most people do not opt for the same phone when they upgrade their device. Likewise, upgrading your search technology to a modern engine can:
- Improve relevance with AI and Semantic Search
- Expand use cases across departments and client offerings
- Operationalize insights from search behavior
The last point is critical for personalization and optimization.
Top innovations in Elastic 9.x
Here are some of the features we think are compelling in the new 9.x Elastic stack:
Better Binary Quantization (BBQ) – Now Default:
This is Elastic’s proprietary vector quantization technique, now generally available and enabled by default in 9.1. It delivers:
- Faster query speeds
- Higher throughput across all recall levels
- 95% reduction in memory footprint
- Seamless integration with semantic search and reranking
ES|QL Joins – The “Holy Grail” of Search
The Elasticsearch Query Language (ES|QL) now supports:
- LOOKUP JOINs for real-time cross-index and cross-dataset queries
- Cross-cluster search (CCS) with resilient architecture
- KQL filters and partial query results
- Advanced text grouping functions for scalable data exploration
Lucene 10 Upgrade
Elastic 9.0 is built on Lucene 10, bringing:
- Improved query performance and lower latency
- Smarter indexing and parallelism
- Hardware-level optimizations
- Simplified index management APIs
LLM Observability - Native GenAI Monitoring
Elastic now offers native observability for LLMs, including:
- Prompt and response tracking
- Performance and latency metrics
- Cost and usage visibility
- Support for OpenAI, Azure OpenAI, Amazon Bedrock, and Google Vertex AI
Elastic Cloud Serverless
Elastic Cloud Serverless is now GA on AWS, GCP, and Azure, offering:
- Instant deployment without managing infrastructure
- Full access to Elastic 9.x features
- Scalable search, observability, and security solutions