Following Google’s announcement that it would be phasing out their enterprise solution for search, known as Google Search Appliance (GSA), a lot of enterprise customers need to start searching for a worthy alternative search engine. If your organization is one of those impacted, an exciting, educational and fun journey awaits you.
As search solution for a website or a company is an essential part of the technology landscape nowadays, one cannot simply replace GSA without thinking about the vision and direction your organization and projects are heading to. To guide you through this exercise, we provide you with an overview of the decision criteria for evaluating your GSA replacement.
What is the Google Search Appliance (GSA)?
Google Search Appliance (GSA) provides universal search for internal business systems or website content through one easy search box. The solution consists of a hardware (a yellow box which you can just drop into your existing IT landscape) accompanied with a pluggable search software. Moreover, it provides an administrator interface in which non-technical users can easily dig into search analytics or configure and modify settings of the search engine.
Being a great solution for full text searching, I rather found it’s querying possibilities rather inflexible and limited. During the lifecycle of your project, you will find out it is very common that you need to tweak and modify search algorithms to bring more value to your users or website visitors. Therefore, better solutions are available on the market.
Key requirements for an enterprise search system
In order to find the right replacement for your search appliance, explore the market for different search engine options. Start by lisitng the required functionalities of the search engine and combine them with the short and long term vision and goals from your business stakeholders. The following four factors should be considered:
1. Data connectivity and search types
How do you want to embed your solution into your enterprise environment? Establish which functionalities do you have to serve to your end visitors, if you require ‘auto-suggestions’ or ‘Did you mean’ features on your website search, and the types of search you wish to provide. For example, do you require faceted search (drill-down) capabilities? You also should consider how to feed the available data within your enterprise into the engine to get added-value for aggregating and querying and which information from which sources should show within the search results.
2. Content processing and indexing
Feeding data into the search engine might require additional processing of the data to meet the business requirements. Which data needs full text searching or exact querying (Do you want to search within the texts of documents or only on the title of a document)? Do you need to tweak and adapt text analyzers (to support multiple languages and certain domain/enterprise terminology)? GSA does this pretty much out of the box for you with a focus on full text search. However a lot more value can be achieved from your search engine when you take control of your indexing process. For example if your website contains time-critical information, you can decide to make these specfic items show up immediately when they are available on your website, instead of indexing them every hour.
3. Querying and search relevance
Some search engine functionalities are getting pretty standard nowadays. A search replacement should support search suggestions (the ability to give a list of most used search terms), faceting (drill-down within search results), synonyms (if you search for ‘television’, the search engine will also search for ‘TV’), pagination, etc. But keep an eye to cutting-edge functionalities and research that made it’s way into modern search engines. For example, elastic search (an open source search engine solution) also supports aggregation to find uncommonly common terms in your data set: these are terms that are statistically unusual and are usually indicative of something interesting in your data.
4. Security, administration and infrastructure
Consider upfront how you will integrate the search solution within the IT-landscape of your organization. While GSA provides you with a point and click interface, other engines require technical configuration and management through xml or json format. Clustering & scaling (to make the search engine work smoothly even when you have high website visitor traffic), high availability (always up and running) and incorporation in your disaster recovery plan should be designed, communicated and planned beforehand. There are numerous options in authentication and authorization with a search engine, ranging from basic authentication to PKI infrastructure. Watch out if you are considering a cloud-based SAAS solution, as the authentication possibilities are often more limited.
Wrapping things up: Managing and migrating from Google Search Appliance should not be taken lightly and will depend highly on the AS-IS and TO-BE situation within your organization and digital strategy. Thinking upfront about the key requirements for your search engine, developing a proof of concept, communicating and planning the project are the key success factors in migrating from your current GSA.
Your digital partner should should have the necessary knowledge in search technology and expertise in planning, migrating and keeping an eye on the roadmap of the tools your organization is using to ensure they are the best fit for your specific context.
Let us help you find the right search solution for your organization’s digital strategy and IT environment