If you want to see how Algolia works, sign up for free and try things out for yourself. Try out one or more of the tutorials, find out how you can use your data, and read about how analytics, relevance, and insightscan help measure the success of your implementation.

Run the Quickstart Assistant tutorial

When you create a new Algolia account, the dashboard’s Quickstart Assistant will offer you several choices. Choose New to Algolia to begin this interactive tutorial.

1

Data options

Follow the Quickstart Assistant prompts. You can use your data, the data provided by the assistant, or import the data used in the ‘Bootstrap’ tutorial.

2

Create demo

Click on Indices in the dashboard’s sidebar menu, click the UI Demos tab, and then the Generate a UI demo button.

3

Customize

Complete the fields you require and click Save.

4

Preview demo

Click the eye icon next to your newly saved demo to preview it.

Try the onboardong tutorial

The two-minute interactive tutorial walks you through the steps required to build an Algolia user interface. It covers choosing a dataset, sending that data to Algolia, examining the effect of changing relevance, and building a user interface.

Create a “bootstrapped” search project

For each library supported by Algolia, there is a ‘bootstrap’ tutorial that takes you through the following:

  1. Installing and initializing the pre-built project
  2. Adding code to display a user interface: a search box with highlighted results
  3. Adding more elements to the UI: filters and pagination
  4. Adding more information and styling to search results.

Once you have explored the bootstrap tutorial, try using your data instead. It will help your overall implementation if you also consider how to measure the success of your Algolia implementation once it’s live by understanding how to interpret analytics, tweak relevance, and capture user insights.

Use your data

In the preceding tutorials, you were probably using pre-built datasets. In reality, you need to transform your data into an acceptable JSON format and then send that data to Algolia using an Algolia API client, no-code connector, CLI, Crawler, or dashboard.

On a live site, you need to create a script to update your index whenever the site content changes: new material, edits, and deletions. However, if you’re using them, the no-code connector and Crawler will handle this synchronization for you.

Analytics, relevance, and insights

Search relevance measures the accuracy of the relationship between the search query and the search results. Algolia has good relevance out of the box, thanks to its ranking criteria. The system also provides tools to help fine-tune your users’ search experience. Of these, the most important ones are:

Review your analytics for details such as popular and ‘no results’ phrases to verify that any changes you made to relevance have positively affected your users’ behavior. These insights let you make informed decisions about adjusting searchable attributes, custom ranking, and other aspects of relevance.

Consider using A/B testing to compare one set of adjustments with another to see which performs best.

Often, search isn’t a standalone goal: it’s an aid to achieving broader aims. For example, the aim of an ecommerce site may be to increase sales, increase profits, or drive a particular product range. By capturing user behavior, you can see what happens after they have performed a search and determine if the search implementation has helped achieve your goals.

Further reading