This article explains how to start with Flagship, including basic concepts and giving you keys to create your first feature.
🧙 Technical implementation
To integrate Flagship into your website, you need a technical team ready to implement some code into your codebase. It usually takes several hours/days of implementation depending on the size of your team and the number of flags that you will need at the beginning.
You will need to carry out the following steps:
- Setting-up the Decision API/SDK with unique visitorId and the user context (workload: around ½ day)
- Defining flags in your code (workload: around 10min/flag)
- Defining business KPIs or performance KPIs in your code (workload: less than 10min/event)
- Setting-up targeting keys in your code (workload: less than 10min/keys)
Step 1: Setting up the Decision API
The Decision API is a REST API that serves as the engine behind Flagship. Implementing it directly into your application will let you and your team fine-tune each call and digest every response without any additional layers between Flagship and your applications.
The Decision API is language-agnostic by design, meaning you can use Flagship even if your stack includes a variety of programming languages, or if you would like to experiment with more specialized languages or frameworks for which we don't offer a dedicated SDK yet. If you'd rather use an SDK with preconfigured methods to implement the Decision API, you can see the full list of available languages and learn more about SDK-specific features.
For more information on the Decision API, refer to Decision API for non-techie users.
For technical information or if you would like to take the implementation further, see the developer portal.
Step 2: Defining your flags
When the Decision API or the SDK is implemented, you need to define your flags. Flags are variables you want to act on with a specific value to manage your features. It’s important to set them in your code otherwise you will not be able to modify them through Flagship or get a report of your feature’s performances.
Once they are defined, you can get an overview of all the flags you have implemented.
Step 3: Defining your KPIs
Defining business metrics is the key to tracking the performances of your features.
In your codebase, the KPIs will be labelled with tracking hits sent to the Decision API so that it can calculate the performance of your feature, such as conversion or transaction rates.
There are 4 types of KPIs available:
- Transaction
- Event
- Pageview (on Web/Server)
- Screenview (on APP/Server)
For technical information or if you want to take the implementation further, see the developer portal (Pss; Every SDKs embed the collect part; if you are using one, you don’t need to carry out a manual POST call).
Step 4: Setting up your Targeting keys
To assign your visitors to a feature, you need to pass some user context keys inside the Decision API call, also used as targeting. Otherwise, you will still be able to use Flagship, but you will be able to target only specific visitorIds or all your visitors.
User context keys can enable you to filter your reports during the analysis of your feature.
For the technical implementation, refer to the Quickstart & integration section of the developer portal.
⛵ Using Flagship
You can now start to use Flagship. Here are the basic steps to start with:
Creating your first use case
To create a use case, apply the following steps:
- Go to the Flagship dashboard.
- Click the + button.
- Choose a name for your project and click Create.
- Select a feature for your use case (for instance, Progressive rollout) and click Choose this template.
Entering the basic information
First, you need to enter the basic information of your feature:
- The feature name: use the most representative name for your feature, because this is the one you'll need to remember in case you want to find it later.
- The feature description: explain exactly what your feature deployment is about and what its purpose for your business is.
Choosing the KPIs to follow
This step is optional, but its goal is to define the objectives of your feature.
You can select a primary KPI, which will serve as a point of reference for your feature and one or several secondary KPIs (Transaction KPI or Event KPI).
For more information, refer to Configuring KPIs.
Defining flags
Here, you can choose the flags which will interact with your feature.
The name of the flag you configure must be the same as the one in your codebase.
Defining targeting
During this step, you need to define which users will see your feature.
The 3 following options are available:
- Defining it to All Users if you want all your users to progressively see your feature.
- Defining it to Users by ID if you want only users with a specific ID to see your feature.
- Defining it as a Key if you want only users matching this key value to see your feature.
Defining deployment steps and rollback option
The deployment step allows you to plan the start date, the action perimeter, and the volume of the first users targeted by your deployment. Then, it allows you to schedule your other deployment steps with fixed steps (defined with the first one) or customs ones.
Finally, you may choose the rollback option for your deployment. I invite you to look at this article about the automatic rollback option. It will help you to understand how it works!
Reading the summary (needs tech supervision)
The summary is the final step. It summarizes the feature configuration. This step is very important to check if you want any information about your feature.
Enabling your feature
Once you have configured all the steps of your feature, it is OFF by default to allow you to check that it is correctly configured and give your marketing team some time to finalize the communication strategy.
You can switch your feature ON when you want it to be visible to your targeted audience.
Analyzing the feature results
After activating your feature, you must wait until you have enough insights from your users to know if your feature is performing. To check the reliability status, click the Reporting button of your feature from the dashboard.
For more information, refer to How can I know my test is reliable and my data significant enough to be analyzed?
Need additional information?
Submit your request at product.feedback@abtasty.com
Always happy to help!