---
title: "Scope a Data Project - Mozenda web scrapping software"
slug: "scope-a-project"
description: "Starting a new data project always can be some questions in mind like why do you want the data, how to make the data usable and many more. Here you can find the answers."
updated: 2024-05-14T11:23:23Z
published: 2024-05-14T11:23:23Z
canonical: "help.mozenda.com/scope-a-project"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://help.mozenda.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Scope a project

Before you begin a new Mozenda project, consider these questions: :

- Why do you need this data?
- How will you interact with the data?
- Who are the project stakeholders?
- What decisions are influenced by the data?

The answers to these questions will guide the flow of your project and help you achieve your end goal. With a high-level understanding in mind, you can start your project.

### Identify and locate the data

Identify the data you need. List each of the required data points, then choose your source website(s). Next, browse the site and verify that you have access to what you need. As long as you have access to the site the Agent Builder does too.

### Document the navigation process

Document the process after you locate your dataset and verify that you have access. Imagine you're providing step-by-step instructions that a colleague could follow.

Include the following information in your instructions:

- Starting URL
- Navigation path
- Menu selections
- Sign in information
- Text entries
- etc.

### Build the agent

Using the list of instructions, it's now time to create an agent - the script the application will use to carry out your harvesting project. First, load the target website into the Agent Builder, navigate to your starting page, and select **Start a new agent**. Now, follow your documented steps to teach the agent how to collect your data. At each step, Mozenda will recommend the appropriate system **action** needed to proceed.

Select **TEST** to preview the data your agent will capture. Once everything looks good, save your agent and open the Web Console.

          Actions

          

Agents use specific processes called *actions* to extract data. As you gain experience, you'll start to understand which actions correspond to each step of the agent-building process.

### Harvest your data

Now that you've built your agent, all that remains is to run it using the Web Console. In addition to manually running agents, the Web Console allows you to create automated harvesting schedules, run projects in a sequence, and manage the data you extract. Data collections can be viewed in the Web Console, downloaded to your device, or connected to other platforms via native publishing integrations or our awesome API.