🔄 What are Dynamic Lists and what is their true purpose?
Dynamic Lists are "living" segmentation engines. They are sets of customers and prospects that group and ungroup completely automatically based on a set of rules (filters) that you define yourself.
Their fundamental nature is prospective: they are designed to start "listening" and capturing data from the exact moment you click "Create". The system constantly evaluates your database; if a customer meets the requirements in the future, they enter the list immediately; if their conditions change and they no longer meet them, they automatically exit.
⚖️ Technical difference with Static Lists:
While dynamic ones evaluate future interactions to feed recurring campaigns, Static Lists have a strictly retrospective purpose. They serve to "take an exact photograph" and frozen snapshot of your database at a specific point in time.
🌟 Operational benefits of dynamic segmentation
Zero manual load: Forget completely about downloading, cleaning and reloading Excel files every week.
Perfect timing (Living segmentation): By updating in real time, you ensure that your campaigns (e.g. abandoned carts, payment reminders) reach people who match the profile today, not those who matched it a month ago.
Hyper-personalization: It allows you to structure communication funnels where you speak differently to a "Cold Lead" than to a "Recurring Customer", optimizing conversion.
🛠️ Step-by-step detailed guide to create your List
Access Route: Go to the Campaigns > Lists module in your left sidebar menu and select the option to create a new "Dynamic List".
Building logics (Condition Groups): The system allows you to stack filters using fundamental logical operators:
AND connector: Use it within the same group if you need the customer to meet all rules simultaneously (Ex. Is from Mexico AND has the VIP tag).
OR connector: Use it to create alternative groups. It's enough for the customer to meet the conditions of any of the groups to enter the list.
Initial Counter (Preview) and Creation: In the side panel, Atom will show you a real-time preview of how many contacts currently meet your rules. Check that the volume makes sense with your strategy and click "Create". From that second on, the system begins indexing.
🧠 Deep Dictionary of Filters and Conditions
To build infallible segmentations, we have divided all variables into two categories based on how the Atom engine processes time.
🔮 1. Prospective Filters (Capturing data in the future)
What are prospective filters in dynamic lists? Prospective filters are conditions that evaluate events and behaviors in real time. They serve to register only the actions that occur after the list is created.
Rules for how prospective filters work:
Not retroactive: The list does not include past events, state transitions, or customers. It only evaluates information from its configuration onwards.
Initial state: When you create a dynamic list with a prospective filter, it will start completely empty (with 0 contacts).
Real-time evaluation: The system monitors behaviors live.
Immediate inclusion: A customer is automatically added to the list in the exact second they perform the action or meet the condition.
Last Classification: Filters based on the exact reason an agent closed the conversation (includes the internal classification of "Automatic Closure" for inactivity).
⚠️ Important rules to consider: If you set a strict time period (e.g. "exactly 24 hours ago"), the customer will enter the list, but after hour 25 passes, the contact won't be taken into account. Legacy lists that used a separate date field will continue to work fine.
Customer Stage (Prospect and Service): Unifies your sales and support segmentation under a single variable.
⚠️ Important rules to consider: Cross prevention. If you're setting up "Prospect" stages and change the selector to "Service", the system will automatically clear previously selected values to avoid confusing the engine. Also, lists configured in the past as "Prospect Stage" will now display under this new format without affecting their function.
Last assignment (group): Filters based on which was the last team (group of agents) responsible for the service.
⚠️ Important rules to consider: Conversations without a group. If a contact started a chat but was never assigned to a group (still with the bot), the system will interpret that they don't belong to your selection, being successfully captured if you use the "Does not contain" exclusion operator. You can only use this variable once within the same condition group.
Last Message (Unified Filter): Create precise exclusions by evaluating when the last communication was, whether sent by the customer or by the company.
⚠️ Important rules to consider: Behavior of the "Company" option. This is a strict inactivity rule; if you add a customer because you wrote to them and they didn't respond, but suddenly the customer replies, the system will remove them immediately from the list. If there has never been a message from the company in that chat, the contact will be ignored.
Flow abandonment: Catches users who left an automated conversation halfway through.
⚠️ Important rules to consider: To be labeled as "abandonment", 3 inviolable rules must be met: the conversation must have expired forcibly due to timeout, the customer must have sent at least one previous message, and the last message in the conversation must have been from Atom. State cleanup: If you manage to contact them and they resume the conversation, their abandonment status is cleared and they exit the list.
Tags: Filters prospects based on the quick tags assigned to them during the service process.
Lifecycle Stage: Segment automatically by evaluating status updates through direct synchronization with the stages of your sales funnel in HubSpot.
🕰️ 2. Retroactive / Mixed Filters (Bringing historical data)
Analyze static information that is already recorded in the user's profile. Ideal for filtering large databases from the past.
Source: Segment based on the channel through which the conversation originated.
Campaigns: Users who entered through previous mass sends.
Web traffic: Contacts coming from your portals or landing pages.
Click to WhatsApp (CTWA): Precisely groups people who started a chat by interacting with your Meta ads.
⚠️ Important rules to consider for Traffic (CTWA and Web): In "Click to WhatsApp", the menu will only show campaigns that have already generated at least one chat in Atom and the search requires precision (if you changed the name in Meta, you'll need to select both versions). In "Web Traffic Source", selecting "Organic Traffic" will group all those who arrived without tracking parameters (UTMs).
Information field: System attributes (Name, Last Name, Phone, Email) and custom ones for your business (e.g. Brand, Model, City, Branch).
🔎 Search Operators for Information Fields
If you use Information Fields to collect custom data from your customers (such as city, model of vehicle of interest, or their profession), you have at your disposal two advanced operators to speed up the creation of your rules without having to repeat conditions.
1. "Contains" Operator
This comparator is ideal when information is completed inconsistently (for example, mixing uppercase and lowercase), as it allows you to search for a specific word within a longer phrase or text without having to create multiple "Is equal to" conditions.
How to use it?: Select a field of text type and choose the Contains operator. Enter the exact word you want to search for (for example, if the field stores the information "Ciudad de Bogotá 2024", you can write "bogota").
Key rules:
Case-insensitive: "bogota" or "Bogotá" will work the same.
Whole words: Searches for matches of entire words (tokens). Searching for "Bogo" won't return results if the full word is "Bogotá".
2. "Contains any of" Operator
This operator allows you to evaluate if an information field matches a list of several possible values within a single rule. This way you avoid having to create a new condition (separated by OR) for each individual option.
How to use it?: Select your field (can be text, number, or currency) and choose the Contains any of operator.
Value entry: You'll be able to write multiple words or numbers separated by a comma ( , ). For example, if you want to segment customers interested in certain vehicles, you can enter: Yaris, Corolla, RAV4. When you write the comma, the value will be automatically saved as a visual tag (or "pill") within the search bar and the comma will disappear.
Key rules:
If the field is text type, the search will evaluate whole words ignoring uppercase/lowercase.
If the field is numeric, the match must be an exact number.
📋 Duplicate a Condition Group
When you build advanced segmentations that involve multiple groups of rules joined by the "OR" connector, it's common to need to repeat certain combinations of similar filters while changing only a couple of data points.
To save time and make your process much more agile, the platform includes a duplication tool:
How to duplicate a complete group?
Within the creation or editing screen of your Dynamic List, locate the card of the condition group you want to clone.
In the top right corner of that card (next to the "Delete" button), click on the Duplicate group icon (represented by the copy documents icon 📄).
Automatically, the system will create an exact copy of all conditions, internal logical operators (AND/OR) and entered values, positioning the new group just below the original.
⚠️ Important rules and Considerations:
Complete independence: Once you duplicate a group, the copy becomes 100% independent of the original rule. This means you can edit, add, or delete tags, classifications, or date ranges in the new group without these changes altering or affecting the source group.
Maximum group limit: The system allows you to structure a maximum of 5 condition groups within the same list. If you've already reached this limit of 5 cards, the duplicate button will be defensively disabled to prevent overloading the segmentation.
Error carry-over: If the original group you decided to clone had any incomplete field or validation error, the duplicated group will be born preserving exactly the same state so you can correct it manually.
🚨 Core System Processing Rules
To master lists, you must understand how Atom thinks:
The Inheritance Rule (Mixed Filters + Prospective): This is the golden rule. If within the same group (joined by AND) you combine a historical attribute (e.g.
Country = Mexico) with a future event (e.g.Classification = Closed Sale), the prospective filter always overrides the mixed one. This means the list will start completely empty and will only begin catching customers from Mexico who are classified from the moment you published the list. It won't bring those classified in the past.Character sensitivity (Case & Accent Sensitive): With the exception of "Custom Information Fields", you should assume that the standard "Contains" condition in the rest of the platform is strictly sensitive. The evaluated text must match your configuration to the millimeter, including respect for uppercase, spaces, and accents.
⚙️ Management, Advanced Actions and Editing
All administration is concentrated in the quick actions menu (the three vertical