Poper

Read projects, domains, campaigns, captured leads, and performance analytics from the Poper AI popup builder. Agents pull captured leads into the systems that will actually work them.

11 actions API key available

Poper website ↗ Platform Documentation ↗ Capability data verified 2026-08-02

Hourly pull of newly captured submissions across the account's campaigns

List Leads returns the submissions for the form, quiz and NPS campaigns in the window

Agent separates completed submissions from pending ones and keeps them apart

Each completed lead is matched against the CRM for an existing record

Completed, unmatched leads are created in the CRM and routed to the owning rep

Marketing reviews the pending submissions and the low-score responses before anyone is contacted

What This Integration Enables

Poper's public API is read-only, and stating that plainly is more useful than dressing it up. There are no create, update or delete endpoints, so a FlowRunner agent cannot build a popup, change a campaign, or alter anything in the account. What it can do is get everything out: projects with their related domains, the flat account-wide domain list, and campaigns filtered by category or explicit type with shortcuts for popup, form, quiz and NPS survey types. It also returns the leads and responses those campaigns captured, and the same aggregate analytics the Poper dashboard displays for a domain or for named campaigns.

That read-only shape decides what this connector is for. Popups capture a lead and then the lead sits there, which is the actual problem: value accrues at the moment the submission reaches the CRM, the sequence and the rep, not at the moment it is captured. An agent that pulls on a schedule, deduplicates against existing records and routes to an owner closes that gap without a human exporting a file every Monday.

One field in the lead response deserves more attention than its size suggests. List Leads can be limited to completed submissions, pending submissions, or both. A pending submission is somebody who opened a form and did not finish it. They are in the data, they are not a lead, and the difference between those two facts is the difference between a normal follow-up and an unwelcome one. The flows on this page keep that distinction intact all the way to the routing step.

Without FlowRunner

Leads sit in the popup tool Captured submissions live where they were captured and reach the CRM on an export cadence

Popup performance reviewed by hand Someone opens the dashboard, reads the numbers, and retypes them into a report

Every submission treated the same A finished form and an abandoned half-filled one arrive in the same undifferentiated list

With FlowRunner

Leads pulled on a schedule New submissions move into the systems that will work them without anyone exporting a file

Performance pulled with the leads Domain and campaign analytics arrive on the same cadence and land in the same report

Completion state carried through Completed and pending submissions stay distinguishable all the way to the routing decision

Use Case Scenarios

Captured leads into the CRM on an hourly cadence

An agent runs every hour. It calls List Form Campaigns to resolve the current form campaign identifiers, then List Leads scoped to those campaigns with a short lookback window and the filter set to completed submissions only. Each result is checked against HubSpot for an existing contact on the same email. New ones are created with the campaign and domain recorded as the source so attribution survives, and existing ones get the submission logged as an activity rather than being duplicated. The rep sees the lead in the tool they already work in, on the same day it was captured.

NPS responses routed by score, not by volume

A team runs an NPS survey through Poper. The agent calls List NPS Survey Campaigns to resolve the campaigns, then List Leads to pull the individual responses and their scores. Promoters are logged and flagged as candidates for a review request. Passives are logged and left alone. Detractors are not routed anywhere automatically, because the right response to a low score is a conversation and not a sequence, so they go to a person with the score and the verbatim comment attached.

Weekly performance reported without opening the dashboard

A reporting flow calls List Domains to resolve the exact domain values the analytics operations require, then Get Domain Analytics for the last seven days per domain and Get Campaign Analytics for the campaigns the team is actively testing. Both accept preset windows or a custom range, so the same flow serves the weekly and the quarterly view. The output lands in Google Sheets alongside the lead counts pulled in the same run, so campaign performance and what the campaign actually produced sit in one table.

Human-in-Loop Highlight

The connector is read-only, so nothing the agent does inside Poper can be wrong in a way that needs undoing. The irreversible act is the first outbound touch, and it happens one system later. When the hourly pull returns submissions, some of them are pending: a person opened the form, typed an address into the first field, and closed the tab. Their address is in the response. They did not consent to anything. Emailing them is unrecallable, it is the fastest way to get a domain reported, and in several jurisdictions it is unlawful. The same caution applies to a detractor NPS response, where a promotional sequence arriving after somebody just told you they are unhappy actively damages the account.

So the agent enrolls nobody from either group. It routes the completed, consented submissions on its own authority, and posts the rest to the marketing owner: "This hour: 41 completed submissions routed to the CRM. Held: 12 pending submissions with an email captured but the form unfinished, and 3 NPS responses scoring 4 or below. Hold as pending, add to the nurture list, or discard?" A person decides what happens to people who did not finish saying yes. That is human-in-the-loop doing exactly the job it exists for, which is standing between an automated system and the people it would otherwise contact without being asked.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via email

Human decides

Agent resumes with decision

Agent Capabilities

11 actions

Account

Verify API Key Validates the configured Poper API key against the ping endpoint and returns the email address of the authenticated account. Used to confirm credentials before running other operations, and as a connection check in a scheduled flow.

Projects

List Projects Retrieves every Poper project available to the key owner. Each project includes its related domains, and the response also carries a flat domains array covering all projects. Used to discover the project and domain identifiers that the campaign, lead and analytics operations require.
List Domains Retrieves the flat list of domains connected to the account. This is the account-wide list the analytics operations draw on, and they require an exact domain value, so resolve it here rather than typing it.

Campaigns

List Campaigns Retrieves campaigns owned by the key holder, optionally narrowed by category, explicit campaign types, project or domain. Campaign types override the category filter when both are supplied. The response includes a per-type breakdown and the list of campaign types the account supports.
List Popup Campaigns Retrieves only popup campaigns for the account, a shortcut for the campaign list filtered to the popups category, accepting optional project, domain and limit filters.
List Form Campaigns Retrieves only form campaigns for the account. Use the returned campaign identifiers with the lead operations to pull the submissions each form captured.
List Quiz Campaigns Retrieves only quiz campaigns for the account. When quiz campaigns are present the response may also include aggregated quiz statistics alongside the campaign list.
List NPS Survey Campaigns Retrieves only NPS survey campaigns for the account. Use the returned campaign identifiers with the lead operations to pull individual survey responses and their scores.

Leads

List Leads Retrieves the leads and responses captured by Poper campaigns. Target specific campaigns by identifier, or select them implicitly with the category, project and domain filters. Results are paginated at 100 per page by default and can be limited to completed submissions, pending submissions, or both, with an optional lookback window in days. The completed and pending distinction is the one to build routing logic around.

Analytics

Get Domain Analytics Returns aggregate performance analytics for a single domain, using the same aggregation the Poper dashboard displays. Choose a preset reporting window or supply a custom start and end date. The domain must match a connected domain exactly, so resolve it with List Domains first.
Get Campaign Analytics Returns aggregate performance analytics for one or more campaigns on a given domain. Both the domain and at least one campaign identifier are required. Supports the same preset and custom reporting windows as the domain analytics operation.

Frequently Asked Questions

What can FlowRunner do with Poper?

FlowRunner agents can run Verify API Key, List Projects, and List Domains in Poper, plus 8 more actions.

Does connecting Poper to FlowRunner require OAuth?

No. Poper connects to FlowRunner with an API key, no OAuth flow required.

Can Poper trigger a FlowRunner workflow automatically?

Poper doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with Poper

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us