Northbeam connector for ChatGPT: ROAS, CAC, models

Platform	Spend	Attributed rev	ROAS
Meta	$12,400	$34,700	2.8
Google Ads	$8,100	$27,500	3.4
TikTok	$9,300	$15,800	1.7
Pinterest	$2,600	$6,200	2.4

01Setup

How to connect Northbeam to ChatGPT

Porter’s Northbeam connector for ChatGPT takes 3 steps, with no code and nothing to host. Create an API key in Northbeam and paste it into Porter with your Data-Client-ID. Add Porter’s MCP URL to ChatGPT as a custom MCP server and ask your first question. MCP (Model Context Protocol) is the open standard ChatGPT uses to connect to tools.

Before you start

You need access to Settings > API Keys in Northbeam, as well as a free Porter account and a ChatGPT plan that supports custom MCP servers.
Custom MCP servers are in beta in ChatGPT on the web, so set up Porter and ask your questions from a browser. On Business, Enterprise or Edu, an admin or owner turns on developer mode; on Pro, you turn on developer mode yourself (OpenAI’s guide to custom MCP apps).
Works in ChatGPT on the web (chatgpt.com, custom MCP servers in beta) and in Codex, with the same Porter MCP URL.

Step 1: Create an API key, then add it to Porter
In Northbeam, open Settings > API Keys, then choose Create new API key to see the key and your Data-Client-ID (Northbeam’s guide to creating an API key and Data-Client-ID). Sign up free at Porter, pick Northbeam as the source and paste both, then pick ChatGPT as the destination.
Step 2: Add the connector to ChatGPT
In ChatGPT, open Customize › Plugins, click Add and choose Add custom MCP server. On Business, an admin does this from Workspace settings › Apps › Create. Name it Porter Metrics, paste this URL as the Server URL and keep OAuth as the authentication:
MCP URL
https://mcp.portermetrics.com/mcp
Tick “I understand and want to continue”, then select Create as a plugin.
Step 3: Ask ChatGPT about Northbeam
Next, select Continue to Porter Metrics, sign in to Porter and click Allow Access. Then select Try in chat and paste this prompt:
Prompt
Using Northbeam, show Spend, Attributed Rev and ROAS by Platform (Northbeam) for the last 7 days on the Clicks + Modeled Views model, 7-day window, accrual accounting.

Guides and videos

Connect Porter to ChatGPT with MCP, the full setup guide
See all 647 Northbeam fields with labels and IDs
Porter’s YouTube channel, setup walkthroughs and use cases on video

Connect Northbeam to ChatGPT free

app.portermetrics.com / create

Create: Northbeam to ChatGPT

Source: NorthbeamSourceDestination: ChatGPTMCP

Select the accounts ChatGPT can use.

ChatGPT / Customize / Plugins / Add / Add custom MCP server

Remote MCP server URL

https://mcp.portermetrics.com/mcp

ChatGPT

Using Porter / Northbeam …

Demo data

02Use cases and prompts

What you can ask ChatGPT

Each use case comes with a prompt to copy, the fields it uses from the Northbeam fields list and who it’s for. Every prompt names an attribution model and accounting mode, plus a window in accrual, because Northbeam returns one row for each combination.

Can ChatGPT show which channels each attribution model credits?

The problemClicks-Only says a prospecting channel pays for itself, Last Non-Direct Touch says it doesn’t, and the budget call stalls on which model to believe.

What ChatGPT doesChatGPT pulls Spend, Attributed Rev and ROAS by platform for several models, then groups them by Attribution Model in one table. Then it flags the platforms that change rank. Spend repeats on every model row, so ChatGPT counts it once.

What you getYou see which channels depend on the model you pick before you move budget between them.

Prompt

For the last 30 days, show Spend, Attributed Rev and ROAS by Platform (Northbeam) and Attribution Model, for Clicks-Only, Clicks + Modeled Views and Last Non-Direct Touch, 7-day window, accrual accounting. Which platforms change rank the most between models?

For: DTC media buyers · agencies · growth leads

Fields used: SpendAttributed RevROASPlatform (Northbeam)Attribution ModelAttribution WindowAccounting Mode

Can ChatGPT show what a new customer costs per platform?

The problemBlended CAC looks healthy while the cost of a first-time buyer on a prospecting channel keeps climbing underneath.

What ChatGPT doesChatGPT ranks platforms by CAC (1st time) together with Spend, Rev (1st time), Transactions (1st time) and Customers (New). It sticks to one model, a 7-day window and accrual accounting, so each platform gets one row. The new vs returning split is accurate when Northbeam gets the same customer ID with every order.

What you getProspecting budgets then get judged on what a first order costs instead of a blended CAC that returning buyers make look better.

Prompt

For the last 30 days on Clicks + Modeled Views, 7-day window, accrual, rank Platform (Northbeam) by CAC (1st time) and show Spend, Rev (1st time), Transactions (1st time) and Customers (New). Flag any platform whose CAC (1st time) is more than double the account's.

For: e-commerce marketers · media buyers · CMOs

Fields used: CAC (1st time)Rev (1st time)Transactions (1st time)Customers (New)SpendPlatform (Northbeam)Attribution ModelAttribution WindowAccounting Mode

Can ChatGPT explain why Northbeam shows lower Meta ROAS than Ads Manager?

The problemAds Manager shows one ROAS while Northbeam shows a lower one, so the creative team quotes whichever looks better.

What ChatGPT doesChatGPT shows Northbeam’s ROAS for Meta campaigns compared to FB ROAS (7D Click, 1D View), Meta’s own figure as Northbeam imports it. Then ChatGPT names the campaigns with the widest gap. It explains the gap by model, window and customer journey (Northbeam’s guide to ad performance discrepancies).

What you getThe ROAS argument ends with campaign names and a reason for each gap instead of two screenshots.

Prompt

For the last 14 days, list Meta campaigns with Spend, ROAS on Clicks + Modeled Views, 7-day window, accrual, and FB ROAS (7D Click, 1D View). Sort by the gap between the two and explain the 3 largest gaps.

For: media buyers · creative strategists · agencies

Fields used: ROASFB ROAS (7D Click, 1D View)SpendCampaignPlatform (Northbeam)Attribution ModelAttribution WindowAccounting Mode

Can ChatGPT show which campaigns pay back on LTV?

The problemA prospecting campaign looks weak on 1-day ROAS, while the buyers it brought keep ordering months later.

What ChatGPT doesChatGPT sorts campaigns by LTV ROAS and LTV CAC, then sets them next to 7-day ROAS and Forecasted Attributed Rev (90d), Northbeam’s own revenue forecast.

What you getYou keep or cut prospecting on what its buyers return over time rather than on their first week.

Prompt

For campaigns with Spend over $1,000 in the last 60 days on Clicks + Modeled Views, accrual, compare ROAS on the 7-day window with LTV ROAS and LTV CAC, and add Forecasted Attributed Rev (90d). Which campaigns look better on LTV than on 7 days?

For: founders · CMOs · media buyers scaling prospecting

Fields used: LTV ROASLTV CACROASForecasted Attributed Rev (90d)SpendCampaignAttribution ModelAttribution WindowAccounting Mode

Can ChatGPT show how much Shopify revenue gets attributed?

The problemThe founder reports Shopify’s Total Sales while the media buyer reports Northbeam’s Attributed Rev, so the weekly meeting opens with the gap between them.

What ChatGPT doesWith Shopify also connected in Porter, ChatGPT sets weekly Attributed Rev, Spend and MER next to Shopify Total Sales. Then it works out the share Northbeam attributes. Totals still differ, because each tool handles taxes, shipping and returns its own way.

What you getFounders and media buyers start the meeting from one weekly table instead of two exports.

Prompt

For the last 8 weeks, show Northbeam Spend, Attributed Rev and MER by week on Clicks + Modeled Views, cash accounting, next to Shopify Total Sales by week. What share of Shopify sales does Northbeam attribute each week?

For: marketing analysts · founders · growth leads

Fields used: SpendAttributed RevMERDateTotal Sales

03Why Porter

Why Porter for Northbeam in ChatGPT

Porter puts Northbeam’s attribution numbers in ChatGPT next to Shopify and your ad platforms. For single orders or saved Northbeam dashboards, use Northbeam’s official MCP, which comes with every Northbeam plan. You can run both.

✓

Every model in one answer

Porter returns each attribution model, window and accounting mode as columns. ChatGPT can then set Clicks-Only, Last Non-Direct Touch and the other models side by side in one table.

⊞

Shopify and ad platforms in the same chat

Ask about Northbeam next to Shopify, Meta Ads, Google Ads and other Porter connectors in one question, so analysts and media buyers work from the same weekly numbers.

⚙

Every client in one question

Agencies connect each Northbeam client with its own API key and Data-Client-ID, then ask one question across all of them. ChatGPT tells the brands apart with the Client field.

Safe for live accounts

ChatGPT reads every platform, campaign and model, then tells you where to move budget. Your Northbeam account stays exactly as you set it, and you can revoke the API key anytime.

↻

647 ready-made fields

ChatGPT gets 647 named fields, such as CAC (1st time), LTV ROAS and Forecasted Attributed Rev (90d), instead of export files you join by hand.

✓

One URL, nothing to host

Add one MCP URL to ChatGPT on the web and Porter runs the rest. The free plan includes the MCP, so your first Northbeam account costs nothing to try.

⊞

Fresh export on every question

Porter runs Northbeam’s export each time ChatGPT asks and stores nothing in between, so every answer reflects Northbeam’s numbers at that moment, including recent days Northbeam is still processing.

⚙

Live dashboards, not only chat answers

Ask ChatGPT for a dashboard and Porter’s AI dashboards publish it at its own link, synced automatically and ready to share. Official MCP servers answer inside the chat and stop there.

04Dashboard

Dashboard template

Start from a ready-made Northbeam template, or ask ChatGPT to build a live one with AI dashboards.

AI dashboards

Build a live Northbeam dashboard from the chat

Describe the dashboard in one sentence and Porter builds it. A client-ready dashboard in minutes, with no chart dragging and no BI tool to learn. Ask ChatGPT:

Prompt

Build a live Northbeam dashboard for my account with the last 30 days, the main KPIs and a weekly trend.

Live URLEach dashboard at its own link: no login, no app, no download

Auto-syncFresh Northbeam data, nobody re-exports each period

Scheduled deliveryEmail or PDF sends on a set day

White labelYour logo and branding for clients

Sharing permissionsView or edit access per person

Or start from a template

Dashboard templates for every source ↗Ads, social, SEO, ecommerce and CRM

Porter's template gallery ↗Every Porter template in one place

05Pricing

Pricing

You can start free, since Porter’s Northbeam connector for ChatGPT comes with every plan. Use the calculator for your number of accounts, or see Porter pricing for every plan.

What you get

Free plan: never expires, with unlimited connected accounts and the last 15 days of data in ChatGPT (30 days in Data Studio and Sheets)
Reporting: from $12.50 per account a month billed annually, or $15 month to month for one account
Billing: each Northbeam client counts as one account, with volume discounts and no per-seat fee
No credits needed: ChatGPT’s Northbeam questions are reads, so they use no AI credits
Trial: 14 days of full access, no card

See full pricing and plans

Price calculator

Data source accounts

$33.33/mo$6.67 per account

An account is one ad account, Stripe account, store, page or GA4 property. The Free plan never expires and has unlimited accounts, with the last 15 days of data in AI chats.

06Data and actions

Northbeam data ChatGPT can use

The Northbeam connector gives ChatGPT 647 fields (632 metrics and 15 dimensions) from Northbeam's Sales Attribution export. They cover metrics such as spend, attributed revenue, ROAS, CAC, MER, LTV, forecasts, customers and platform-reported metrics, from platform down to ad level (full list of Northbeam fields).

12 Northbeam fields ChatGPT uses most

Field	What it's for
Attributed Revnorthbeam_rev_attributed	Revenue the model on the row credits
ROASnorthbeam_roas	Attributed revenue per dollar of spend
CACnorthbeam_cac	Spend per customer acquired
CAC (1st time)northbeam_cac_ft	Cost of a new customer
MERnorthbeam_mer	Blended revenue divided by spend
Spendnorthbeam_spend	Spend across connected platforms
Transactionsnorthbeam_txns	Attributed orders
LTV ROASnorthbeam_roas_ltv	ROAS on the lifetime window
Forecasted Attributed Rev (90d)northbeam_forecasted_rev90	Northbeam's 90-day revenue forecast
FB ROAS (7D Click, 1D View)northbeam_meta_roas7_dclick1_dview	Meta's own ROAS as Northbeam imports it
Attribution Modelnorthbeam_attribution_model	The model on the row
Accounting Modenorthbeam_accounting_mode	Accrual or cash

Product stats

What	Northbeam connector
Fields	647 (632 metrics, 15 dimensions) from Northbeam's Sales Attribution export
Actions	Reads only; your Northbeam account stays as you set it
Freshness	Export runs on every question; nothing stored in between
History	Northbeam allows up to 2 years per fixed date range; no Porter limit on paid plans; 30 days on the Free plan, 15 in AI chats
Speed	Each export takes a few seconds to several minutes; Northbeam allows 60 export requests per minute
Levels	Platform, campaign, ad set and ad, by day; no hourly data
Access	Northbeam API key plus Data-Client-ID

See all 647 Northbeam fields

07Security

Is it safe?

Yes. Porter's Northbeam connector for ChatGPT reads data through the API key and Data-Client-ID you create, and your orders, spend records and settings in Northbeam stay exactly as you set them.

ChatGPT goes through Porter

Porter runs Northbeam's export only when ChatGPT asks a question, and ChatGPT never connects to Northbeam directly.

What an API key can reach

One Northbeam API key covers the whole client, because the export, orders and spend endpoints accept the same two headers. Porter only calls the export, but every request needs both the key and its Data-Client-ID.

08FAQ

FAQ

Media buyers and agencies usually ask about several brands, attribution models and windows, as well as freshness and dashboards or orders. Then come 3 fixes: numbers that don't match Northbeam, totals that look too high and slow answers.

Can I connect several Northbeam dashboards or brands?

Yes. Each Northbeam client, with its own API key and Data-Client-ID, connects as one account and also bills as one. Agencies add each brand the same way, then ask one question across all of them. ChatGPT still tells the brands apart with the Client field.

Which attribution models and windows can ChatGPT use?

As long as your Northbeam export offers them, ChatGPT uses any of Northbeam's 7 attribution models. Northbeam's attribution windows run 1, 3, 7, 14, 30 and 90 days plus LTV, but exist only in accrual accounting. So name the model in every prompt, and the window in accrual.

How fresh is the Northbeam data, and how far back does it go?

Porter runs Northbeam's export each time ChatGPT asks and stores nothing in between, so answers reflect Northbeam's data at that moment. Recent days can still change while Northbeam processes new orders and spend. Porter's paid plans set no history limit of their own, while the free plan reads 15 days in AI chats (30 days in Data Studio and Sheets).

Can ChatGPT see my Northbeam dashboards or orders?

The Northbeam connector for ChatGPT reads Sales Attribution metrics: spend, revenue, ROAS, CAC and LTV by platform, campaign, ad set and ad. For a specific order or a saved dashboard view, connect Northbeam's official MCP next to Porter, since it lists dashboards and reads orders.

Troubleshooting

Why don't the numbers match Northbeam?

ChatGPT's numbers match the Northbeam dashboard only when both use the same account, date range, timezone, attribution model, window, accounting mode, level and export options. Recent days also shift while Northbeam processes orders, so compare a date range that ends yesterday.

Why do Northbeam totals look too high?

Ask for one attribution model, window and accounting mode, or group by all three. Northbeam returns one row per combination and Porter passes every row to ChatGPT, so adding those rows up counts the same revenue several times. Northbeam's accounting modes explain why cash and accrual differ.

Why does a Northbeam question take a minute?

Ask for several metrics in one question. Porter runs an asynchronous Northbeam export for each question, which takes a few seconds to several minutes, and Northbeam allows 60 export requests per minute (Northbeam's Data Export API rate limits).

09Next

More ChatGPT connectors

Attribution makes more sense next to the ad platforms it measures, so Porter brings each of them in through the same MCP URL. Add the platforms your budget runs on, or browse all ChatGPT connectors instead:

→

Meta Ads connector for ChatGPT

→

Google Ads connector for ChatGPT

→

Shopify connector for ChatGPT

→

WooCommerce connector for ChatGPT

→

Amazon Seller connector for ChatGPT

Not only ChatGPT. Send Northbeam data anywhere

The same Porter account and connection also send Northbeam data to these destinations, at no extra cost.

→

Northbeam to Claude

→

Northbeam to Google Sheets

→

Northbeam connector for any AI tool (MCP)

Try the connector in ChatGPT

Setting up the Northbeam connector for ChatGPT takes one API key and one URL. Then ask ChatGPT which channels each attribution model credits, what a new customer costs and how much Shopify revenue Northbeam attributes. Since the free plan includes the MCP, your first account costs nothing to try.

Free plan · 14-day trial · No card needed

Media buyers: model comparisons and Meta ROAS gaps in one answer instead of three dashboard views.

Founders: new-customer CAC and LTV payback by platform before the budget meeting.

Agencies: every client's Northbeam numbers through one connector, so one question covers all of them.

About the author: Porter Metrics team, who build Porter's connectors for ChatGPT and Claude.Last updated October 20, 2026

How to connect Northbeam to ChatGPT

Before you start

Step 1: Create an API key, then add it to Porter

Step 2: Add the connector to ChatGPT

Step 3: Ask ChatGPT about Northbeam

Guides and videos

Create: Northbeam to ChatGPT

Remote MCP server URL

ChatGPT

What you can ask ChatGPT

Can ChatGPT show which channels each attribution model credits?

Can ChatGPT show what a new customer costs per platform?

Can ChatGPT explain why Northbeam shows lower Meta ROAS than Ads Manager?

Can ChatGPT show which campaigns pay back on LTV?

Can ChatGPT show how much Shopify revenue gets attributed?

Why Porter for Northbeam in ChatGPT

Every model in one answer

Shopify and ad platforms in the same chat

Every client in one question

Safe for live accounts

647 ready-made fields

One URL, nothing to host

Fresh export on every question

Live dashboards, not only chat answers

Dashboard template

Build a live Northbeam dashboard from the chat

Or start from a template

Pricing

What you get

Price calculator

Northbeam data ChatGPT can use

12 Northbeam fields ChatGPT uses most

Product stats

Is it safe?

ChatGPT goes through Porter

What an API key can reach

FAQ

Can I connect several Northbeam dashboards or brands?

Which attribution models and windows can ChatGPT use?

How fresh is the Northbeam data, and how far back does it go?

Can ChatGPT see my Northbeam dashboards or orders?

Troubleshooting

Why don't the numbers match Northbeam?

Why do Northbeam totals look too high?

Why does a Northbeam question take a minute?

More ChatGPT connectors

Not only ChatGPT. Send Northbeam data anywhere

Try the connector in ChatGPT