goal_N — conversion / goal events

goal_N is a family of materials, not a single one. Each site can define up to ten goals, and each goal gets its own material named goal_1, goal_2, …, goal_10. A row in goal_3, for example, is "someone triggered goal #3 on this site".

Use goal_N whenever the question is about conversion: how many goals fired, which pages drove them, which referrers, what the path to conversion looked like. If the user says "conversion", "signup", "purchase", "lead", "contact form submission", think goal_N.

Which goal is which?

The /guide endpoint returns a sites array, and every site has a goals list containing id, name, and a human-readable condition. That is where you look up which goal number means what. Never guess that goal_1 is "signup" — it depends on how the site owner configured their site.

A typical sites[0].goals response:

[
  { "id": 1, "name": "Contact form submit", "type": "gtype_page",
    "condition": { "url": "/contact/thanks", "match": "exact" } },
  { "id": 2, "name": "Newsletter signup",   "type": "gtype_click",
    "condition": { "page": "/", "selector": "#newsletter-submit" } },
  { "id": 3, "name": "Free trial started",  "type": "gtype_page",
    "condition": { "url": "/trial/start",   "match": "exact" } }
]

If the user asks "how many signups did we get this week?", you look at that list, pick the goal whose name is "Newsletter signup" (or whichever matches intent), and query the corresponding goal_N material.

Sample rows

Hand-crafted sample for goal_3 ("Free trial started"). Only page_id, session_id, pv_id shown as IDs.

pv_id	session_id	page_id	access_time	url	referrer	device	browse_sec
91030	48221	1305	2026-04-14 09:14:22	/trial/start	google / organic	desktop	38
91033	48224	1305	2026-04-14 10:02:11	/trial/start	(direct)	mobile	72
91041	48228	1305	2026-04-14 11:48:55	/trial/start	twitter.com	mobile	24
91048	48230	1305	2026-04-14 13:22:07	/trial/start	google / ads	desktop	19

Things to notice at a glance:

• Every row has the same page_id (1305) and URL (/trial/start). That is expected: goal_3 is a page-type goal, so every goal firing is a page view of the thanks/start page.
• The rows differ in who got there and how: a mix of organic, direct, social, and paid. That is exactly what makes this material useful — you can break down conversions by any of the allpv-style columns.

Every goal_N material shares the same shape as allpv but is already filtered to the page views that triggered that goal. That means you do not need to JOIN to get source, device, referrer, or engagement columns — they are already there on each row.

Common JOIN keys

• pv_id → click_event.pv_id (what did the user click on the conversion page view?)
• session_id → allpv.session_id (what other pages did they see in the session before converting?)
• page_id → gsc.page_id (did they come from organic search?)

Good first questions for goal_N

• "How many conversions did we get this week, by source?"
• "Which referrers drive the highest-quality conversions (longest browse_sec pre-conversion)?"
• "What is our conversion rate by device type?"
• "Which landing pages lead to the most conversions two steps later?"

Where to go next

• /guide reference — to see the structure of sites[].goals in a live response.
• allpv — the base material goal_N is derived from.