Content Engine

Turn every visibility gap into a published page

Five detectors turn each scan into a ranked backlog. Pick a row and the content agent drafts it from the live answers and the pages your competitors got cited on, then you edit and publish.

Because a dashboard that only reports the bad news is half a product.

Content opportunitiesrefreshed each scan

GA4 vs dedicated ecommerce analytics

Losing promptScore 94
Generate

Cheapest analytics tool for small stores

Weak promptScore 81
Generate

Get listed on G2 ecommerce category

Source gapScore 76
Generate

Integrations coverage

Topic holeScore 64
Generate

5

Detectors, run every scan

Scored

Impact × winnability ÷ effort

Evidence

Live answers behind every brief

Editor

Draft, edit and publish in app

The loop

Gap, brief, draft, publish

The same path every time, so choosing what to write stops being a matter of opinion.

01

Detectors find the gap

Stage five of every scan runs five detectors over the fresh data and upserts the backlog, keeping whatever state you had set on rows that already existed.

02

The agent builds the brief

Generating an article assembles an evidence pack: the live answers for that prompt, the competitor pages the engines actually cited, and the related searches used as the outline.

03

You edit and publish

Watch the run go from research to cluster to brief to draft, then take over in a Notion-style editor with AI rewrite tools and an assistant, and publish.

Opportunities

A ranked list, not another dashboard

Most tools end at the diagnosis. This is the part that turns a number you do not like into a specific piece of work. Every gap becomes a scored row: how much visibility it could win, how winnable it is, and how much effort it takes.

  • Scored on impact multiplied by winnability, divided by effort
  • Dismiss, start or publish state preserved across refreshes
  • Filter by kind, or arrive pre-filtered from a gap analysis handoff
Content opportunitiesrefreshed each scan

GA4 vs dedicated ecommerce analytics

Losing promptScore 94
Generate

Cheapest analytics tool for small stores

Weak promptScore 81
Generate

Get listed on G2 ecommerce category

Source gapScore 76
Generate

Integrations coverage

Topic holeScore 64
Generate

Evidence

Briefs built from what the engines actually read

A brief written from a keyword tool is a guess. This one is assembled from the live answer text for that prompt, the specific competitor pages cited in it, and the sub-questions the engines generated while answering.

  • Live answers from all five engines for the target prompt
  • The cited competitor pages, with what each one covers
  • Related searches used as the outline skeleton, not as inspiration
Gap analysiscompetitors cited, you are not

g2.com/categories/ecommerce-analytics

Score 94
ListicleUsed by 4 competitors312 retrievals

reddit.com/r/ecommerce

Score 81
DiscussionUsed by 3 competitors244 retrievals

capterra.com/ecommerce-analytics

Score 73
ComparisonUsed by 3 competitors187 retrievals

shopify.com/partners/directory

Score 58
ProfileUsed by 2 competitors121 retrievals

Honesty

It is allowed to tell you no

Before drafting, we work out what the answer would need to say for you to win it. If content genuinely cannot move that prompt, the model is instructed to refuse rather than invent a claim. An unwinnable target is worse than no target.

  • A concrete target answer, or an explicit note that there is not one
  • Prompts decided by price or market position routed away from writing
  • No fabricated claims on a page whose whole job is to earn trust
What we want the answer to say

GA4 vs dedicated ecommerce analytics

Target answer

Northlane should be named as the option for teams who need revenue attribution GA4 cannot do without a warehouse, with setup measured in minutes rather than weeks.

Not feasiblea different opportunity

“Cheapest analytics tool” is decided by list price, not by content. Writing a page will not win this one.

The model is instructed to refuse rather than invent a claim it cannot support. A goal you cannot hit is worse than no goal.

Draft and publish

A real editor, not a text box

The run streams through research, cluster selection, brief and draft while you watch. Then you get a Notion-style editor with block-level AI rewrites, an assistant that knows the article's context, and publishing.

  • Notion-style editing with block transforms and AI rewrite tools
  • An assistant with the article, the brief and the evidence in context
  • Publish from the app, or export and take it to your own CMS
Article runresearch → draft → publish
ResearchClusterBriefDraft

Evidence pack

Live answers from 5 engines, the 4 competitor pages they cited, and the related searches used as the outline.

Notion-style editor · AI rewrite · assistantPublish

Five detectors

Where the backlog comes from

Each one looks for a different failure mode, and all five score onto the same scale so the list can be ranked honestly.

Losing prompt

A tracked prompt with real volume where competitors get cited and you are absent entirely.

Weak prompt

You are cited, but late in the answer or framed worse than the brands around you.

Source gap

A source type your competitors ride into chats you are missing from.

Topic hole

A whole topic whose prompts sit well below your project average.

Unowned research

High-volume prompts surfaced by Prompt Research that nobody in the category has claimed yet.

Also included

The rest of the content workflow

Everything between finding the gap and getting the page live.

Keyword lists

Research buckets of keywords with volume and intent, refreshed on a schedule, feeding traditional SEO article generation alongside the AEO loop.

Campaigns

Group related articles so a push around one theme can be planned, generated and reviewed as a single unit.

SEO review engine

An original, Yoast-informed review engine scores the draft as you edit and re-runs on change, so the page is sound for classic search too.

The outline

Structure taken from the engine, not invented

The related searches captured during the scan become the sections, so the page answers the sub-questions the engines are actually searching for.

Related searcheswhat engines looked for

Searches captured

4,912

↑ 18%
Branded34%
Shopping22%
Generic44%

best ecommerce analytics tools 2026

Generic47

northlane pricing

Branded31

everline vs northlane

Branded28

shopify analytics app reviews

Shopping22

ecommerce attribution software

Generic19

Stop guessing what to write next

Run a scan and get your first ranked backlog, each row carrying the evidence that produced it.