lighthouse logo
lighthouse logo
AI Lab

Blog SEO (+GEO) Assistant

AI assistant that audits, optimises, and enhances new and existing blog posts for SEO and Generative Engine Optimisation (GEO)
New cohort starts in
Days
Hours
Minutes
Get a quote
AI Enablement
Get a quote

The Opportunity

Marketing and communications teams spend hours optimising content manually, if they do it at all. Many published blog posts underperform because they were written without a clear SEO strategy or have since become outdated. Meanwhile, creating new content is expensive and time-consuming. This experiment explores how an AI-powered SEO assistant can analyse both draft and live content to identify optimisation opportunities, re-write sections, and add new openings or conclusions (“topping and tailing”) that align with current search intent, tone, and calls to action. The goal is to increase visibility and engagement while extending the lifespan and ROI of existing content.

Hypothesis

We believe that an AI SEO assistant capable of auditing and improving existing blog posts will increase organic traffic by at least 25%, reduce manual editing time by 40%, and help older content re-enter top search positions.
Tags

Design & Method

This pilot compares performance before and after AI-assisted optimisation of ten published articles and five new drafts.

Process

The step-by-step process that guides the setup, delivery, and measurement of the experiment.
Content Ingestion
Import draft or live article text into the AI workspace for analysis.
SEO Audit
Assess keyword alignment, search intent match, readability, and tone consistency.
Optimisation Recommendations
Generate headline, meta description, and structure suggestions aligned with target keywords.
Topping and Tailing
Add or rewrite introductory and concluding paragraphs to improve relevance and conversion opportunities.
Implementation and Scoring
Apply accepted changes, rescore for SEO and readability, and record deltas.
Performance Tracking
Monitor ranking, click-through rate, and dwell time over a four-week period.

Inputs & Prompts

Examples of user queries used to train and evaluate the AI's understanding.

Blog Text Input

Full article text (draft or published) provided for analysis.

Target Keyword Cluster

Primary and secondary keywords to guide optimisation.

Desired Call-to-Action

Defines the thematic or campaign-specific end goal (e.g. newsletter signup, demo).

Performance Baseline Data

Pre-experiment traffic, rank, and engagement metrics for comparison.

Experiment type
  • Assistant
Maturity
Complexity level
Programme Finder ChatBot
Conversational assistant that helps executives find the right programme faster, increasing engagement and click-throughs.
View experiment
Partner Prospecting Agent
An AI agent that automates employer research to surface high-fit partnership leads by course, region, and CSR alignment.
View experiment
Student Support Agent
An AI-powered triage agent that helps student support teams prioritise, route, and resolve queries faster.
View experiment
Blog SEO (+GEO) Assistant