Marketing
YouTube and video content in AI answers: from transcript to citation
Industry studies show Google's AI surfaces cite YouTube heavily. What engines actually read in a video, and how captions, chapters, transcripts and VideoObject markup make it citable.
Getting into AI's "top 5": winning best-of and recommendation queries
Best-of prompts work differently from head-to-head comparisons. We examine how engines assemble shortlists, what the evidence says about self-promotional listicles, and how to measure inclusion.
Local GEO: how AI assistants choose the "best" local business
AI assistants build local recommendations differently from map rankings. We set out what Google, ChatGPT, Apple and Bing document about local data and which signals a business can manage.
Original data and research: creating the content AI chooses to cite
What the GEO study really found about statistics, quotations and citations, where its limits lie, and a rigorous method for producing original data and research that AI engines can cite.
Source gap analysis: the pages AI cites instead of you
A five-step method for finding the pages AI engines cite instead of yours: a frozen prompt set, URL classification by owner and type, a gap matrix and the legitimate response for each source type.
Automatic budget allocation: ads that tune themselves
When the winning ad set stops at nine on a Saturday night while the losers keep spending, the problem is not decision quality but the frequency of the decision.
Channel mix analysis: which channel should the budget move to?
A working method for reading the gap between spend share and outcome share, costing each channel honestly, and setting budget-shift thresholds before the meeting.
Audience builder: turning CRM data into ad audiences
A practical walkthrough of what to clean, what to check on consent and which audiences to exclude before you push a CRM segment out to an ad platform.
From ad creative to A/B test hypotheses
How to turn a finding in a creative report into a sentence you can test: the four-part hypothesis, thresholds written before launch, and the learning a test leaves behind.
Landing page section analysis: which section breaks conversion?
Score the headline, value proposition, button and trust block separately and the break in the chain becomes visible. A method, thresholds and a test queue.
Client-ready SEO reports: white-label output for agencies
A monthly report is a data pipeline problem, not a slide problem. The three layers of white-label, weekly versus monthly cadence, and writing a bad month.
Content health audit: which post should you update?
Not every declining post earns a rewrite. Here is a repeatable scan that reads the signals, ranks the candidates and settles update, merge, redirect or remove.
IndexNow: instant index notifications that actually work
Stop waiting for a crawler to notice your new pages. Key setup, which URLs to submit, what each response code means and how to measure the real effect.
Internal linking opportunities: connecting the posts you already have
Orphan pages, buried posts and titles competing for one query: how to earn visibility by rewiring the archive you already own, without writing anything new.
AI readiness audit: can AI engines actually read your site?
robots.txt, llms.txt, rendering, schema and citability: the checkable items that reveal whether AI engines can actually read the pages you publish, and in what order.
Share of voice in AI search: measuring visibility against rivals
One answer is an observation, not a measurement. Here is how to set the denominator, design the question set, choose a cadence and read the competitive matrix.
Mention or citation? Two different wins in AI answers
Your name in the text and your domain in the source list are separate events; here are the four cells you can be in and the first move each one calls for.
SERP tracking: where did your keyword ranking go?
A position is not a single number. How to build a tracking list, date the drop, tell a competitor's rise from your own decline and keep alerts out of noise.
Competitor website change detection: seeing a price move the same day
A competitor's pricing page moved yesterday and nobody noticed. Here is the setup that closes the gap: a short watch list, sane thresholds, honest limits.
True ROAS from your CRM, not the ad platform
The four reasons platform conversion value and closed-won revenue disagree, how cohort maturity changes the answer, and the tests a rule should survive first.
Rolling out an ad rule in stages: start with 10%
A sound rule can still damage an account when it opens everywhere at once. Here is how to start on ten percent of campaigns and widen scope stage by stage.
The data health gate: never decide on ads with broken data
A stale sync, a partial day and a mismatched currency are enough to mislead a rule. Here is how to build the gate that checks data before any ad decision runs.
One-click rollback in ad automation
A rule paused the wrong campaign overnight. Here is how reversing that decision works, what the 24-hour window covers and which guardrails prevent a repeat.
Conflicting ad rules: when two automations touch one budget
If one rule opens a budget overnight while another cuts it, the defect is in the portfolio, not the rule. Separate scope, metric and window to end the loop.