Chang Chen26 November 2024

How She Scaled an AI App to 3M Visitors/Mo Using AI SEO ft Chang Chen

2Frameworks

8Insights

Frameworks in this episode

Marketing7 steps

Human-in-the-Loop AI SEO Workflow

Automate repetitive SEO research while humans supply judgment and originality

Read →

Marketing6 steps

Keyword-Intent-Quality Gate

Validate opportunity, page format, and originality before scaling AI content

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster17:00

AI SEO's Advantage Is Throughput, Not Better Rankings

The mixed AI-and-human pages described in the episode perform roughly on par with traditionally produced content. The advantage comes from reducing research and drafting effort, enabling humans to publish more while retaining responsibility for the work that influences quality.

Mixed-production content can perform on par with human-only work
AI does not automatically improve rankings
The primary gain is greater human throughput
Repetitive tasks are the strongest automation candidates

“We do see most of the content we write as performing on-par with the traditional or human owning content”
Chang Chen · 17:00

“It's a, it's a productivity gain, versus, like being able to rank things better.”
Kieran Flanagan · 17:30

#seo performance#productivity#content operations#automation

Hot Take· 2

Hot Take03:00

AI Search Will Shift Content Measurement from Clicks to Influence

The hosts argue that answer engines may satisfy queries without generating substantial referral traffic. Content created for these environments may therefore need to be evaluated through brand influence and impressions rather than the click-based model used for traditional search.

Answer engines can remove the need to click
Referral traffic may remain limited
Brand influence may matter more than direct visits
Traditional click attribution becomes less reliable

“Like, we used to create content for clicks, and now we kind of create content for influence, and I think that's the that's the change…”
Kieran Flanagan · 04:00

#ai search#attribution#brand influence#answer engines

Hot Take20:00

Personal AI Agents Could Replace Manual Customer Segmentation

Chang argues that manually designed segments are imperfect proxies built from experience and aggregate modeling. An AI agent can instead understand and engage each customer individually, creating proactive one-to-one experiences without forcing users into broad predefined groups.

Manual segments may encode inaccurate assumptions
AI can model users at the individual level
One-to-one engagement exceeds broad personalization
Agents can adapt how each customer is approached

“we're actually just skipping that step to rely more on the AI agent to understand the user on the one-on-one level”
Chang Chen · 21:00

“be able to engage customers in the exact way that they want to be engaged.”
Chang Chen · 21:30

#segmentation#personalization#ai agents#customer experience

Explainer· 1

Explainer04:00

Why AI Companions Create an Attribution Blind Spot

Chang explains that a companion trained on an individual's preferences can become a trusted recommendation source even when it supplies no direct link. This creates personalized influence that conventional referral analytics may fail to capture.

Companions learn individual preferences
Personalized recommendations can earn trust
Recommendations may omit direct links
Influence becomes difficult to attribute

#ai companions#personalization#attribution#recommendations

Tool· 1

Tool10:30

Turning One Product Feature into Keyword-Specific Landing Pages

For structured landing pages, Chang's team converts keywords and page-section content into tabular data, with each row representing a page. The resulting CSV can be uploaded to Webflow, allowing one product feature to support multiple keyword-specific pages while retaining a common template.

Represent each page as a table row
Use one structured workflow across many keywords
Upload generated page data through the CMS
Create multiple keyword angles for one feature
Retain human review before publication

“And we use Webflow, so we can just upload the CSV and be able to create landing pages as well”
Chang Chen · 11:00

“for one feature, we actually identified 10 plus different keywords that we can target to create a one template landing page”
Chang Chen · 12:00

#programmatic seo#landing pages#webflow#automation

Takeaway· 3

Takeaway01:30

The Three AI Growth Use Cases Working Now

Chang Chen identifies localization, SEO, and content repurposing as the areas receiving the most attention. She also reports that AI can reduce localization and influencer-campaign costs, while AI-native referral channels remain comparatively small.

Prioritize localization, SEO, and content repurposing
Use AI to accelerate content creation and translation
Treat AI-native referral traffic as an emerging channel
Measure operational savings as well as traffic

“Right now, we are spending a lot time on localization, SEO and the content repurposing”
Chang Chen · 02:30

“Influencer campaigns as well. It helps us save at least 80% of our costs and be able to help to reach more micro influencers with…”
Chang Chen · 02:00

#ai marketing#localization#content repurposing#growth

Takeaway15:00

Five Sources of Originality for AI-Assisted Content

Chang lists practical sources that can distinguish AI-assisted pages from generic search copy. These include editorial perspectives, product messaging, proprietary market research, user-generated material, and insights taken from subject-matter experts.

Capture author and editor perspectives
Integrate product messaging where relevant
Publish proprietary market research
Use reviews and community discussions
Repurpose expert interviews and podcasts

“The unique content, the unique perspective, is still the key.”
Chang Chen · 15:00

“the unique content piece is still something that AI cannot do by itself.”
Chang Chen · 16:30

#information gain#original content#ugc#expert insights

Takeaway19:30

AI Assistants Can Beat Static Product Onboarding

Chang reports conversion gains when shoppers or SaaS users receive help from a personal AI assistant. Unlike a fixed tutorial, the assistant can answer early questions, explain the offering, and guide users according to their immediate needs.

Use assistants to help visitors explore the site
Answer early product questions conversationally
Guide SaaS users through free-product options
Replace rigid rules with responsive assistance

“The AI agent works better than the traditional tutorial - a very standard and traditional rule-based onboarding flow”
Chang Chen · 19:30

#conversion#onboarding#ai assistants#saas