Both 'Building' and 'Operating' Are Being Replaced by AI. What Work Remains for D2C Marketers?

Introduction | What is happening quietly, yet surely

In the two fields of web production and ad operations, the same thing is now happening rapidly and simultaneously.

The 'work of experts' is being replaced by AI.

The process that used to take weeks for LP production is now completed in minutes from a prompt input. AI autonomously handles keyword selection, bidding, and creative optimization for ads 24 hours a day.

This is not a story about 'AI surpassing humans.' It is a story about 'the areas where AI surpasses humans in terms of volume, speed, and cost becoming clear.'

In this article, I will structurally organize what AI replaces and what remains for humans in LP production and ad operations.

1. The structure where 'building' is replaced by AI

The current state of LP production

LP production used to be a project that designers, engineers, and directors spent weeks on.

It is different now.

AI website builders output layouts, copy, page structures, and SEO settings all at once just by entering a business description. Tasks that previously required a three-person team and several weeks can now be replaced by a few minutes of operation by a non-expert.

The difference in cost comparison is also obvious.

Furthermore, there is data showing that 78% of users in blind tests cannot distinguish between AI-generated sites and professionally produced sites.

In other words, 'visual quality' as a form of production value has already begun to commoditize.

The scope of 'building' that AI replaces

• Generation of layouts and designs

• Basic copywriting

• Responsive design and display speed optimization

• Mass production of A/B test variations

• Routine content updates

The task of 'creating the form' is an area where AI excels.

(2) The structure where 'operating' is replaced by AI

The current state of ad operations

In Google Ads, the following are already automated by AI.

Bid management Smart Bidding (tCPA / tROAS) has become standardized, and manual bid adjustment has become a minority practice.

Delivery optimization Performance Max (P-MAX) automatically delivers ads across Search, Display, YouTube, and Shopping using AI. AI determines both channel selection and budget allocation.

Search query matching In September 2025, Dynamic Search Ads (DSA) will transition to AI Max. AI reads the content of landing pages and real-time search intent to automatically match them to the optimal query.
According to Google's official data, enabling all features results in an average 7% increase in conversions while maintaining the same level of CPA and ROAS.

Creative generation In responsive ads, AI optimally combines multiple headlines and descriptions. With AI Max, AI automatically generates the text itself.

The scope of 'operating' that AI replaces

• Keyword selection and negative keyword management

• Bid price adjustment and time-of-day optimization

• Execution of ad copy A/B tests

• Numerical aggregation and standard formatting of reports

• Budget allocation between channels

The operation of 'manipulating the management screen' is an area where AI excels.

(3) So, what remains for humans?

This is the essential question.

When the meansof designing the purposethat humans should take on when 'creating (production) and operating' are replaced by AI.

Remaining Work ① | Interpretation of 'Why'

AI is good at showing 'what is happening.' CVR dropped, bounce rate increased, CPA worsened—AI detects these facts instantly.

However, interpreting 'why it happened' is a different story.

• Why is there a drop-off at the first view of this brand?

• Why does this appeal not resonate with this target segment?

• Why is the ad click-through rate high but the CVR low?

These can only be answered by comprehensively understanding the brand's context, competitive environment, and target psychology. It is the job of 'interpreting' data, not just 'reading' it.

Remaining Work ② | Brand-Specific Structural Design

AI is good at creating 'generally good things,' but it cannot design 'appeal points unique to this brand.'

In terms of LP structure design, it means this:

• Which appeal axis to place at the top (differentiation point from competitors)

• In what order to unfold information (designing the reader's psychological flow)

• What words to use to create the CTA (verbalizing the brand promise)

These are tasks designed by humans who deeply know the brand's products, customers, and competitors. Even if AI can output an 'averagely good structure,' it cannot yet derive a 'structure where this brand can win.'

Remaining Work ③ | Decision-making for the Entire Funnel

Ads and LPs cannot be thought of separately.

• When inflow increases with AI Max, which LP should be used for guidance?

• When CVR is low, should the ad appeal axis be changed, or should the LP structure be changed?

• To maximize LTV, which customer segment should be prioritized for acquisition through ads?

These are judgments that span Acquire (customer acquisition) and Convert (conversion). While AI optimizes each phase, the overall design spanning these phases is done by humans.

④ What 'commoditization of means' implies

There is an important paradox.

The more AI replaces 'creation and operation,' the more the value of 'designing and judging' increases.

When the man-hours for LP production approach zero, the remaining difference is the design ability of 'what to create.' When the automation of ad operations progresses, the remaining difference is the strategic design of 'what to aim for.'

In a market where means have become commoditized, 'guidance toward a goal' becomes the most scarce and highly valued asset.

This is not a threat, but a structural change.

To those who differentiated themselves with the skills of 'being able to create and operate,' it appears as a threat. To those who defined their value along the axis of 'being able to design and judge,' it acts as a tailwind.

Summary | The role of D2C marketers in the AI era

'Creating and operating' is handed over to AI. 'Designing, interpreting, and judging' is kept by humans.

Whether or not one can consciously design this division of labor will be the turning point for D2C marketers from now on.

Why not start with a 'diagnosis' first?

For those who want to request LP structure design and improvement but feel that making a large investment right away is high risk—

NextSTUDIO is currently accepting D2C brand LP structure diagnoses at a monitor price on Coconala.

'I'm spending money on ads, but the CVR isn't going up.' 'I don't know where the problem is with my LP.'

We will identify the structural problems of that LP and deliver a report with improvement priorities and estimated impact simulations.

Currently accepting for only 3 companies at a monitor price.

→ Click here for requests on coconala

I will diagnose your D2C-specialized LP structure and CVR. Identify the structural causes of why your advertising spend isn't turning into profit | coconala

For those who need practical tools for structural design

I sell LP diagnostic checklists, funnel design worksheets, and market analysis reports. If you would like to try diagnosing and designing it yourself first, please make use of these.

→ NextSTUDIO Content Store

About NextSTUDIO

I specialize in LP structural design, CVR improvement, and funnel design for D2C brands.

I take the time to decipher the unique structure of your brand.

I am the only one in charge—that is precisely why I limit the number of projects and dive deep into each company. While large support firms handle multiple companies in parallel, I promise to deeply decipher the unique structure of your brand.