A New Operating Model for Campaign Planning

I met Jessica Herrin at an e-commerce founders event earlier this year.

Almost every presentation that day made me feel slightly behind. People were building AI tools, vibe-coding, creating agents and connecting systems. I remember thinking that perhaps I had missed an entire season somewhere this year. :)

Then Jessica took the stage.

Her view of AI was ambitious, but grounded. She was not telling every founder or marketer to become a technologist. She was asking a more useful question. What should technology allow us to do better?

Jessica has been building businesses through technology shifts for most of her career. Her first startup became WeddingChannel.com, and she later built Stella & Dot. Today she is founder and CEO of Marklo, a campaign planning and analytics platform for Shopify DTC brands.

Our conversation made me think about something that has been building in marketing for years.

We have transformed the tools we use to execute marketing, but the way we plan campaigns, connect information and learn from what we have done has changed much less.

AI may finally force that operating model to catch up.

Marketing has become fragmented

Marketing teams have more tools, channels and data than they did ten years ago.

That has not necessarily created better marketing.

Shopify made sophisticated e-commerce infrastructure accessible to almost anyone. Klaviyo did something similar for email. Add Meta, Google, TikTok, analytics platforms, calendars, spreadsheets, creative tools and Slack.

Each tool solves something useful. Together, they can create a very fragmented way of working.

Jessica described marketers today as “spread-thin executors,” constantly tab-switching and being asked to create more content in more places, often for the same or lower results.

I have seen versions of the same problem both inside large marketing organizations and now working with smaller founder-led teams.

Planning happens in one place. Execution happens somewhere else. Performance sits across multiple platforms. Historical learning may live in an old presentation, a spreadsheet or simply in the head of someone who was there last year.

The problem is not lack of information.

It is the lack of a shared view connecting what the business is trying to do, what it has already learned and what should happen next.

Campaign planning is the missing layer

This is the problem Jessica is trying to solve with Marklo.

Instead of replacing Shopify, Klaviyo or Meta, Marklo connects information across them and organizes it around campaigns.

That distinction matters.

A marketer rarely thinks about an email in isolation. You think about a launch, a promotion, a product, a customer, the inventory, the creative idea and the timing. The channels are expressions of that larger campaign.

Yet most marketing technology has been built channel by channel.

Jessica’s view is that AI becomes much more useful when it understands the business around the channel. What did we run last year? What sold? What is in stock? What is arriving next month? What kind of promotion does this brand use? What does this brand never do?

Only then does “what should we run next?” become an intelligent question.

A generic AI model can generate endless ideas. Generating ideas is not the hard part.

Knowing which idea makes sense for this brand, this customer, this inventory situation and this moment is.

The team around the campaign will change too

Jessica pointed to what is already happening inside technology companies.

Roles that were once clearly separated across product management, project management, engineering, QA and design are beginning to converge because AI allows one person to operate across a broader range of tasks.

She expects something similar in retail and marketing.

Marketing, merchandising and inventory planning may become less separated as people gain access to more information and more execution capability.

That does not make expertise less important. It changes where expertise creates value.

If AI can pull the data, surface comparable campaigns, suggest ideas and create the first draft, the marketer spends less time assembling information and more time deciding what deserves to happen.

Strategy, taste and discernment become more important as execution becomes easier.

There is also some urgency here. Jessica’s point was that AI-native companies are already being built with these new ways of working from day one. Established companies will eventually compete with them.

So the question is not simply which AI tools to add to the current organization.

It is whether the organization itself should work differently.

Brand founders do not need to become software developers

At one point this year, after watching enough people on LinkedIn building agents and custom tools, I spent a day with Claude trying to understand what exactly I should be vibe-coding.

I eventually concluded that I probably should not be vibe-coding anything.

Jessica’s position is even stronger. She thinks telling product people, brand owners and merchants to build their own technology is fundamentally the wrong use of their time.

Her argument is that AI will increasingly become a utility. Access to it will not be what differentiates a brand.

For retailers, she brought the focus back to three things.

Create a distinctive product for a specific customer with a clear point of view, manufacture, deliver and service it with excellence, and build a real social connection with the customer community.

That is where founders and brand teams should be spending their energy.

Understanding AI matters. Knowing what it can make possible matters. Building every tool yourself does not.

Efficiency needs a purpose

I asked Jessica what we should deliberately not delegate to AI.

Her answer was “the brand magic.”

Her analogy was a dinner party. You can simplify parts of the preparation, but someone still needs to decide what kind of evening they want to create.

Marketing needs the same distinction.

Automate repetitive reporting, information gathering, routine customer-service requests and first drafts.

But do not use efficiency to move further away from the customer.

Use the capacity you create to understand them better, develop stronger products, make better decisions and build experiences people actually value.

That, to me, is the more interesting promise of AI in marketing.

Not doing the same work with fewer people, but redesigning campaign planning so teams have more time for the work that creates value.

As Jessica said during our conversation:

AI should enable the best experience your customer has ever had.