Marketing in Two Worlds (Claude)
The Old World
Ria opened her inbox on Monday morning to find seventeen promotional emails. She deleted sixteen without reading them. The seventeenth—from a running gear brand she’d bought from twice—offered her 20% off “popular items.” She’d bought trail shoes both times. The email showed road running trainers.
She ignored it.
Later that week, Instagram showed her an ad for the same brand. Different products, same problem: compression socks, when what she actually needed was a replacement headlamp for early morning trail runs. The ad took her to a generic landing page with a search bar. She closed it.
The brand had her purchase history. They had her email engagement data. They knew she opened emails about trail running but ignored road running content. They even knew she bought before dawn, suggesting early-morning runs. Yet every touchpoint felt like they were guessing.
They were.
At the brand’s office, Maya stared at her dashboard. Ria was in the “Active Customer” segment—purchased twice in six months, opened 40% of emails, never clicked an ad. The segment had 8,000 people in it. Maya’s job was to move them to “Loyal Customer” status.
She set up an A/B test: version A offered 15% off, version B offered free shipping. She’d know which worked better in two weeks. She’d been running variants of this test for three years.
Maya knew this approach was broken. She was optimising for open rates and click-through rates, not for what actually mattered: whether Ria would buy again, how much she’d spend over two years, whether she’d become an advocate. But those answers required seeing into the future. All Maya could see was the past.
The New World
Eighteen months later, Maya’s world looked different.
The brand had implemented world models and created digital twins for every customer. Ria’s twin—RiaBrandTwin—wasn’t a database record. It was a simulation of how Ria made decisions about running gear.
The twin knew Ria was a trail runner, but more importantly, it understood why she trail-ran. By analysing the cadence of her purchases (always before local trail race seasons), her email engagement patterns (higher response to endurance and challenge content than speed or aesthetics), and her browsing behaviour (she researched extensively before buying), the twin had developed a model of Ria’s underlying motivation: she ran trails for the achievement, not the aesthetic.
This changed everything.
When Ria’s headlamp was likely approaching failure (the twin simulated usage based on purchase date, typical trail runner training patterns, and equipment lifecycle), Maya didn’t just send an email. First, she tested inside the twin.
She ran fifty simulated interventions: different subject lines, sending times, offer structures, and landing page designs. She tested immediate discount offers versus “here’s what to consider when replacing a headlamp” educational content. She simulated what happened if she sent the email Tuesday morning versus Saturday evening. She tested whether including trail race calendar content increased conversion.
The twin showed her that Ria would respond best to Saturday evening (when she planned her training week), that she’d engage more with “upgrade your night running” content than “20% off,” and that featuring a comparison guide would increase consideration time but dramatically increase purchase confidence and basket size.
Maya sent one email to Ria, but she’d already tested fifty versions.
When Ria opened it Saturday evening, the email felt like it had been written specifically for her. Because it had. The subject line read: “Planning night runs? Here’s what changed in headlamp tech.” The content wasn’t selling; it was teaching. It acknowledged that her current headlamp was probably fine but explained what the new generation offered: better battery life for ultra-distance training, improved beam patterns for technical terrain.
Ria clicked through to a landing page that didn’t look like a shop. It looked like a guide. The twin had predicted she’d want to compare specifications, so the page offered a comparison tool. When she selected “ultra-distance trail training,” it recommended two models and explained why, with links to reviews from runners whose training patterns matched hers.
She bought the premium model. The twin had predicted an 73% probability she would.
But the transformation went deeper. The twin was now simulating Ria’s next twelve months. It predicted she’d enter a 100km trail race in eight months (based on her training progression pattern and local race calendars). It forecasted she’d need new shoes in four months, a hydration vest in six, and nutrition products throughout her training block.
Maya could see this entire journey before Ria took a single step. More importantly, she could test interventions: What if we offered a “100km training bundle” in month three? What if we sent weekly training tips? What if we introduced a premium membership with early access to new gear?
The twin let her run all these scenarios. It showed her that Ria would respond to structured training content but ignore bundle offers (she liked choosing her own gear). It revealed that a membership offer would work, but only if positioned as “early access” not “discounts”—because Ria’s motivation was performance, not savings.
Over the next year, Ria received eleven touchpoints from the brand. Every single one felt relevant. Not because they guessed right, but because they’d tested inside her twin and only sent what worked.
She never felt marketed to. She felt understood.
The Difference
In the old world, Maya optimised campaigns. In the new world, she orchestrated futures.
In the old world, Ria was a segment of 8,000. In the new world, she was a simulation of one.
In the old world, the brand tested on customers. In the new world, they tested on twins.
In the old world, marketing was spray and pray. In the new world, it was simulate and send.
Maya’s metrics transformed: Ria’s lifetime value increased 240%. Marketing cost per pound of revenue decreased 60%. But the real shift was philosophical: marketing stopped being about persuading customers to buy and started being about understanding what customers actually needed, then being there at precisely the right moment.
The twin made the invisible visible. It turned “what might work” into “what will work.” It transformed marketing from an expense optimised for efficiency into an engine optimised for customer lifetime value.
For Ria, the brand stopped being noise and became signal. For Maya, customers stopped being segments and became individuals.
That’s what happens when you can simulate the future before you live it.
