Every cross-border seller needs a sandbox for failure
You have run the playbook: push price up 12%, watch conversion drop 20%. Ship 5,000 units without final QA, watch your listing fill with one-star reviews. Watch a competitor clone your bestseller, undercut by 40%, and run it at a loss for two months until you blink. That gut-punch of a decision–outcome loop is what makes or breaks a cross-border operation. Yet most of us learn it the expensive way—on live listings with real inventory on the water. That is why a free, browser-based simulation game called San Fran Sim caught my attention. Created by Daniel Kempe, it models the exact cause-and-effect tradeoffs that e-commerce operators face every day: pricing elasticity, quality defects, churn, competitor dynamics, and running out of money. It is not a business school case study and it is not a spreadsheet. It is a playable mental model, and I think it can make you a better seller—without burning a single dollar of ad spend.
The simulation that mirrors your margins
San Fran Sim is deliberately set in a satirical SaaS startup world—think “Goggle” and “Chirper”—but the underlying mechanics map directly to e-commerce metrics. The game’s core loop: you decide how much to charge for your product (or service), how much to invest in quality assurance, how aggressively to spend on growth, and when to cut costs. Each decision has a real consequence. Raise prices and conversion drops. Ship without QA and bugs breed, which feeds into churn. A rival can undercut you to “free, for now.” The maker describes it as “Theme Hospital meets RollerCoaster Tycoon,” and the comparison holds because every lever has a hidden second-order effect.
The kicker is the comment from Kempe when asked about failure modes: “running out money and youre dead”. That is the most honest description of e-commerce I have read all year. Whether you are on Amazon Seller Central or running a Shopify store, the single biggest risk is not a bad product—it is a cash crunch. You buy 10,000 units of a new SKU, you launch with aggressive PPC, conversion underperforms, and now you have a container sitting in a warehouse with storage fees ticking. The game forces you to feel that pressure in a loop that can take hours or end in minutes when you hit zero.
What makes San Fran Sim different from a spreadsheet model is that the metrics are dynamic. Players asked whether churn and LTV are calculated based on decisions or follow fixed formulas. Kempe’s answer: “yes based on decisions, timing, competitors, all sorts”. In the e-commerce world, that is the difference between a static P&L and a living one. Your excess inventory costs change when a competitor drops price. Your return rate spikes if you switch factories. Your ad cost per acquisition jumps when seasonal demand shifts. The game models this fluid, interconnected system rather than a linear cause-effect.
Why this game beats the usual business sim
There are plenty of business simulation games out there—Kariosoft’s mobile titles, Capsim for classrooms, even the old “Entrepreneur” board game. What sets San Fran Sim apart for the cross-border operator is its focus on operational tradeoffs rather than high-level strategy. Most sims let you build a factory and set a price and watch revenue go up. This one forces you to fight churn, deal with competitors who undercut to zero, and manage internal morale (“restocking the kitchen… affects morale and therefore how much work gets done”). That kitchen restocking is a proxy for the hundred small decisions that eat your margin: supplier relationships, employee training, customer support response time.
Compare that to the real tools we use. Helium 10 and Jungle Scout give you data on competitors, keyword volume, and estimated sales. They do not simulate the dynamic feedback loop of raising prices and watching your buy box share erode. Klaviyo can automate flows, but it won’t show you the second-order effect of a discount code on your full-price conversion rate. San Fran Sim, for all its cartoonish parody, creates that feedback instantly. And because it runs in the browser with no account required and is free, you can play it on your lunch break.
The dry announcer commentary is another differentiator. One commenter called it “the dry announcer narrating your burn rate decisions.” In e-commerce, we get that feedback from numbers on a dashboard—not exactly thrilling. The game anthropomorphizes your metrics, making the cost of a bad decision feel personal rather than abstract. That emotional hook is why a simulation can teach something a spreadsheet cannot.
What cross-border operators can steal from a startup simulator
You do not need to play the game to learn from its structure. But if you do, here are three mental models you can export directly into your daily operations.
1. Pricing experiments without the risk. The game’s pricing slider is immediate. Raise by 10% and you see the impact on conversion in the same turn. In reality, you wait days or weeks for enough data to know whether a price change worked. Use the game to internalize the shape of the demand curve: small price increases often have outsized conversion drops on price-sensitive categories (think low-cost impulse buys on Etsy or TikTok Shop). Conversely, on B2B-heavy eBay or Amazon B2B, demand may be stickier. The game rewards you for understanding product-specific elasticity before you touch a real price.
2. Quality investment vs. return rate tradeoff. The decision to “ship without QA” in the game mirrors the real choice between cheap manufacturing with higher defect rates and premium manufacturing with longer lead times. Cross-border sellers on Temu and SHEIN are already in a race to the bottom on cost, but the hidden cost is return rates and customer service overhead. The game makes that tradeoff visceral: you see bugs multiply, churn rise, and eventually you run out of money. For Amazon sellers, that translates directly to the “refund rate” metric in Seller Central that eats into your profitability—and can trigger listing suppression. Play the game, and you instinctively start building a buffer for quality.
3. Competitor dynamics and “free for now.” The game allows a rival to undercut you to zero as a pricing strategy. In e-commerce, this is the Temu playbook: take massive losses to steal market share, then raise prices later. The game shows you that the optimal response is rarely to match their price, because they have deeper pockets or a different cost structure. Instead, you need to differentiate (better listing, better reviews, better packaging) or find a niche where they cannot follow. That insight is worth a hundred hours of MBA case studies.
Why Amazon sellers should care more than Shopify ones
Amazon sellers operate under a set of constraints that mirror the game’s mechanics more closely than a typical direct-to-consumer Shopify store. On Amazon, you have a dozen levers you can pull simultaneously: pricing, PPC budget, coupon, lightning deal, inventory levels, FBA vs. FBM, product image, A+ content, review velocity, and competitive positioning. Each lever interacts with the others in non-linear ways. The game’s simulation of “bugs breeding” when you skip QA is exactly the same as shipping a poor-quality product on Amazon—negative reviews multiply, buy box placement drops, PPC cost rises. Amazon’s algorithm is itself a simulation, and San Fran Sim trains your intuition for it.
Shopify merchants, by contrast, control more of the customer experience but face fewer dynamic competitive levers. Your main tradeoffs are pricing, ad spend, and email flows. The game still useful—especially for learning churn dynamics and cash flow—but the density of interdependencies is lower. For Amazon sellers, the game is a condensed training ground for managing the chaos of the marketplace.
Where the math breaks
For all its usefulness, San Fran Sim has blind spots that a cross-border operator must account for. The game is built on a SaaS cost structure: zero variable cost of goods, zero shipping, zero returns logistics, zero customs and duties. In e-commerce, your unit economics are dominated by COGS, freight, storage, and returns processing. The game lets you “restock the kitchen” as a morale expense; it does not model a 40-foot container stuck at Long Beach or a tariff war changing your cost base overnight.
It also lacks any advertising platform simulation. In the game, your “growth” is a monolithic slider. In reality, a seller on Amazon manages PPC campaigns by keyword, adjusts bids by time of day, and analyzes impression share versus competitors. A TikTok Shop operator optimizes for video virality, not a simple budget line. The game’s abstraction of growth to a single lever is too coarse to train you for the nuance of ad optimization.
Another gap: inventory management and lead times. The game’s model runs on a monthly or quarterly tick. E-commerce inventory cycles span 60–90 days from order to shelf. The game does not force you to place orders 90 days in advance, predict demand, or handle dead stock. For anyone who has experienced a $50,000 clearance write-down, this is a critical missing dimension.
Finally, the game assumes a single product or service. Most cross-border sellers run a portfolio of SKUs with cross-sell dynamics. Raising prices on one product affects demand for a related product. The game’s competitors also focus on one product—unlike the real world where a competitor might attack your entire category with a single loss leader.
What I’d watch / test next
I am not suggesting you replace your weekly P&L review with a video game. But I do have three concrete experiments for any operator reading this.
1. Run a team lunch-and-learn with San Fran Sim. Have everyone in your operations team—from sourcing to ads to logistics—play one round together. After the game ends (when you run out of money or get acquired), discuss which decisions each person would have made differently. The goal is not to win the game; it is to surface assumptions about cause and effect that you all carry. You will be surprised how often your inventory planner and your PPC manager disagree on the price elasticity of a product.
2. Build a custom spreadsheet that mirrors the game’s logic for your real product. Take your actual cost, price, return rate, and churn (repeat purchase rate). Create a simple interactive model where you can slide pricing, defect rate, and ad spend. Use the game’s mental framework to decide which levers to pull first. The spreadsheets exist in tools like Fathom or even Google Sheets, but the discipline of mapping the game’s mechanics to your real numbers is where the learning lives.
3. Use the seed system to run “challenge” scenarios with other sellers. The game allows you to share a seed so others can play the same starting conditions. That is a perfect format for a slack group or a mastermind. Each week, share a seed and see who can survive the longest with the highest profit. Then debrief the strategies. This turns abstract business theory into a repeatable, competitive learning loop.
San Fran Sim is not a replacement for Helium 10 or SellerSprite—it is a prerequisite. Before you touch real data, you should train your gut on a system where mistakes cost nothing but your ego. The cross-border sellers who thrive are the ones who treat decision-making as a skill, not a byproduct of experience. This game gives you cheap reps. Use them.
