What I Got Wrong at Gorillas

A post-mortem on hyper-growth marketing.

There’s a version of the Gorillas story people love.

A group of ambitious operators builds a rocket ship, raises billions, scales across Europe in record time, and becomes the blueprint for hypergrowth marketing.

Some of that is true. Gorillas scaled insanely fast. The brand became culturally relevant. The app was everywhere. Competitors copied the playbook. Investors kept funding the machine.

But looking back, I realize I misunderstood what was actually happening.

I ran the channels well. I misunderstood the nature of the growth.

And if you’re building in a category that’s currently “hot” (AI, consumer health, fintech, climate, whatever gets subsidized right now), Gorillas is a useful warning.

Because the uncomfortable truth is simple:

A lot of what looked like marketing performance was just venture-funded economics in disguise.

---

The context people forget

Gorillas wasn’t selling a lifestyle brand or a long-term identity. It was selling speed and convenience.

That sounds great until you realize what it implies:

• switching costs are close to zero

• loyalty is fragile

• differentiation is thin

• retention depends on habit

• habit depends on frequency

• frequency depends on incentives

So growth becomes less about persuasion and more about brute-force repetition.

At the time, I thought this was normal early-stage scaling. Now I think it was something else.

---

The biggest misconception: we thought we were building leverage

In hypergrowth companies, marketers chase leverage. The idea that every euro spent compounds through retention, word-of-mouth, and brand memory.

In reality, Gorillas often didn’t build leverage. We bought demand.

That distinction matters because buying demand works incredibly well in the short run and collapses brutally once subsidies stop.

---

What actually worked (and what it really did)

1. Promotions were the engine

Promotions weren’t a tactic. They were the business model.

Promo codes were everywhere. Influencers, CRM, referrals, paid ads, flyers, partnerships. Sometimes as extreme as €10 off your first 10 orders.

It worked — driving acquisition, reactivation, and order frequency — but it also created the core problem:

Users learned to treat grocery delivery like a bidding war.

If Flink paid more, they switched. If Getir paid more, they switched again. If Gorillas came back with a better deal, they returned.

That isn’t loyalty. It’s arbitrage.

Once you train users to expect subsidies, you don’t “optimize discounts away.” You go through withdrawal. And in a category with low differentiation and low switching cost, withdrawal is brutal.

---

2. CRM was the closest thing to a real growth machine

If you forced me to pick one area where marketing created compounding value, it was CRM.

Email, push, and in-app messaging were insanely powerful because they were cheap, direct, fast to test, and fully controllable.

The best teams weren’t “creative.” They were systematic. They built lifecycle flows, segmentation, and reactivation loops. If you knew what you were doing, CRM printed money.

But the uncomfortable detail is obvious in hindsight: CRM didn’t work because messaging was magical. It worked because it distributed incentives efficiently. A great push notification without a discount rarely solved retention.

---

3. Referrals scaled, but mostly as incentive distribution

Referrals worked extremely well: “Give €10, get €10.”

It was low friction, easy to understand, and matched the product’s behavior perfectly. But referrals didn’t create community-led loyalty. They scaled incentives through users.

Again — growth, yes, but never durable loyalty.

---

4. Paid social drove installs, but not economics

In the Facebook + Instagram era, you could buy volume. Installs worked. First orders worked. Even repeat orders worked for a while.

But CAC was always uncomfortable because LTV was structurally capped by low margins, high fulfillment costs, low switching costs, and promo-driven behavior.

Paid social wasn’t a compounding engine. It was a volume dial. And volume dials are expensive.

At the time, I thought we just needed better creatives and better targeting. Today, I think the category was the problem. You can’t out-optimize a broken equation.

---

5. Influencers gave reach, but weak brand value

Influencers looked good on dashboards: awareness, clicks, installs, promo code redemptions. They were also a perfect way to push discounts without being penalized by social algorithms.

But influencer marketing didn’t build brand in the way people romanticize. It built short-term spikes correlated with the offer. In the free-money era, pricing inflated massively. Many “ROI-positive” deals were only positive because the discounts were insane.

---

6. Trade marketing was a hidden engine because it was economics

Trade marketing was one of the most underrated levers. Suppliers co-funded promotions, paid for placement, and subsidized visibility. That allowed Gorillas to spend aggressively while protecting gross margin.

If you want to understand why q-commerce survived as long as it did, look at trade budgets.

---

7. OOH worked until it became an arms race

OOH was the signature move. City domination. Bold creative. High frequency. It worked because it made Gorillas feel unavoidable — perception turned into reality.

But the advantage didn’t last. As soon as every competitor copied the playbook, OOH became table stakes. The channel turned into an arms race instead of an edge.

---

The real post-mortem: we confused subsidized demand with brand-driven growth

This is the hard lesson:

A lot of Gorillas’ “growth” wasn’t marketing performance. It was artificially cheap groceries. VCs paid for customer acquisition by literally paying for the customer’s basket. That is exactly why the model scaled as fast as it collapsed.

The market didn’t shrink because marketers got worse. It shrank because the equation stopped being artificially supported.

---

What I would do differently today

If I could go back, I wouldn’t start by changing channels. I would change the questions.

1. Focus less on CAC, more on switching cost

The real question was never “how cheap can we acquire a customer?” It was “how do we make leaving painful?” That could have meant memberships earlier, subscription mechanics beyond discounts, exclusive product selection, or personalization that actually compounds. Instead, we fought for the cheapest basket.

2. Treat promotions like a drug, not a lever

Discounts create dependency. Promotions should have been tightly controlled — used to create habit loops, phased out systematically, tied to behavior rather than volume, and replaced by real value. Instead, we treated them as fuel. Fuel works until the tank is empty — and ours was never as full as the dashboards suggested.

3. Separate “real growth” from “subsidy growth”

The biggest strategic mistake wasn’t running promotions, but failing to measure what share of our growth was real behavior versus artificial pull-forward. The question should have been simple: what happens if we remove the discount tomorrow? If the answer is “everything collapses,” then you’re not scaling a product.

---

The takeaway

Gorillas taught me more about marketing than any job before or after. But the most important lesson wasn’t about channels. It was about incentives.

Marketing is powerful when it amplifies a product that creates its own pull. Marketing becomes dangerous when it masks a business model that cannot stand on its own.

For a while, q-commerce looked like a revolution. In reality, it was a venture-funded experiment in subsidized convenience. When the money stopped, the business had no floor to land on.