The Peltzman Effect: How Save Flows Can Create the Cancellations They Save

The Peltzman effect says that safety features can increase risky behavior.

Original observation: drivers with seatbelts drove faster because they felt safer. Net safety improvement was less than expected because the risk-taking offset the protection.

In SaaS, the same effect shows up in save flows.

The retention Peltzman effect

Here is how it plays out:

1. You build a good save flow. Customers who click cancel get thoughtful offers.
2. Save rate rises. Great.
3. Some customers learn that clicking cancel triggers offers.
4. They start clicking cancel more casually. Not because they want to leave. Because they want to see what shows up.
5. Cancellation attempts rise. Some fraction of those trigger successful saves. Others result in actual cancellations from customers who might have stayed if they had never clicked.

Save rate looks great in isolation. Net retention may be worse.

The signals that you have a Peltzman problem

Watch for these:

Cancellation attempts are rising even as save rates rise

If your save flow is working, some customers who would have cancelled outright are now staying. Good. But if your total cancellation attempts are also climbing month over month, something else is happening. Customers are clicking cancel more.

Repeat cancellation attempts from the same customer

A customer clicked cancel in month 3, took the discount, then clicked cancel again in month 6 for another discount. This is Peltzman behavior. They have learned the cycle.

Save offers are becoming expected, not surprising

Customer support starts hearing "just tell me what discount I can get if I cancel." Or your NPS follow-ups reveal customers who say "everyone knows to click cancel for a deal."

The best save offers get taken by unprofitable segments

Your generous discount is taken by price-sensitive customers who would have cancelled without it, not by the loyal customers you actually wanted to keep. This is classic adverse selection.

Why great save flows can hurt long-term retention

Three mechanisms:

1. Training a specific customer behavior

Repeated exposure teaches. Customers who use the save flow twice learn to use it every renewal. It becomes part of their annual routine, not an emergency.

2. Adverse selection into discounts

Loyal customers rarely trigger save flows. Price-sensitive customers do. Over time, your discount recipients become a self-selected group of price-sensitive customers who will churn eventually anyway. You are subsidizing the wrong segment.

3. Damaging the pricing anchor

If a meaningful portion of customers pay less than sticker price after triggering saves, sticker price stops being real. New customer acquisition becomes harder because prospects hear "you can get 30% off if you push back."

How to avoid the Peltzman problem

Do not advertise save offers

The customer should be surprised by the offer, not planning for it. If your marketing or help docs mention "we can offer discounts if you consider leaving," you are training the behavior.

Vary offers to prevent gaming

If everyone gets the same 30% off, customers know the price. If offers vary by tenure, plan, and behavior, they cannot game as reliably.

Track cancellation attempt rates, not just save rates

Save rate is a ratio. Attempts are the denominator. If both are rising, you may be creating what you save.

Reserve the best offers for real risk, not routine cancellations

A customer who clicks cancel with declining product usage over 3 months is at real risk. A customer who clicks cancel while still using the product heavily is probably just testing the flow. Different offers for different behaviors.

Consider a firmer "you cancelled, we mean it" experience

Some SaaS companies deliberately make save flows minimally attractive to reduce Peltzman behavior. The tradeoff: lower save rate on legitimate cases, but net retention can be higher because you are not training the game.

The counterintuitive lesson

Do not optimize save flow rate in isolation. Optimize net cancellation impact, which requires tracking both the numerator (saves) and the denominator (attempts) over time.

A save flow that produces 25% save rate on 500 attempts (net save of 125 customers) is worse than a save flow that produces 15% save rate on 300 attempts (net save of 45 customers) if the difference in attempts came from Peltzman behavior. The net is 45 vs 125 saves. But the second case may have fewer real cancellations because customers were not testing the flow.

Related concepts

• How to write a cancellation save offer
• AI cancellation save flows
• FTC Click-to-Cancel rule - the regulatory context

To score whether your save flow is producing the retention you think it is, take the 60-second Churn Health Check.