Should You Discount to Save a Renewal? The NRR Trap Says No. You Probably Should Anyway.

The Discount That Would’ve Kept Us

So we’re going to churn off Adobe Marketo.

The irony: I was one of the first 10 Marketo customers. Ever. I’ve been a customer (with a break) for goodness, the better part of two decades. And a reference account for years. I did the calls. I told other B2B companies why they should buy it. Back then, Marketo was the scrappy challenger taking on Eloqua, and it was genuinely better. I believed it.

And now, finally, we’re leaving.

Part of it is simple: our AI VP of Marketing simply actually use it. Broken unsubscribe handling, daily API rate limits (the biggest issue), a 6-month history cap, general hostility to agents. When I asked three foundation models what they’d use for marketing automation, all three mocked Marketo. The consensus was that an agent will never send an email through Marketo, because it will just write a better one itself.

But even with all that, if they’d offered us a large discount to stay, we’d probably stay. We pay way too much for a 3-person + 21-Agent org.

At least for another year. And then who knows? Maybe another. And another. Another year at least buys you a chance to improve and really re-earn the business. Another year keeps the customer in play.

Instead, they just increased pricing at our renewal. For zero new features, and in fact, an even more limited API.

This is a tough one for legacy and older B2B companies.

Offering big discounts on renewals really does work. Leaving is work. Migrating is work. Retraining the team, rebuilding the workflows, re-integrating everything. A big enough discount makes all that work not worth it.

But discounts can be tough on the org, and it’s hard to know how low to go. Sales reps hate discounting into renewals. Finance hates the ASP hit. Nobody wants to be the one who set the precedent that squeaky customers get 40% off.

And it’s even tougher on the metrics.

We all sort of assume B2B runs at 100%-130% NRR. Every deal that renews lower drags that down. So the instinct is to hold the line on price, protect the number, and let the churn-risk customers walk if they’re going to be difficult.

But that instinct is backwards for a lot of legacy vendors right now.

There are so many customers that could be saved today if expensive legacy vendors just lowered pricing at renewals. Customers who don’t actually want to leave. Customers who are only shopping because the price stopped matching the value.

And a discount doesn’t just save the logo. It buys you another year to actually make your product competitive again. Another year to ship the AI features. Another year to close the gap on the newer, cheaper, faster tools eating your base.

The NRR Trap

Median private B2B NRR has slipped to about 101%, barely above the break-even line. And the market pays enormous premiums for the companies above it. SaaS companies with NRR above 120% trade at roughly a 63% premium to the median. Companies below 100% trade at close to a 46% discount. The relationship isn’t linear either, it’s exponential at the top. So every point of NRR feels like it’s worth defending, because in valuation terms it is.

That pressure is exactly what pushes legacy vendors to hold the line on renewal price. Discounting a renewal drags NRR down this quarter. Letting the difficult account walk, and backfilling the story with expansion elsewhere, keeps the number clean. You want NRR up. Public markets want NRR up. So you protect it.

But look at what you’re actually protecting it with. You’re hitting the number by letting churn-risk customers leave rather than saving them at a lower price. You’re harvesting the base to make a quarter look good.

The customers most likely to walk on a bad renewal are often the ones giving you the most signal about where the product is falling behind. Lose them to protect ASP and you don’t just lose the logo. You lose the year the discount would have bought you to fix the product. You lose the account that would have expanded once you did. The metric looks fine right up until the base is gone and there’s nothing left to expand into.

That’s the trap. You may want NRR up. Public markets may want NRR up. But in a case like this, holding price to protect the number is trading away your future to hit it this quarter. The honest move is to take the near-term NRR hit, keep the customer, use the year, and rebuild the product so the next renewal expands instead of churns. Protecting the metric and protecting the company are not the same thing, and legacy vendors confuse them constantly.

The Only Real Answer Is a Better Product, And Building a Winning Agentic Product

The discount is a stall, not a fix. The real answer is a better product.

If you save a customer, at least you bought yourself a year to make the product more competitive. Use the year. Don’t just bank the renewal and move on to the next fire.

But if nothing else, use AI to actually read all your customer signals. You have more of them than you think: usage trending down, support tickets piling up, logins dropping, the champion who stopped showing up to the QBR. And the ones who tell you repeatedly they’re going to leave, or whose signals say the same thing, probably aren’t playing games. They mean it.

So ask the real question: is it better to keep them? Sometimes the answer is no, and you let them go. But you should be making that call on the signals, not missing it because no one was reading them.

Understand the Incentives

And this is the part most legacy companies get wrong. Understand the incentives.

If there’s no incentive to save these customers, they won’t be saved. It’s that simple. You can build the churn model, light up the dashboard, run the QBRs. If nobody actually owns the save and gets paid on it, the customer walks.

Moving your entire customer success team into Sales often means no one will save them. Now everyone’s comped on new logos and expansion, and the quiet work of pulling a wavering customer back from the edge belongs to no one. The renewal that could’ve been saved with one honest pricing conversation just… isn’t, because no one’s job was to have it.

What a 14-Year Customer Walking Should Tell You

Marketo could keep us. The math to keep us is obvious. They’re just not doing it.

And I’m not a hard customer to keep. I’m the guy who used to sell Marketo for them, for free, for years. If a founding-era reference account is walking and no one at the company has picked up the phone with a real offer, that’s not a me problem. That’s a signal about how the whole base is being managed.

For legacy B2B: the customer telling you they’re leaving is often the cheapest customer you’ll ever have a chance to save. The discount feels expensive. Losing them, and everyone like them, costs far more. But the discount only matters if someone is incentivized to offer it, and if you use the year it buys to build something worth staying for.