The NRR Trap Is Real, But Discounts Are Just a Band-Aid—Data Shows the Bigger Opportunity

What if the biggest threat to your SaaS portfolio isn’t the customer who screams about leaving, but the one who quietly downgrades to a cheaper plan without telling anyone? That’s the question I kept coming back to after reading Jason Lemkin’s recent piece on SaaStr about discounting to save renewals. His argument is visceral: when a longtime customer like Marketo faces a price hike with dwindling value, a big discount might be the only thing stopping them from walking. But our data at PainSignal suggests we’re missing an even larger, more insidious problem.

Lemkin frames the issue as a trap: protect NRR at all costs, and you let salvageable customers churn; discount to keep them, and you wound the metric that investors prize. It’s a genuine trade-off. But what his article doesn’t surface is how many other customers are already halfway out the door without anyone noticing. I see it in the 134 problems and 89 app ideas on PainSignal specifically around churn prediction and retention. Founders and operators aren’t just worried about the big renewal conversation—they’re desperate for ways to spot small, cumulative value leaks that erode revenue over time.

Consider the phenomenon of “shadow churn.” Think of a mid-market SaaS tool used by a 50-person team. The champion leaves, usage drops from 50 seats to 20, but the account doesn’t technically churn. The NRR hit is gradual, spread across quarters, and no alarm bells ring. Lemkin’s Marketo example is dramatic and relatable, but our data shows that most revenue loss isn’t coming from the loud, referenceable accounts. It’s coming from the silent downgrades. In fact, PainSignal tracks over 47 problems just in the Marketing Automation category with an average severity score of 3.8 out of 5, many of them centered on pricing-to-value misalignment. That’s not a legacy problem; it’s an industry-wide signal.

Lemkin is right that the discount conversation is broken. He points out that if nobody is incentivized to save a customer, the customer walks. Our data reinforces that: we see a cluster of problems where users of B2B tools complain that their success manager never has the authority to offer a meaningful concession. But here’s the twist that builders and investors should lean into: the solution isn’t just better incentive design or a permission slip for discounting. The solution is data. Specifically, operational data that can predict the need for a discount before the customer ever thinks about asking.

There’s a clear market gap. We have 23 verified problems on PainSignal where SaaS buyers say they have no way to know if they’re paying a fair price. These aren’t just legacy enterprise buyers; they’re indie hackers, agency owners, and SMB founders. At the same time, an app idea for “SaaS Pricing Intelligence” has been gaining traction—it’s an automated benchmarking tool that would let vendors proactively adjust pricing based on real-time market signals. Imagine a customer success platform that ingests not just product usage data, but external signals like crowdsourced pricing complaints and competitive shifts, and then recommends a personalized discount before the renewal call. That’s not a feature, it’s a category creator.

Why does this matter for a seed investor? Because the next wave of retention startups won’t be built around better QBR decks or health scores. They’ll be data infrastructure plays. They’ll ingest signals from places like PainSignal and turn subjective churn risk into an actionable, quantitative pipeline. Lemkin says the real answer is a better product. I agree, but I’d add: the real answer is a product that is constantly aware of its market value and can adapt. The future isn’t a CS rep armed with anecdotes; it’s an AI-native system armed with market intelligence.

This also reframes the NRR trap. Instead of viewing discounting as a defensive move to protect an increasingly fragile metric, what if companies could optimize NRR by eliminating revenue leakage before it shows up in the numbers? Our data suggests that the most painful churn is often preceded by months of subtle signals: feature requests that go ignored, support tickets about pricing, a champion who stops logging in. The problem is that these signals are trapped in isolated systems. A platform that unifies them—from support tickets to market pricing data—would let companies intervene strategically, not reactively.

Lemkin’s article is powerful because it gives permission to break the metric-driven myopia that plagues so many SaaS leaders. Discounts are a tactical tool, and sometimes the right one. But for the builders in our audience, the takeaway should be bigger: the data is telling us that retention is a pipeline problem. And pipelines can be engineered. The companies that turn churn prediction into a data product will not only save their own NRR—they’ll sell the picks and shovels to everyone else.

For investors, the takeaway is equally clear. A startup pitching “better retention through AI” is a yawn. A startup that comes with a proprietary dataset of real-time customer pain signals—like what PainSignal aggregates—and turns that into a retention co-pilot for B2B sales teams? That’s a moat. And it’s not a niche for legacy vendors. It’s for any SaaS company that competes in a crowded market where pricing and value perception shift fast.

So go ahead and discount if the math works. But use the year it buys you to build a data moat, not just a product update. The trap isn’t NRR. The trap is thinking you can manage retention without a real-time, external view of your customers’ pain.