How I built a new composite qualitative metric - The Lovable Score

Why did no one tell me this was a thing??

Growth teams love metrics they can screenshot in a dashboard.

Usage.
Conversion.
Retention.
Revenue.

A much less comfortable question: do users actually give a shit?

Because a lot of “healthy” products are really just products people tolerate:
their boss told them to use it, they have not found a better alternative yet, or switching feels too annoying.

None of that means they love it.

And the second a real alternative shows up, they’re gone. There goes your retention.

If you want durable growth, the distinction between “I have to use it” and “I love using it” matters a lot.

---

Hey, quick heads-up! I’m going to be doing something new with my sponsors - specifically focusing on sharing products that I use myself or personally recommend to others. More on this soon, but this week I’m excited to share about one of my faves: Passionfroot.

Passionfroot helps b2b marketing/growth teams work with creators (like me!) And they just released their new AI agent, Zest, which can run your entire creator-led GTM from just a prompt. You get:

• A full campaign strategy (based on your ICP, channels, content format preferences & budget)

• Suggested channel mix (across LinkedIn, YouTube, X, Insta, TikTok)

• A list of the specific creators you should work with (even if they’re not on Passionfroot)

• Personalized outreach drafts

• Full campaign management & reporting

They really do it all for you. If you’re planning a big launch, announcement, or other campaign and want to amplify it with creator partnerships - I’d definitely recommend trying Passionfroot and Zest. Learn more, here.

---

As I’ve been talking more and more about the importance of customer trust and emotional connection in product development/growth (see: Brand is a product job now and Minimum Lovable Product era), I can just see the looks of unease/disgust on the faces of my super quantitative peers. ‘You want us to optimize for… feelings?’

I see it all over: Most really strong Growth orgs struggle with integrating qualitative insights. The same thing that makes a team elite at rigorously tracking, analyzing, and applying quantitative data can make them avoid the squishier parts of non-numerical inputs. In today’s tech landscape, feature differentiation is collapsing. Growth is becoming a trust problem, and the products making the biggest splash are the ones people genuinely love.

So, where do you start? The good news is that there are already a bunch of tried-and-tested qualitative metrics… the bad news is that most of them don’t work very well.

• NPS is widely used, but weak on its own (which Leah Tharin has very subtly pointed out in NPS is an awful, useless Metric for product teams)

• CSAT is useful, but often scattered over some crappy looking survey from Qualtrics that’s like ‘We value your time, please take 10 minutes to tell us’

• CES is strong, but usually too narrow on a feature level

• Sean Ellis PMF is excellent, but strangely underused by later-stage companies (I believe every company tried out the Sean Ellis score after reading that one First Round Review post with the Superhuman founder… but for some reason, stopped doing it entirely after the company got a bit of traction. WHY? Your PMF is not static!)

Is it so surprising that teams are not getting a clear picture of what users actually think and feel?

Help me, Notion

Earlier this year, I wanted to fix that for Lovable.

Our quantitative North Star for 2026 is Daily Active Apps. It’s been incredibly useful. But it doesn’t answer a different question I care about a lot: How do people actually feel about building with Lovable?

That led me to Notion, arguably one of the most beloved tech brands out there. I heard they have a composite metric called LUV: Love + Use + Value (how cool is that acronym!?)

Use and Value come from engagement and revenue metrics. Love comes from CSAT surveys led by user research.

Here’s what Rachel Hepworth (Notion’s ex-CMO) told me:

We introduced LUV score because we had this odd issue where our NPS was low, even though our satisfaction scores were very high. Our theory was that it was hard to get started, so people didn't think their friends and colleagues would get past the hump. That meant they wouldn't necessarily recommend it to others even though they personally loved it. Just leaning on NPS wasn't enough. We needed a combined score to show the complete picture.

That was the unlock for me.

Because a composite score solves the biggest failure modes:

• No qualitative metric at all

• Overindexing on one flawed metric

• Inventing some custom nonsense nobody can benchmark

So I built our own: the Lovable Score!

The Lovable Score

So I built my own. Our new ‘Lovable Score’ combines the following components:

• NPS: How likely are you to recommend Lovable to a friend or colleague?

  This one was obvious, because Word Of Mouth is so important to us. This wouldn’t be as important for other companies that don’t rely on this, but for us, a user’s likelihood to spread the word is crucial. But! I also don’t like NPS because of the lack of actionability - someone saying that they think they’re likely to talk about your product does not guarantee that they actually will. So, we added a bonus question: “Did you actually refer Lovable to a friend or family member in the last 30 days?”

• Sean Ellis PMF score: How would you feel if you could no longer use Lovable?

  If people wouldn’t miss us, we’re at risk.

• CSAT: Overall, how satisfied are you with Lovable?

  Also straightforward, because we want it to be satisfying to use.

• CES: How easy was it to accomplish what you came to Lovable to do?

  Same thing. If people don’t find the product easy to use, there will be problems.

“By your powers combined…”

We already have PostHog running, which has a survey functionality. It just pops in-app 7 days after signup for a customer, if they’re still active.

Then we standardize everything to a 100-point scale and weight it based on what matters most to us. For Lovable, that means overindexing on NPS and PMF. Here are our weights:

• 35% NPS (advocacy intent)

• 25% PMF (would miss it)

• 20% CSAT (overall satisfaction)

• 20% CES (ease)

Your mix may be different. That’s the point.

But one thing I’ll say: Don’t leave out the Sean Ellis PMF score. Of all of the metrics we included, that’s actually the one I feel everybody should be tracking all the time. Why? In case you haven’t noticed, every software and AI company is on a PMF treadmill, sprinting just to stay in place. If you’re not continuously seeing that people would go ballistic if they lost access to your product, that’s really bad.

I’ll also challenge teams to raise the bar a bit. Sean described 40% answering ‘extremely disappointed’ as the target for this - that’s the benchmark that’s supposed to objectively define having product-market fit. But I actually don’t think that’s enough anymore. I’d say it should be 50% or even 60%.

Whatever standard you set for yourself, watch this number like a hawk. Especially in the face of the SaaSpocalypse, you do not want to be the last to know if your PMF is starting to collapse.

The results

Our current numbers are:

• NPS: 60+ (happy about that)

  • Referred within last 30 days: 70%+ (very happy about that)

• PMF score: 60% would be very disappointed without Lovable (very very happy about that)

• CSAT: Very satisfied (from 1-7, scored 7) at 70% (happy about that)

• CES: Very easy (from 1-7, scored 7) at 50% (solid)

That gives us a composite Lovable Score of 80+.

One interesting thing: We’re not trying to move it. Instead, I’m focused on protecting it. As we grow, I don’t want us to lose this. As organizations scale, this stuff always gets worse. More features, more complexity, more crap. So I’m super proud of the standard that we’ve established, but I know that we’ll be fighting gravity to stay at this level. Which is why we’re not trying to optimize things to force this score up… we’re tracking it month-to-month to make sure it doesn’t dip. And letting the team know that we can’t drop the ball on this stuff.

That may be the biggest advantage: reinforcing these expectations within the team culture. It reinforces what matters:

• Are we easy to use?

• Are we creating happy users?

• Are we irreplaceable?

• Do people actually want to talk about us?

And if the score starts to slip, it gives us a much clearer read on why.

Our North Star tells us if the business is moving. Our Lovable Score tells us if users still care.

Quant North Star + Qual Composite Score = BFFs

You’re welcome to use the exact Lovable Score I created as a starting point - but do not leave it there. Adjust the metrics you use and the weights you apply, to fit the specific context of your company. Your qual composite score needs to

• Balance out your quantitative North Star metric

• Lean into your org’s Growth priorities and product superpowers

• Be applied across the entire product experience (not just a single feature or flow), to tell the whole picture.

Now, go get your composite on.

---

Edited by Jonathan Yagel.

---