Share
From a recent live webinar conversation
AI pricing is evolving faster than most teams can spell artificial intelligence. This conversation captures how one of the fastest-growing AI companies is actively thinking through pricing, packaging, and monetization in real time.
Below is a lightly edited Q&A with Elena Verna, Head of Growth at Lovable, pulled directly from the live discussion during a recent webinar. It covers who should own pricing, how to run pricing experiments, how to think about retention in a nascent AI business, and what the next frontier of AI monetization might look like.
Q: Who owns pricing at Lovable, and how does that work organizationally?
At Lovable, growth sits inside product, so pricing and packaging are deeply intertwined with how the product is built and shipped. While there’s a single owner responsible for driving decisions—Elena and her team—pricing is not a solo function.
“Growth function for us sits with product. So it’s very intertwined.” —Elena Verna, Head of Growth at Lovable
Hear Elena’s live answer
When new capabilities are developed, like Lovable’s backend and cloud offerings, the teams building those features also propose how they should be packaged and priced. Those proposals come from the bottom up, grounded in usage and cost realities, and then get folded into the broader system.
As the company has grown, Lovable has also introduced a monetization council that includes product, engineering, marketing, and sales leadership. Elena’s team ultimately owns pricing, but bigger model changes run through that group to ensure alignment, since pricing decisions ripple across the entire company.
“This is a team effort… I just try to basically herd the cats to run in the same direction.” —Elena Verna, Head of Growth at Lovable
The key, Elena emphasized, is having one clear owner, paired with broad input. Elena owns pricing, but she works closely across leadership and stakeholder teams to drive monetization strategy and decisions. Pricing can’t be a committee decision, but it also can’t live in a silo.
Q: How do you actually test pricing and packaging changes? Who do you test on, and what do you measure?
All pricing experiments at Lovable start with new users. When existing customers already have an established mental model of pricing, that makes it harder to isolate the effect of a change. New users provide a cleaner signal.
“We always start with new users, because existing users already have a set perception of what your pricing is.” —Elena Verna, Head of Growth at Lovable
Importantly, these aren’t “painted door” tests. Lovable builds full pricing experiences and runs real A/B tests through live checkout flows. Experiments can be global or region-specific—for example, testing lower entry prices in markets where U.S. pricing doesn’t align with local purchasing power.
“We don’t do painted door tests. We actually create full experiences.” —Elena Verna, Head of Growth at Lovable
Before launching a test, the team defines sample size and acceptable error up front. Once enough users have been assigned, they stop new assignments and then observe how cohorts behave over time.
The decision isn’t made based on conversion alone. Pricing changes are evaluated across:
- free-to-paid conversion
- retention
- pay-to-play expansion
- overall product engagement
Pricing is treated as a lever that affects the entire growth system, not just the purchase event.
Q: How do you think about long-term retention when the company itself is so young?
Lovable only introduced annual plans a few months ago, and those are intentionally de-emphasized. Monthly is the default, because it lowers the barrier to entry and increases initial conversion. Annual plans can improve retention, but they also require much tighter control over utilization, which the team isn’t ready to optimize for yet.
“We’re not trying to optimize for hype. We’re really trying to optimize for value and overall retention.” —Elena Verna, Head of Growth at Lovable
More broadly, retention at Lovable is anchored to one core question: Are the apps people build actually successful?
“Yes, we make many millions of dollars in revenue, but I still think that we’re on what I call the product–market fit treadmill.” —Elena Verna, Head of Growth at Lovable
Rather than optimizing for hype or short-term usage spikes, the team focuses on whether customers continue building, improving, and expanding their use of the product. Elena pushed back on the narrative that AI products are inherently “leaky,” noting that Lovable’s retention is on par with, or better than, well-known consumer and prosumer benchmarks.
The north star is sustained value creation, and not revenue velocity.
Q: How do credit rollovers work, and how does self-serve convert into enterprise?
Credit rollovers at Lovable are intentionally limited. Users can roll unused credits forward one month, effectively doubling their available balance, but credits don’t accumulate indefinitely. This prevents runaway liability while still accommodating uneven building patterns.
“We introduced credit rollovers, which actually did not impact our conversion or pay-to-play upgrades. In fact, it just improved retention.”
Rollovers also function as a retention mechanic. If a user churns, they lose their rollover balance. Staying subscribed preserves accumulated value, giving customers a reason to remain engaged even during quieter periods.
On the enterprise side, Lovable is still experimenting. Early enterprise pricing leaned toward per-user models because that’s what buyers expected, but that approach limited internal adoption. The team is now exploring models that combine platform fees with usage-based pricing to unlock company-wide access while preserving predictability.
“All of our enterprise deals today are closed on top of self-serve.” —Elena Verna, Head of Growth at Lovable
One thing is already clear at Lovable: every enterprise deal starts from self-serve. Usage, validation, and expansion all happen before sales enters the picture. In this environment, enterprise isn’t a parallel motion; it’s an extension of product-led growth.
Q: How do you manage customer frustration around credits and unpredictability?
Credits introduce real complexity. Each action can consume a different number of credits depending on compute intensity, and planning or chat modes also draw from the same balance. That makes it hard for users to predict exactly how much something will cost.
Lovable absorbs certain costs entirely (security fixes, failed builds, detected bugs) so users aren’t penalized for things that go wrong. But the broader challenge remains: customers don’t always know how far their credits will go.
“There’s unpredictability in credits. It’s like, ‘Oh my gosh, I spent this many credits—where did they go?’” —Elena Verna, Head of Growth at Lovable
Interestingly, perception varies widely. Users who’ve worked with engineering teams tend to see Lovable as dramatically cheaper than competitors. Nontechnical users sometimes struggle with the idea that building anything at all costs money.
Internally, the team is actively exploring ways to reduce the cognitive burden of cost calculation for users, including whether credits should evolve into a more straightforward wallet model over time.
Q: How should early-stage companies think about freemium without subsidizing an unsustainable user base?
Elena argued that the starting point is understanding your alternatives. Freemium reduces friction and increases perceived value, but if you don’t invest there, you’ll invest elsewhere—often in paid acquisition.
“If you’re not going to do freemium, then you’re going to have to spend a lot more on your marketing and sales.” —Elena Verna, Head of Growth at Lovable
The question isn’t whether freemium is expensive. It’s whether it’s more effective than the next-best channel.
“People on freemium usually have the highest NPS and talk about your product the loudest.” —Elena Verna, Head of Growth at Lovable
In many products, free users are also your best advocates. They drive word of mouth, create content, and introduce new users—often more credibly than any marketing campaign. If that loop exists, freemium can be a powerful and efficient growth lever. If it doesn’t, it may not be worth the cost.
Q: Looking ahead 12–24 months, where do you think AI pricing is going?
One idea was key here: today’s AI pricing models are temporary.
“We’re very immature in how we’re pricing AI—and that’s because the costs are so high.” —Elena Verna, Head of Growth at Lovable
High costs have pushed many companies toward pass-through or credit-based pricing. As underlying model costs fall, entirely new approaches may become viable, whether that’s outcome-based pricing, internal “currencies,” or a return to simpler per-user models.
“Whoever will be on the frontier in that evolution is going to be the winner in the market.” —Elena Verna, Head of Growth at Lovable
What will separate winners from losers is whether they’ve built the internal capability to change pricing quickly, safely, and transparently as the economics evolve. In the AI era, having the ‘right’ pricing is moot, and being ready for the next pricing iteration is what’s going to keep companies ahead of the pack.
.png)
.png){kind=avatar}
