How'd you fix Bubble's revenue issues in 2026?

The two paths in front of a commoditized no-code platform

Every horizontal builder facing the 2024–2026 AI-app-generation wave lands on the same fork, and Bubble is the clearest example of it. Path one is defend the horizontal position: keep the platform general-purpose, keep courting the non-technical founder, and try to out-ship the AI-first tools on generation speed. Path two is retreat to defensible ground: accept that scaffolding an app from a prompt is now a commodity, and monetize what happens *after* the prototype exists — production hosting, compliance, integrations, vertical workflow depth, and the multi-year contracts that come with them.

Path one is the emotionally comfortable choice and the financially worse one. The horizontal builder's core promise was compressing time-to-first-app. When tools like Lovable, v0, and Bolt collapsed that window from weeks of learning a visual editor to minutes of prompting, the promise stopped being scarce. A platform whose entire acquisition funnel rests on "you don't need to code" is now selling a benefit its prospect can get free, and the RevOps consequence is brutal: top-of-funnel volume may hold steady while conversion and expansion quietly rot, because the people arriving are tire-kickers who will churn the moment a faster option appears. You see this in the data before you see it in the revenue — trial-to-paid stays flat for a quarter, net revenue retention slips two or three points, and the sales team blames pricing.

Path two is uncomfortable because it means saying no to revenue you're currently collecting. Concentrating on dental practices, HVAC contractors, home services, and salon chains shrinks the addressable market on paper. It also means telling the roadmap committee that four categories get 60% of engineering and everything else gets maintenance. But it converts a race you can't win — generation speed against well-capitalized AI-native competitors — into a race where your accumulated platform depth is the advantage: database logic, workflow branching, permissions, payment integration, hosting reliability, and the boring operational surface that a scaffolded prototype doesn't have.

There is a third framing worth naming, because it's where most of the actual money sits: be the execution layer for the tools that beat you at generation. An agency scaffolds a client's booking app in an AI builder in an afternoon, then hits the wall every prototype hits — auth that survives real users, a payment flow that reconciles, audit logs, uptime commitments, a client who wants a change in month seven. That wall is a product category. Bubble sitting on the far side of it, taking a revenue share on the productionized app, is a better business than Bubble losing the scaffolding race on the near side.

The same fork shows up outside no-code. Design tools faced it when AI generation hit mockups. Email service providers faced it when AI copy tools commoditized drafting. In each case the winners moved downstream toward deliverability, compliance, and integration depth — the parts that are tedious to rebuild and expensive to get wrong — while the losers kept optimizing the thing that had just become free.

How to decide between them

The decision isn't a taste call. It's four diagnostics, run against your own data, and each one points cleanly at a path.

First: where does your revenue actually concentrate? Pull ARR by customer segment and by use case. If more than half your revenue already comes from a handful of recognizable industry patterns — appointment booking, field-service dispatch, client portals with payments — the vertical pivot is a description of your existing business, not a bet on a new one. Most horizontal platforms discover their long tail is thinner than the pitch deck implies.

Second: what's your net revenue retention by cohort, and is it moving? NRR is the honest scoreboard for whether your moat is real. Flat or declining NRR in the segment you claim to serve means customers aren't deepening — they're maintaining until something better appears. NRR above 110% in a specific vertical, even a small one, is a signal to pour resources there.

Third: what does win/loss say about why you lose? If you lose on price, pricing is fixable. If you lose on speed-to-first-app against AI-native tools, that's structural and you should stop fighting it. If you lose on compliance, security review, or "our IT won't approve it," that's a product gap with a direct dollar value attached — and it's the gap that opens enterprise contracts.

Fourth: how much of your churn is pricing-surprise churn? Consumption billing that produces a $500 development bill and a $3,000 production bill doesn't just cost you the churned account. It costs you the reference, the case study, and the community sentiment that fed your organic acquisition. Quantify it: what share of cancellations cite bill volatility, and what's the ARR attached?

Run these four in a two-week sprint, not a quarter-long strategy review. The inputs are all in systems you already own: billing, CRM closed-lost reasons, support tickets, and cancellation surveys. The failure mode is treating this as a positioning exercise for marketing rather than a resource-allocation decision with a headcount attached.

What the numbers look like under each option

Strategy documents float on adjectives. Here's the arithmetic that decides it, using ranges rather than invented precision.

Horizontal defense economics. Self-serve no-code pricing generally lands somewhere between roughly $30 and $400 per month per app or workspace, and the crucial problem is that the price doesn't scale with the value the app delivers. An app serving twelve internal users and an app serving eight thousand customers land in the same tier. That's a monetization ceiling baked into the packaging. Blended ARPU in the low-to-mid double digits monthly means you need enormous logo volume to move ARR, and every one of those logos is acquired through a funnel that AI-native competitors are now siphoning at the top. The CAC math only works if organic acquisition holds, and organic acquisition is exactly what pricing-surprise backlash destroys.

Vertical concentration economics. A multi-location service business — five dental offices, a franchised home-cleaning operation, an HVAC contractor with three crews — buys an operational system, not a builder. That buyer prices against what they're replacing: a scheduling tool, an invoicing tool, a client-portal tool, and the manual glue between them, which routinely totals several hundred to a few thousand dollars monthly. Landing at a few thousand per month per multi-location account is not a stretch; it's often a savings. The ARPU delta versus horizontal self-serve is roughly an order of magnitude, and the retention profile is completely different because you're now embedded in how the business books revenue rather than being a tool someone learned once.

The cost side matters just as much. Vertical templates are expensive up front — you're building booking, invoicing, payment reconciliation, client messaging, and the industry-specific compliance bits *once per vertical* — and then nearly free to deploy per tenant. That's the shape you want: high fixed cost, low marginal cost, and a sales motion that gets easier every deal because the reference customers are in the same trade association as the prospect. Sales cycles compress when the demo is a working dental-practice system rather than a blank canvas.

Enterprise and regulated-industry economics. Mid-market enterprises with compliance requirements pay serious money for low-code, and they pay it grudgingly, because the incumbent options are expensive and slow. A compliance-capable workspace tier — audit logging, data residency, field-level access control, SSO, documented security posture — is an add-on that unlocks buyers who are currently structurally unable to purchase. The revenue per account jumps to five or six figures annually, and the gross margin is excellent because most of the cost is one-time engineering plus audit fees rather than per-account service.

The honest trade-off: enterprise deals bring enterprise sales cycles. Procurement, security review, legal redlines, and a pilot phase mean six to nine months from first conversation to signature unless you shorten it deliberately — with pre-completed security questionnaires, a published trust page, reference architectures, and outcomes-based pilot terms. Every one of those artifacts is a RevOps deliverable, not a product deliverable, and they're the difference between a nine-month cycle and a four-month one.

Marketplace and ecosystem economics. Plugin and template marketplaces rarely become the primary revenue line, and pretending otherwise is how platforms waste a year. Their real function is switching cost. When an agency's client portfolio depends on certified integrations that exist on your platform and nowhere else, migration stops being a weekend project. Model the marketplace as a retention lever with a modest revenue tail, and it earns its keep. Model it as a growth engine and it will disappoint you.

Agency revenue share economics. This is the highest-leverage line and the most operationally demanding. An agency that builds and maintains client apps on your platform is a distribution channel with its own sales team that you don't pay for. Revenue share on their client billings aligns the incentive properly — they win when their clients expand, you win when they win. The demand is partner enablement: certification, template libraries, support SLAs that let them promise something to their own clients, and a partner portal that doesn't feel like an afterthought. Partner programs fail from under-investment far more often than from bad economics.

Sequencing the fix without breaking what still works

Order matters more than ambition here. Every one of these moves is defensible in isolation; run them in the wrong sequence and you'll churn your base before the new engines are producing.

Quarter one: stop the bleeding on pricing. Publish transparent, predictable tiers. Grandfather every existing account — no exceptions, no fine print — because the goal is repairing trust, and a migration that raises anyone's bill undoes the entire exercise. Ship a real-time consumption dashboard even if you're moving off consumption billing, because the anxiety was never really about the dollars; it was about not being able to forecast them. Offer annual prepay at a genuine discount. Cash flow improves and churn drops in the same motion. Measure success by two numbers: share of cancellations citing bill volatility, and trial-to-paid conversion.

Quarter one, in parallel: pick the verticals and say no publicly. Choose three, maybe four. The selection criteria are concrete — existing customer density, average business size large enough to support a meaningful monthly price, workflow complexity high enough that a prompt-generated prototype won't cut it, and a trade-association ecosystem you can market into. Then actually deprioritize everything else on the roadmap, in writing, to the whole company. The half-committed version of this strategy is worse than either pure option, because you carry the cost of vertical depth without the focus benefit.

Quarter two: ship one vertical end-to-end before starting the second. Not 60% of three verticals. One complete operational system — booking, client portal, invoicing, payments, reporting, and the industry-specific pieces — with a reference customer running production on it. The temptation to parallelize is enormous and it's a trap: three half-finished verticals give you three unconvincing demos.

Quarter two: launch partner certification with the first cohort. Twenty to fifty agencies, hand-picked, with the founders in the room. Their feedback shapes the program before you scale it to hundreds. Get the revenue-share mechanics, the support escalation path, and the template licensing terms right at small scale, where fixing them is cheap.

Quarter three: open the compliance tier. Audit logging, data residency options, granular access control, SSO, and the documentation package that lets a security reviewer approve you without a six-week back-and-forth. Sell it to design partners first, at a discount, in exchange for the case study and the willingness to be a reference. The regulated-industry pipeline builds slowly and then all at once — the first three logos in a vertical make the next fifteen dramatically easier.

Quarter four: position as the production layer for AI-scaffolded apps. Build clean import paths from the popular generation tools. Publish the message plainly: prototype wherever you like, ship it here. This is only credible after the vertical templates and compliance tooling exist — announce it in quarter one and it reads as spin.

The sequencing principle underneath all of it: fix trust before you fix monetization, and prove one vertical before you fund three. Platforms that invert this end up with a beautiful new pricing page and a support queue full of people asking why their bill changed.

The RevOps machinery this pivot actually requires

Strategy decks skip this part and it's where the pivot lives or dies. Moving from horizontal self-serve to vertical contracts plus enterprise plus partner revenue means running three fundamentally different revenue motions simultaneously, and the operating infrastructure for that doesn't exist by default.

Segmentation has to be rebuilt from scratch. Your CRM almost certainly segments by plan tier and company size. Neither predicts anything useful now. You need industry vertical, multi-location versus single-location, whether the account came through a partner, and whether it has compliance requirements. Every one of those fields drives routing, pricing, and expansion motions differently. Retrofitting them onto existing records is unglamorous data work that has to happen before the sales team can act on the strategy.

Forecasting breaks in a specific, predictable way. Self-serve revenue forecasts from cohort curves. Enterprise revenue forecasts from pipeline stages and win rates. Partner revenue forecasts from partner-reported pipeline, which is optimistic and late. Blending all three into one number produces a forecast nobody trusts. Run them as separate models with separate accuracy tracking, then roll up — and expect the partner model to be your least accurate line for at least three quarters.

Compensation design decides what actually gets sold. If reps earn the same on a self-serve upgrade as on a vertical contract, they'll chase the fast close every time, because it's rational. Weight the plan toward multi-year vertical contracts and enterprise ACV, and pay partner-sourced revenue at a rate that makes partner enablement worth a rep's afternoon. Comp is the most direct lever you have on rep behavior, and it takes a full quarter to see the effect — change it early, not after the first disappointing quarter.

Attribution gets genuinely hard. A prospect who scaffolds a prototype in a competitor's tool, hits a wall, searches for a production path, and lands on you through a partner referral has a journey no single-touch model captures. Accept directional attribution and instrument the two moments that matter most: the wall (what broke in the prototype) and the referral source. Those two data points drive more product and partnership decisions than a perfect multi-touch model would.

Churn definitions need surgery. A hobbyist canceling a $30 plan and a five-location franchise canceling a $3,000 contract are the same event in a naive logo-churn metric and completely different events in reality. Split the metric by segment on day one of the pivot, or your dashboards will show improvement that isn't real — or panic that isn't warranted.

Support becomes a revenue function. In vertical SaaS the support ticket volume per account is higher and the stakes are higher, because you're inside the customer's revenue path. A booking system that fails on a Monday morning costs a dental practice real money and costs you the renewal. Staff for that before you sell into it, and route vertical accounts to specialists who know the industry workflow rather than to a general queue.

What to watch for while the pivot is running

Two failure modes account for most botched platform pivots, and both are visible early if you're looking.

The first is hedging. The company announces vertical focus, but the roadmap still funds horizontal features, sales still takes any deal that walks in, and marketing still runs the old generalist message because it produces more leads. Six months in you have vertical-tier pricing, horizontal product depth, and a positioning statement nobody in the company can repeat. The tell is a roadmap where the deprioritized categories keep getting "small" allocations. Audit engineering time against the stated allocation monthly, and treat drift as a leadership problem rather than a planning error.

The second is abandoning the base too fast. The hobbyist and individual-builder tier looks like a distraction on a spreadsheet: low revenue, high support load, loud in the community. But that tier is where your evangelists live, where your organic search surface comes from, and where the next generation of agency partners learns the platform. Starving it saves a little money and quietly kills the acquisition engine that feeds the enterprise pipeline three years later. Keep it healthy and unglamorous — stable, well-documented, honestly priced — just stop building roadmap around it.

A third, subtler risk: the AI-tool relationship is a moving target. Positioning as the production layer for prototypes generated elsewhere works only while those tools stay weak at production concerns. They won't stay weak forever. The durable moat isn't the import path — it's the vertical workflow depth and the compliance posture, both of which take years to replicate and neither of which a generation tool is naturally positioned to build. Lean on those, and treat the import path as a distribution tactic rather than a strategy.

Finally, watch the leading indicators rather than ARR, which lags everything by two quarters. Trial-to-paid conversion inside the target verticals. NRR by vertical cohort. Sales cycle length on enterprise deals. Partner-sourced pipeline as a share of total. Share of cancellations citing bill volatility. If those five are moving the right direction, the revenue follows whether or not this quarter's number cooperates.

Related questions

Does narrowing to verticals shrink the company's ceiling?

Near-term addressable market shrinks; realistic revenue usually grows. A defensible position in four verticals at high ARPU and high retention beats a contested position across forty at low ARPU. The ceiling comes from adding verticals later, once the template-building process is repeatable and cheap.

How do you move off consumption pricing without a revenue hit?

Grandfather everyone, price the new tiers at or slightly above current effective spend for typical usage, and push annual prepay with a real discount. Short-term revenue is roughly flat; churn drops and forecasting improves immediately. The gain shows up in retention, not in the first quarter's bookings.

Is the compliance tier worth the audit cost?

Usually yes, if win/loss shows deals dying at security review. SOC 2 and similar audits cost real money and months of preparation, but they convert an entire buyer segment from unable-to-purchase to purchasable. Check your closed-lost reasons before committing.

What happens to existing agency partners during a vertical pivot?

They're the biggest winners if handled well. Vertical templates make their delivery faster and their margins better. Bring the top partners in before the announcement, let them shape the certification program, and they become the pivot's loudest advocates rather than its first casualties.

Can a platform run all three revenue engines at once?

Yes, but not from day one. Each engine needs its own segmentation, comp plan, forecast model, and support tier. Sequence them a quarter or two apart so the operational infrastructure for each is genuinely working before the next one starts competing for attention.

FAQ

Why not just build a better AI generation feature and compete directly?

Because you'd be entering a race late, against companies whose entire architecture, funding, and hiring were designed around it. Shipping a generation feature that reaches rough parity with free alternatives creates marketing debt, not a moat — you've spent a roadmap cycle to arrive where your competitor started. The asymmetry runs the other direction: production reliability, workflow depth, and compliance are hard for a generation-first tool to bolt on, and easy for an established platform to already have.

How many verticals is the right number?

Three to four for the first year. Fewer than three concentrates too much risk in one industry's economic cycle. More than four and you can't build genuine depth in any of them, which puts you back in the horizontal position with a smaller market. Add the fifth only after the first three are each producing predictable revenue and the template-building process has become routine rather than heroic.

Does the marketplace revenue share actually matter financially?

Directly, rarely — marketplaces are usually a modest revenue line. Indirectly, substantially. Certified integrations that exist on your platform and nowhere else turn migration from a weekend project into a quarter-long one. Judge the marketplace on retention impact and partner engagement, not on its own P&L contribution, and you'll fund it correctly.

How long before this shows up in reported revenue?

Expect eighteen to twenty-four months for the full picture. Pricing changes show in churn within a quarter. Vertical contracts start closing in months four through six and compound from there. Enterprise pipeline takes two to three quarters just to reach first close. Anyone promising a turnaround inside two quarters is describing a pricing change, not a strategy change.

What's the single biggest execution risk?

Hedging. Companies announce focus and then keep funding everything, which produces the costs of both strategies and the benefits of neither. The mechanical defense is an engineering-allocation audit each month against the stated split, reviewed by leadership, with drift treated as a decision to reverse rather than a variance to explain.

Does any of this apply to platforms outside no-code?

The pattern generalizes cleanly. Any horizontal tool whose core value proposition gets commoditized by AI faces the same fork — defend the commoditized layer or move toward where value concentrates: integration depth, compliance, reliability, and industry-specific workflow. The specifics differ by category; the diagnostic questions and the sequencing logic don't.

Sources

• https://www.gartner.com/en/information-technology/glossary/low-code-application-platform-lcap
• https://www.forrester.com/blogs/category/low-code-development-platforms/
• https://hbr.org/2016/12/whats-your-data-strategy
• https://a16z.com/the-new-business-of-ai-and-how-its-different-from-traditional-software/
• https://www.bvp.com/atlas/state-of-the-cloud-2024
• https://techcrunch.com/category/enterprise/
• https://www.aicpa-cima.com/topic/audit-assurance/audit-and-assurance-greater-than-soc-2
• https://stripe.com/guides/atlas/business-of-saas
• https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/the-b2b-elements-of-value

Related on PULSE

• How'd you fix Aston Carter's revenue issues in 2026?
• How'd you fix CyberCoders's revenue issues in 2026?
• Is AI a bubble in 2027?
• Should I open or buy a Vivi Bubble Tea franchise in 2027?
• How Many Employees Should I Schedule Each Shift at My Bubble Tea Shop?
• How'd you fix Illinois's NIL & athletic revenue issues in 2026?