Subscribers no longer want to set and forget. Our UK subscription trends 2026 research found they increasingly expect to scale their commitment up or down to match the value they’re getting. Consumer software subscribers go further; they want to shape how they buy. They also expect the AI features on their bill to earn their keep, and they're twice as likely as the wider market to use AI to research a subscription before they commit. Here's how to give them that flexibility without giving away revenue.
A fixed set of tiers works like a ladder: the only way is up. But software customers’ needs don’t always change in neat steps. A team might grow in one area and shrink in another, or need one premium feature without paying for the whole tier that comes with it. That’s why more than two-thirds (68.9%) of software subscribers surveyed want flexible ways to upgrade. Providers that offer more ways to adjust a plan, rather than making them choose between climbing a ladder and walking away, can better match each customer’s needs, giving them a reason to stay.
So how do you build pricing that provides fluidity and flexibility, without losing customers or revenue?
Ask software subscribers how they’d prefer to upgrade and you won’t get one answer. You’ll get four, almost evenly split:
A single upgrade path, however well designed, will suit only a minority of your customers.
Many of them are already paying for something they don't use (26.9%), against 16.5% across the market. That’s a significant churn risk.
The good news is that software subscribers can be persuaded to stay. Two-thirds (66.8%) have changed their mind about cancelling after being offered a pause or discount, the highest share of any segment we surveyed.
Leaving is about to get easier
From January 2027, the Digital Markets, Competition and Consumers Act, subscriptions will be easier to understand and cancel. Consumer software is already less sticky than B2B: average monthly churn for B2C SaaS runs at around 5.6%, against roughly 3.1% for B2B (Focus Digital). When cancelling is simple, plans that don’t fit will be dropped.
That leaves three things SaaS leaders need to get right:
[Chart: upgrade routes (published report) and reasons to stay (raw survey data), software subscribers, base 193]
The same options that help them grow also stop them leaving. Feature add-ons, usage-based pricing, and bundles appear in both lists, so building them once supports both expansion and retention.
The strongest retention lever, a multi-product bundle (43.5%), isn't an upgrade at all but a sideways move. The wider market is already moving this way: hybrid pricing was used by 61% of SaaS companies by 2025, up from 49% in 2024 (Metronome, Flexera), and usage-based pricing is linked to monthly churn of 2.1%, against 5.6% for single flat-rate (Focus Digital). These benchmarks draw mostly on B2B companies, and the churn gap partly reflects which businesses adopt usage pricing, but the direction matches what consumers told us.
Every new way to downgrade or unbundle can be a way for ARPU to fall. Every extra option on a self-serve paywall is one more decision a new customer might not make.
The businesses that get this right do three things:
AI features that change what customers get out of the product earns loyalty. Personalisation on its own barely registers, and nearly half say no AI feature would change their mind. That matters because customers are increasingly being asked to pay for it: 44% of SaaS companies now charge for AI features (Maxio, GetMonetizely). If customers can’t see the value, they won’t accept the price.
The cost behind every prompt
AI features earn loyalty when the result is tangible, but they aren’t free to run, so how you price them matters as much as what they do. Today, flat plans dominate: half of US consumers paying for AI are on a flat monthly plan while 15% combine a subscription with usage charges as their use outgrows the plan (Menlo Ventures). A flat price is simple, but it overcharges light users and under-recovers from heavy ones.
That’s where usage pricing can earn its place, as 35.3% of software subscribers said that AI-based usage pricing would make them more likely to stay.
Software subscribers are more likely to use AI to choose a subscription, and more likely to act on what it recommends. Your pricing is increasingly interpreted and compared before customers ever reach you. Ensuring that your pricing, bundles, and offers are machine readable is becoming more and more important as consumers increasingly turn to AI to find them the best deal for a given feature rather than rely solely on brand recognition and brand-owned channels.
Clear pricing builds trust
There is a tension between consumers’ cost-consciousness and their distrust of opaque personalised pricing. 71.2% would be less likely to renew if they found their provider was pricing dynamically using algorithms, AI, or personal data, however 37.4% would share their personal data if it meant their subscription price fell when their usage fell. Dynamic pricing that customers understand is welcome; opaque algorithmic pricing imposed on them is not.
Start with actions that protect revenue, then build the capabilities to offer customers more flexibility.
Start with Phase 1 now to protect revenue while you build the rest. The providers who keep their customers will give them a move in every direction, so there is always another hold to reach for.
Unless stated otherwise, figures come from UK subscription trends 2026: Subscriber behaviour in the age of AI (Clarasys and Stripe, September 2026). The research is a YouGov survey of 1,082 UK adults with at least two subscriptions, carried out online in June and July 2026. The software subscriber base is 193.
The retention option figures (43.5%, 37.3%, 35.3%, 28%) and AI feature figures (37.3%, 34.7%, 34.2%, 18.1%, 45.6%) come from the same survey's raw data and are not published in the report.
External sources: