Manny Medina, who previously gained recognition as the founder of Outreach, a sales automation company valued at $4.4 billion, has once again impressed investors with his latest venture, Paid.
Paid recently secured an oversubscribed seed investment of $21.6 million, led by Lightspeed. Combined with the €10 million pre-seed funding it obtained in March, the London-based company has now accumulated $33.3 million in total funding, all before reaching its Series A. According to a source close to the transaction, Paid’s valuation has surpassed $100 million.
Emerging from stealth mode in March, Paid introduced a novel approach to the AI agent ecosystem: rather than providing agents themselves, the company enables agent developers to monetize their algorithms by charging customers based on the measurable value their agents deliver. This approach, often referred to as “results-based billing,” is gaining traction in the AI sector.
Paid aims to empower agent creators to “begin billing for the margin improvements their clients achieve,” as Medina explains.
This represents a fresh pricing model for software in the era of AI, moving away from the traditional SaaS structure of unlimited usage or per-user fees, as well as the one-time purchase model of the client/server age.
The per-user pricing model is ineffective because agent developers incur costs from both model and cloud service providers. Allowing unlimited usage could result in significant financial losses. (This is a common challenge for startups focused on coding automation.)
Instead, agent providers “must demonstrate the tangible benefits their agents bring to customers, since these agents typically operate behind the scenes,” Medina told TechCrunch. When agents perform as intended, they are assigned more tasks, and their increasing workload often goes unnoticed.
“If your agent isn’t actively contributing, it won’t generate revenue,” Medina states. “There needs to be a system in place that allows agents to bill for the extra work they perform,”
However, adopting a monthly subscription for a fixed number of credits—similar to the approach used by model developers and coding startups—also poses risks for agent creators. This is because businesses are reluctant to pay for subpar AI output, which remains prevalent. According to a recent MIT study, after billions invested in AI pilot programs, only 5% of enterprise projects reached production, with the remaining 95% failing to deliver value.
Companies are unwilling to pay for agents that generate unnecessary emails that go unread.
Among Paid’s early adopters is Artisan, a sales automation startup that has gained viral attention. (Artisan’s CEO, Jaspar Carmichael-Jack, will be discussing this topic at TechCrunch Disrupt next month.)
Paid is also gaining traction with SaaS companies exploring agents as their next major growth driver. The company recently announced ERP provider IFS as a new client.
Alexander Schmitt of Lightspeed shared that the firm has invested “over $2.5 billion in AI infrastructure and application companies over the past three years,” and has observed that most AI pilot projects do not succeed.
“The main issue is that it’s difficult to quantify the value agents are currently providing,” Schmitt commented.
Schmitt believes Paid stands out with its distinctive strategy, noting, “we haven’t seen anyone else develop something like this.” If this results-based billing model truly helps agents become widely adopted, more competitors are likely to emerge.
The funding round also saw participation from new investor FUSE and existing backer EQT Ventures.
