❓ Problem

At a ten-person company, there is no one whose job is to check the bills. The electricity, the insurance, the card processing fees, the waste contract and the eleven software subscriptions all sit on auto-pay, quietly renewing at whatever rate the vendor picked on day one. Mar Hershenson's framing in Pear's RFS nails the mechanism: the savings on any single line item never justify the time, so nobody looks, and in aggregate the leakage is enormous.

Card processing is the cleanest example because every business has it and nobody understands it. Most small businesses pay between 1.5% and 3.5% per transaction, of which only the processor markup is negotiable. One firm that audits merchant statements for a living puts the gap between what a business thinks it pays and what it actually pays at 30 to 80 basis points, and widening. Across the economy, processing fees cost merchants $148.5 billion in 2024.

Energy runs the same play in the 14 states where businesses can actually shop. Let a commercial contract lapse and the provider rolls you onto holdover rates that run 2-3x what you were paying, and most owners find out on the next bill. An entire intermediary layer exists to farm that gap, with over 600 licensed brokers operating across the deregulated markets, paid out of a markup you never see itemised.

Software is the same disease with nicer branding. Vertice's 2026 data puts 66% of SaaS licences as either untouched or surplus to requirements, of which 15% is pure shelfware that survives precisely because auto-renewal fires and nobody is watching the calendar.

Then there's the kicker, which is really an arithmetic problem. Clawing back $40 a month on a waste contract is not worth your afternoon. That's exactly why the incumbent fixers price the way they do: Billshark takes 40% of first-year savings, BillFixers 50%, and Rocket Money lets you choose between 35% and 60%. Those aren't greedy prices. They're the cost of a human being on hold. The unit economics of caring about a small bill have always been terrible.

Agents don't sit on hold. They don't have an afternoon. And they don't need 40%.

✅ Solution

An always-on procurement agent that reads every recurring bill a business pays, then works the ones no human would ever bother picking up.

• Connect the bank feed and the inbox. The bank feed is ground truth on what actually leaves the account. The inbox is where the contracts live. Between them, the agent maps every recurring charge to a vendor, a category, an amount and a renewal cadence, which for most SMBs is the first time that list has ever existed.
• Pull the terms out of PDFs nobody has opened since signing. Contract end date, notice period, auto-renewal clause, unit rates, annual escalator. This is the unglamorous core of the product and the hardest bit to copy, because the leverage isn't in knowing the price, it's in knowing the date.
• Start the switch 90 days out without being asked. The agent fires before the notice window closes, runs the market comparison, drafts the paperwork, files the cancellation and comes back with two or three options and a recommendation. The owner approves or ignores.
• Begin with one boring category. Card fees, energy or insurance. One category done autonomously and proven with a real savings number beats a dashboard covering everything.
• Then spread across the whole bill stack. Telecoms, waste, insurance renewals, payroll services, software. Same machinery, new category, near-zero marginal cost.

📊 Key Numbers

Market size

• The US has 36.2 million small businesses, but the real target is the 5.58 million employer firms with staff, premises and a bill stack worth working.
• A 10-50 person business carries somewhere between $75k and $150k a year of recurring non-payroll vendor spend once you total energy, insurance (typically $500-$2,000 for general liability alone), telecoms, card fees, waste and software. Across 5.58 million employer firms that's north of half a trillion dollars of recurring spend that literally nobody negotiates.
• The UK adds another 5.7 million businesses as a later expansion market, roughly 1.6 million of them with employees.
• For reference, the business spend management software market is $31.93bn in 2026, heading to $49.83bn by 2030. Almost all of it is sold to companies large enough to employ a procurement team.