Story

Harvest price increase: what Bending Spoons changed and how indie builders are responding

Kyle Kingsbury has billed his contracting work through Harvest for about a decade. On 8 November 2026 he wrote, under the name aphyr, that Harvest raised his price “from $13.32 to $133.98 per month”. Bill figures in this story are what each customer reported. That jump is the Harvest price increase behind a BBC story, a pricing page of starting prices, and a long Hacker News thread.

The post is dated 8 November 2026. He describes himself as a self-employed contractor, and Harvest as time tracking plus invoicing, with a Stripe integration for credit and ACH payments. The new charge, he writes, came “by moving everyone to a new “enterprise” plan and “unlimited” usage billing model”, which is, “as far as I can tell, exactly what I had before.”

What the Harvest price increase changed

The public record of the change is a short calendar.

When What is on the record
2006 Iridesco launches Harvest. Founders: Danny Wen, Shawn Liu.
July 2025 Bending Spoons acquires Harvest.
2026 Seat prices gain a charge for how the account is used.
20 August 2026 BBC reports customer renewals, including $130 to $2,110 a month.
22 August 2026 Subscription Insider names Flex and Unlimited.
8 November 2026 aphyr publishes $13.32 to $133.98.
9 November 2026 Hacker News, 01:58 UTC (442 points and 147 comments when checked on 12 November).

The 2006 launch and the 2025 sale are on Wikipedia, which dates the acquisition to 2025. July 2025 is the month in Subscription Insider, 22 August 2026. The August bills are in Laura Cress's BBC article of 20 August 2026. The November invoice is aphyr's post. The thread is Hacker News, posted by zdw.

The wider bills are in the BBC piece. Richard Haldenby, of the UK consultancy Salentis, told Laura Cress that his monthly bill had risen from $130 (£95.50) to $2,110. He said he was “in shock”. The same article says a customer in the US reported an annual charge going from $2,800 to $23,000. The BBC described rises of as much as 1,500%.

Haldenby told the BBC the firms had used Harvest for at least 15 years. After he objected, he was offered $1,309 for the next year, paid up front. He called it a way to “delay the inevitable”, and said he is “planning to migrate to a similar but less good service”.

Subscription Insider treats those renewals as individual experiences. The amount, it writes, depends on the account and the billing option.

aphyr’s $133.98 drew a direct challenge on Hacker News. asutekku listed the public annual prices they could see: Free, Teams at $9 a month, and a plan they called Enterprise unlimited at $14 a month. A $133.98 plan, they wrote, “does not even exist” on the site. The same comment calls that gap “Something is a bit fishy here”, and guesses that an extra 3 was added to $13.98, or that a yearly price was read as monthly.

Aurornis answered with the line above those prices: “You need to read the “Starting from” text on their pricing page right before those prices you see.” The lower monthly fee, in that comment, comes “if you pay a higher percentage of your billings.” Aurornis also wrote: “This is happening to all of their customers.” Harvest has not published a count of who was moved.

Why did Harvest raise its prices?

Harvest’s account is in the BBC article. A spokesperson said customers on “outdated” legacy plans, some dating back to 2011, faced the larger increases. Customers are notified “30 days” by email and then “10 days before their renewal date”, and they can switch plans “in one click”.

Mark Peacock, head of the pricing consultancy PriceMaker, told the BBC: “Harvest has completely failed at the transparency test.” His point, in the same article, is that a customer cannot work out the cost until the bill arrives at the end of the month.

On the November thread, Centigonal offered a buyer’s sketch and opened with “Not saying I like their business model”. If a buyer takes a company, multiplies revenue by ten, loses half the customers, and keeps the margins, Centigonal wrote, they’ve “quintupled that company’s profits.” Harvest has not published a tenfold revenue figure, or how many customers left.

kjksf argued from the sale itself: “very profitable businesses don’t sell themselves to Bending Spoons.” The next move in that comment is the caveat. “This might not apply to Harvest.” The sale price and the state of the business beforehand, kjksf wrote, are not public. The comment then guesses at a company that could not control its spending, or owners who wanted one cheque and a way out of the daily work.

aphyr’s own try at an explanation was a support email. He writes that they don’t read customer emails any more, and that “they have a Large Language Model repeatedly vomit thousands of words at you.”

The pattern people reached for is older than this invoice. The BBC says Bending Spoons has acquired more than 50 firms since 2013, among them Evernote, Vimeo, and WeTransfer, and that it has raised prices at companies it has bought. After Evernote in 2022, the BBC reports sharp subscription increases and layoffs in the US and Chile. The BBC’s line on Vimeo and WeTransfer is tighter limits on free tools, with legacy users moved onto higher tiers.

Who owns Harvest now?

Wikipedia says Harvest is a web time tracker launched in 2006 by Iridesco LLC. Danny Wen and Shawn Liu built it because their own studio needed to track time and invoice clients. The page records the Bending Spoons acquisition in 2025.

Subscription Insider puts that deal in July 2025. It also writes that Harvest says more than 70,000 companies use the software.

aphyr dates his worry to the 2025 sale. He calls the buyer “the company that gutted Vimeo”, and writes that he was busy and did not go looking for a replacement then. The BBC’s report is the narrower one: limits on free Vimeo tools, and legacy users on more expensive tiers.

The BBC calls Bending Spoons an Italian tech company, founded in 2013. On Hacker News, throwaw12 groups it with private-equity buyers. The question attached to that is below, in the section on building a replacement.

How does Harvest’s new pricing work?

The pricing page, read on 12 November 2026, leads with “starting from”.

Plan on the page Listed starting price What that start includes
Free $0 1 seat, 2 projects
Teams $9 / seat / month annually, or $11 monthly Unlimited seats, then usage
Enterprise $14 / seat / month annually, or $17.50 monthly Teams, plus reports, approvals, SSO, then usage

Listed starting prices on getharvest.com/pricing, read 12 November 2026. Teams and Enterprise both say that additional invoices, projects, clients, and tasks are billed based on what you use. The Free block says the usage of some features may require fees, and points to a help centre. Enterprise Plus is also listed, for custom work, with no public starting price on that read.

The sentence under Teams and Enterprise is Harvest’s: “additional invoices, projects, clients, and tasks are billed based on what you use, so you’re never overpaying.”

The names at the top of that page are Free, Teams, Enterprise, and Enterprise Plus. Flex and Unlimited are the names Subscription Insider uses for the usage choice. With Flex, the bill can change with activity inside Harvest, including projects, clients, tasks, and invoicing. With Unlimited, the customer pays a fixed usage fee on top of the per-seat price. The article says Harvest does not publish a standard Unlimited fee on the pricing page, which makes the full cost hard to compare from outside an account.

That is the gap between asutekku’s list and Aurornis’s “Starting from”. The number in the column is a seat price. The invoice can add use.

The BBC describes the same shift from the old flat fee per user. Under the new model, it says, a business can also be charged for active projects, clients, and tasks, and for revenue invoiced through the platform.

A “flex” plan, he writes, costs $12.65 per month, and “skims an additional 2% on top of the usual ~3% card transaction fees.” He writes that the usual card fee pays for the card network, the people at Stripe and Mastercard, and the risk of chargebacks and fraud, and that he does not need to spend 2% of gross revenue “to send a couple PDFs a month and wrap a Stripe web form.” Reading a price down to the line that actually bills you is the fight on this page. A different 2026 price page with that same kind of close read is whether Stripe Atlas is free.

Digory posted an email from the account. It put a Basic plan at $40.00 a month with 5 seats, renewing on an Enterprise Unlimited plan at $586.50 a month. Digory called them seat fees plus the new usage fees, for “unlimited” use. The next sentence in the comment is “Same product.” After dropping spare seats and leaving unlimited, Digory wrote, the bill was about $75 a month per seat.

herewulf wrote of using Harvest for 10 years. “They tried to 10x my monthly fee with these defaults also.” After the switch, herewulf went over the usage limit “using too many labels on my tracked time (tasks)” and “got hit with an additional 2.5x of my normal monthly spend.” The comment gives no dollar amount under those multipliers.

hubraumhugo started snapshotting the pricing pages of Bending Spoons companies, to keep a time series. The comment treats the page as something that will move.

What are the alternatives to Harvest?

The products below were posted by their builders, or by a reader recommending one, on aphyr’s post and in the Hacker News thread.

On aphyr’s post, Lucas Cherkewski recommended Cushion. He had switched years earlier, while freelancing. He wrote that it includes invoicing integrated with Stripe, “with no markup”, and described the company as “Kind of a one-man shop, but dependably so”.

waffl posted Zebu after what the comment calls nearly 15 years on Harvest. The price waffl posted is €5 a month, with no usage fees, a 30-day free trial, and a full Harvest migration. “My only real difficulty has been marketing/gaining traction”, waffl wrote, because buying advertising for a project at that fee “has proved to be very difficult/impossible”.

basketbla built billable.to for a family business, and wrote that a handful of other practices use it too. The pricing line in the comment is “Flat rate pricing not seat based!”

abejora stated the bind in two sentences: “The cost of moving is high because you have been on their platform for years.” And: “The cost of staying is high because of the price increases.” They moved to a platform they built, at abejora.com. “Staying would have been cheaper in the end, but this one is more fun.” The same comment says the current plan is to add invoicing next, including a Peppol integration. Invoicing, on their own account, was still ahead of them.

Years of data are why the move hurts. A host can also end the product on a calendar. Val Town’s shutdown, and the migration dates that came with it, is that other bill: the service stops, and the export becomes the work.

Digory, after posting the $40 to $586.50 email, wrote: “if someone wanted to take over angry Harvest customers, it would probably be a good little business.”

Can you vibe-code your own Harvest replacement?

throwaw12 asked a question that has been circulating all year. Once a buyer like Bending Spoons acquires the company, can you start the same product, finish it by the time the pricing breaks, and have the customers waiting?

njovin answered: “Building a thing and selling a thing are very different actions,” and “the selling is the hard part.” The same comment grants the customer’s side of the idea. If a buyer takes a tool you rely on, njovin wrote, start a private replacement, because even a rough one may beat what arrives later. They named the sale as the hard part. That split is also the one in whether indie hacking is dead: the build got cheap, and finding a buyer did not.

2arrs2ells posted a private version that stays on their own machine. “Fable built me a version that covered all of my needs in an hour.” It “Runs on localhost with SQLite”. Clients pay by ACH or wire, and the comment says a Stripe integration might have taken a day. The spec, they wrote, was a Harvest data export and the PDF invoices they already had.

wrs told Claude to use Beancount, “I don’t want probabilistic accounting!”, and described a small Go, SQLite, and React app plus Python scripts, on a laptop: hours by project, a PDF invoice, bank imports, and folders of receipts. The limit is in the same comment: “I’m still using Gusto, because they take some responsibility for getting my payroll taxes right, and filing them.”

A public ledger of the same impulse, with the founder’s own prices attached, is Pieter Levels replacing a stack of paid tools with software he generated. The argument that followed, about what a software company can still charge for, is the replies to his “software is dead” post.

mingus88 posted the warning in one line: “Go ahead and vibe code your payment system. What could go wrong?”

BowBun wrote: “Lots of software companies are sitting on no moat.” The next sentence begins “Not talking about Harvest”. The example in the comment is Jira, and products that, in BowBun’s view, pack in features people do not use and then raise the price.

Customers reported sharp jumps on renewals they had paid for years. Harvest’s spokesperson said the larger increases hit outdated legacy plans, with notice by email and a switch in one click. Builders in the thread posted alternatives, and njovin answered that selling is harder than building.

What do you think? If a tool you have used for years sent a renewal like one of these reported bills, would you switch, build your own, or pay?

FAQ

Why did Harvest raise its prices?

A Harvest spokesperson told the BBC on 20 August 2026 that larger increases hit outdated legacy plans, some dating to 2011, after 30 days' email notice and another note 10 days before renewal, with a one-click switch. Centigonal sketched ten times the revenue and half the customers as quintupled profits, and was not endorsing it. kjksf wrote that very profitable businesses do not sell to Bending Spoons, which might not apply to Harvest. Mark Peacock said Harvest failed the transparency test.

Who owns Harvest now?

Wikipedia says Danny Wen and Shawn Liu launched Harvest in 2006 through Iridesco. Bending Spoons acquired it in 2025, and Subscription Insider dates that to July. Harvest says more than 70,000 companies use it. The BBC calls Bending Spoons an Italian company founded in 2013, with more than 50 acquisitions, including Evernote, Vimeo, and WeTransfer. aphyr calls the buyer the company that gutted Vimeo. The BBC says Vimeo limited free tools and moved legacy users onto higher tiers.

How does Harvest's new pricing work?

On 12 November 2026: Free $0 (1 seat, 2 projects); Teams from $9/seat/month annually or $11 monthly; Enterprise from $14 or $17.50 the same way. Extra invoices, projects, clients, and tasks are billed on use. Flex varies with activity. Unlimited's fee is not on the public page. Customers reported $12.65 flex plus an extra 2% on ~3% card fees, $13.32 to $133.98, and $40 for 5 seats to $586.50. asutekku said $133.98 is unlisted. Aurornis pointed at Starting from.

What are the alternatives to Harvest?

Builders posted these in the thread. Lucas Cherkewski recommended Cushion and called it a one-man shop. waffl posted Zebu at 5 euros a month, with no usage fees and a Harvest migration, and said traction was the hard part. basketbla posted billable.to, flat-rate, for a family business. abejora described a self-built move and the high cost of leaving or staying. Digory said taking over angry Harvest customers would probably be a good little business.

Can you vibe-code your own Harvest replacement?

2arrs2ells wrote that Fable built a localhost SQLite app in an hour. wrs described a Go, SQLite, and React app from Claude and Beancount, and still uses Gusto for payroll taxes. mingus88 asked what could go wrong vibe-coding payments. BowBun said many firms sit on no moat, and was not talking about Harvest. throwaw12 asked if you can rebuild once a buyer like Bending Spoons acquires the product. njovin said selling is the hard part.