RIP, Harvest

Can you not 1-shot prompt an agent to build an equivalent piece of software for yourself?

• mingus88 · 2 hours ago

  Go ahead and vibe code your payment system. What could go wrong?

  • skinfaxi · 2 hours ago

    Harvest is a stripe wrapper with time keeping. They are not suggesting to vibe code stripe.

  • dhosek · 2 hours ago

    I’ve found that while vibe coding can be effective, anything with complexity takes more than a 1-shot prompt to get there. The amount of work that comprises a single scrum story is about the right amount to do with a single prompt and even that requires some review and revision. So 1-shot your payment system, hell no, but writing a payment system without writing code is possible although I would think it still would take enough time that I wouldn’t choose that over a reasonably priced SaaS.

  • neya · 1 hour ago

    Come on now, we need a blog post blaming the LLM model when this happens. It always crackw me up when people want to vibe code finance related applications. Let's not devoid ourselves of entertainment, please.

• ChuckMcM · 2 hours ago

  I believe your question answers itself, if you could do this then people would[2]. The next question might be 'What prevents the current LLM based coding tools from being able to do this?" Now there has always been software that people have used that they "could" write for themselves but they didn't because they didn't want to maintain it. Seems like things that are like that would be good candidates for this right? But this question shares at least a spiritual geneology with the response to some new startup code that does some thing, and the comment here is "Gee, I can't believe that got this funded, I could write that in a weekend." The problem is, there are many things that are easy to describe but difficult to pin down, and reality is complicated. So many times I've gone into a project thinking it would be "easy" only to start uncovering layer after layer that are much more involved than I expected. So no, you cannot '1-shot prompt' pretty much any significant software, and that goes doubly true if it affects your life in any potentially devastating way. One of the things I'm constantly amused by[1] is the Solitaire program on iOS that is clearly being vibe coded these days because it comes up with daily challenges like 'win a hand in -2739551 moves'. Like no human would ever write code that screwed up basic math but LLMs do it all the time. [1] Obviously I am easily amused :-). But it is sort of the 'pelican riding a bike' test where you can easily tell what a human would do vs what an LLM does. [2] EDIT: another commentator says they vibe coded it in a weekend. That would support the hypothesis that the application could be at lease a small number of shots vibe coded. That said, I'd love to see that code and the prompts that led to it.

• ares623 · 2 hours ago

  So many one shot opportunities and yet so few actual links

  • bdangubic · 2 hours ago

    A lot of people use "one-shot" things not to sell it or provide links but for personal usage and to replace tools etc they used to pay for (as few did in this thread).

    • ares623 · 1 hour ago

      Hmm right. Personal usage I can get behind. But if I had a coworker/employee just constantly finding ways to one-shot tools we're paying for I'd be pretty miffed. If that's all they're doing why are they even there. Just use their salary to pay for the services at that point.

I vibe coded my own harvest or toggl app in one weekend.

• ChuckMcM · 2 hours ago

  Do you have a repo you could point people too? That would be interesting to see.

• 2arrs2ells · 2 hours ago

  Fable built me a version that covered all of my needs in an hour. (Clients pay me via ACH/wire - maybe it would have taken a day if I needed a stripe integration?). Runs on localhost with SQLite - incredibly basic, but I enjoy it twice as much as Harvest. I used my Harvest data export and generated PDF invoices as the majority of the “spec”. Feels silly to open source this, given how little effort it was… (and because I’m certainly not going to respond to others PRs) but I’m happy to if anyone wants.

  • syr3fx · 1 hour ago

    Same. Feels like there's no point of open source it. I mean, everyone can basically do it on their side.

  • Barbing · 42 minutes ago

    > but I’m happy to if anyone wants. Post for the planet? (power/water usage) An unregistered lurker somewhere is too lazy, shy, or scared to ask, but they’ll start prompting soon :)

• syr3fx · 1 hour ago

  Same, works great.

• BowBun · 1 hour ago

  Love this. Lots of software companies are sitting on no moat. Not talking about Harvest, but your JIRAs just jam a bunch of crap that most people don't need and then have to raise prices to justify it, before PE comes along. A software company's value is not just providing software, it's providing reliable software at a good price. The 'good price' for a ton of stuff has plummetted.

• wrs · 31 minutes ago

  Me too. I told Claude to use Beancount for the ledger (I don't want probabilistic accounting!), and very quickly I had a little Go/SQLite/React webapp and some Python scripts that do everything I needed. Since then I've added little tweaks as necessary. There's a time tracking page where I enter hours by project, and an invoice generation page where I select what to include and get a nice PDF invoice. The Python scripts import my bank statements and auto-reconcile expenses and invoice payments. They connect to Ramp, Expensify, and Gusto APIs for expenses, and maintain a consistent folder structure for reciepts. There's a script to give me monthly cash revenue figures for local taxes and let me know how much excess cash I have for owner distributions. It's exactly what I need and nothing I don't, has no recurring cost, and it's completely private (all runs on my laptop). Note, however, that I'm still using Gusto, because they take some responsibility for getting my payroll taxes right, and filing them.

The Harvest website lists the following (annually billed) plans: - Free (obviously free) - Teams $9 a month - Enterprise unlimited $14 a month The writer of this article claims they got upgraded to $133.98 a month plan, but looking at the site, such plan does not even exist. Something is a bit fishy here or they accidentally added an extra 3 to the $13.98 or mistook the yearly price for a monthly price.

• Aurornis · 2 hours ago

  You need to read the “Starting from” text on their pricing page right before those prices you see. You can have those lower monthly fees if you pay a higher percentage of your billings. The article acknowledged that. > There’s also a “flex” plan which costs $12.65/mo… but that plan now skims an additional 2% on top of the usual ~3% card transaction fees. The problems with the Harvest acquisition driving up prices has been discussed a lot among consultants and contractors. This isn’t one person reading their billing page wrong. This is happening to all of their customers. It’s how the acquiring company operates: Acquire companies, increase billing, accept that most customers will leave but some will stay and make up for the losses.

  • benatkin · 2 hours ago

    Indeed, if aphyr billed $5000 a month, just to use a conservative figure if it's a small but significant part of business, that would be $12.65/mo + $100/mo. Oh, that would actually be cheaper. That hike is a doozy. And I think I went way too conservative on the figure.

• benatkin · 2 hours ago

  I think it's including pay for usage. From the pricing page, beneath both Teams and Enterprise: > Your base rate includes core features. As your team grows, additional invoices, projects, clients, and tasks are billed based on what you use, so you're never overpaying. That's a bit bewildering, lol. Especially since they have annual plans. This clarifies it: > On Flex billing your first (sign-up) invoice usually covers seats only, because no projects, tasks, clients, or invoices have been tracked yet. Usage-based charges are calculated from the previous billing period, so they first appear at your renewal — not at sign-up. This is why a bill can jump (e.g. from a seats-only $44 to $404 once usage is added). https://support.getharvest.com/hc/en-us/articles/31447072608... That's roughly a 10x multiplier, just like aphyr's jump from $13.32 to $133.98! You're right something doesn't ad up, and seems fishy (to me, bizarre even). But that isn't the author, it's Harvest after the Bending Spoons aquisition.

• Digory · 1 hour ago

  Those are seat fees, which do not include the new usage fees. I got this email: > “Looking at your account, you're currently on the Basic plan at $40.00/month with 5 seats, and you're scheduled to renew … on the Enterprise Unlimited plan at $586.50/month…” So, from $40 to $586.50 for “unlimited” use. Same product. I think the goal was to force everyone onto usage based plans. I dropped spare seats and shifted off unlimited, and its about $75/mo per seat. But if someone wanted to take over angry Harvest customers, it would probably be a good little business. Guess I just need to rip the bandaid off, and let a prompt run overnight.

You could try plinx

Wow, harvest still exists? My first tech job ever pretended we were supposed to track our hours in harvest. Which of course no one did. So long ago.

We recently went through a version of this with Asana when we discovered their "per seat" pricing is sold in groups of 5 seats. Going from 5->6 users adjusts your bill to 5->10 seats, doubling your price. It isn't about the money, the principle is gross. Thankfully Asana's core business is now a prompt. We are nearly done replacing it with a native, self hosted client. Will post about it once it is hardened through use. Cancelling our Asana subscription this week. To SaaS providers (myself included) it has never been easier to leave your service due to mistreatment. Take note.

• tra3 · 2 hours ago

  > Asana's core business is now a prompt What do you mean? Like a claude code one shot prompt?

  • kxrm · 2 hours ago

    > one shot prompt Not OP, but definitely not a one shot prompt. However because of AI it takes far less resources to replicate services like these rather quickly.

    • tra3 · 1 hour ago

      That’s what I thought. I wonder if that’s true though. I’m sure 80% take no time but actually productizing it takes forever. Now you’ve got a maintenance burden too. I’m really curious how long it would take to implement harvest and the ecosystem around it these days.

      • joefitzgerald · 55 minutes ago

        It took me about 10 days to update an existing business app that I used to include functionality replacing harvest time tracking, invoicing (with a far superior QBO sync, saving us days of people time per month), and forecasting (Forecast was never well integrated with Harvest). Yes, we are picking up a maintenance burden but the 10x price increase in our case made it a very easy cost/benefit equation, and I decided that as a primary technology in a consulting business it was worth controlling it.

        • tra3 · 43 minutes ago

          How much time do you expect you’ll spend improving it? In other words, are you going to keep adding features? Trying to understand what the LLMs mean to the saas world. Will you have to become a time tracking/invoicing pro? Where’s the break even I wonder.

          • hdjrudni · 21 minutes ago

            If it's an internal app, you stop adding features when you have all the features you need. I've vibe-rebuilt several apps for myself and after a certain point I don't feel like anything is missing so I just stop touching it entirely except to fix a rare bug here and there.

      • gofreddygo · 22 minutes ago

        Its a problem but it isn't. Home grown product isn't going to match a saas product's features. But you dont need most of it anyway. so home grown gets the job done. Till it hits entropy or the roi is gone. Problems will show up a few months / years down till it isn't worth the effort or the savings. Saas products that survive this AI wave are going to come out stronger on the other end.

  • bredren · 6 minutes ago

    I think it is a figure of speech. I built a markdown editor and viewer to better than the best I've bought or used in the macos App Store in 81 prompts. I've built a very good ticketing system and it took a lot more prompts than that.

• 1123581321 · 2 hours ago

  So, you’ve spent a good amount of time on it, to where you hope there’s only a week left, it does a lot less, is less accessible, you can’t fully use it yet, and it’s buggy. I think the only lesson for SaaS providers, yourself and myself included, is how willing people are to cut off their noses to spite their faces. :)

• pwython · 2 hours ago

  Similar story here, except we went from Jira to a custom Plane.so build. Everyone from product to engineering loves it, gets features they want added quickly. And for me, I mainly have Claude keep track of everything (mcp).

• tartieret · 18 minutes ago

  same, I transitioned my teams from asana to linear this week, saving money for better product

Meanwhile in October... Berkshire Hathaway invests further $330M in Bending Spoons https://investors.bendingspoons.com/newsroom/berkshire-hatha... Just keep rewarding these guys....

• Centigonal · 1 hour ago

  Not saying I like their business model, but if they acquire a company, 10x revenue, lose half their customers, and keep margins the same... they've quintupled that company's profits.

I've had a Harvest account for over a decade, and it was always admittedly a little bit too cheap. But two months ago, they changed the pricing based on what you gross, so essentially putting a flat tax on my earnings. I knew it was going to be more expensive, but when I got the bill this month and saw the price was over $200 and then noticed the Bending Spoons LLC tagline at the bottom, that was the end. Claude is backing up all of my data via the API and will be vibe coding my next invoice.

I started monitoring/snapshotting the pricing pages of all bending spoon companies, which could become an interesting timeseries dataset.

• macintux · 2 hours ago

  Judging by the conversation elsewhere in the thread, the pricing page isn't nearly clear enough to capture the real impact. I'd wager that'll be true elsewhere as well. Example: https://news.ycombinator.com/item?id=50420733

Mercury bank handles invoicing quite well. If you're manually preparing invoices (as I assume OP is given the need to include hours), invoicing is free. Obviously payments are included (cards via Stripe). They have recently expanded into bookkeeping, but I haven't used that. If you're just using the bank account & manual invoices, it's free. Used Bonsai happily for years before that. I see they have been bought by Zoom, so I would expect some changes.

• epistasis · 1 hour ago

  I started with mercury recently and absolutely love it. The Books stuff seems to be OK, which is a huuuuge improvement over QuickBooks Online which is now an unwieldy pile of crap that nobody should use. As of Jan 1 Mercury says they will start charging $30/month, which is cheaper than other alternatives I've found. My vibecoded Beancount-based system probably cost me more in time than using a year of Mercury Books.

Hall of fame.

Bending Spoons did the same thing with Evernote over the last couple years. That seems to be their playbook.

Is this the company that bought AOL. lol are they just pe for millennials (or by millennials)

I loved there JavaScript library chosen last updated 13 years ago. Thanks y'all https://github.com/harvesthq/chosen

"Unavailable Due to the UK Online Safety Act"

• PenguinRevolver · 1 hour ago

  https://web.archive.org/web/20261109035326/https://aphyr.com... Also, hope your night is going well. I'm quite tired!

Bending Spoons recently bought MileIQ, the drive tracking, quick classification & team drive reporting app that we use for our vehicle logbooks for the Australian Fringe Benefits Tax record-keeping requirements. Going to have to find a new solution now! Mostly likely something OSS. I might even ask Claude/Codex to integrate that app’s background tracking features into our own staff iOS app, so I don’t need to rely on an underpaid OSS iOS dev for the App Store publishing!

I know a lot of people who have been laid off when Bending Spoons acquired their company so I'm a little miffed when I see this. I had to use Harvest when I did some consulting last year. I don't know the revenue numbers for Harvest. But, the company that asked me to use Harvest was a tiny software consultancy. I always assumed Harvest was something small companies used, not big ones. Those small companies would be very sensitive to price changes, and very motivated to use Claude to build their own alternative if it meant keeping $100 in the bank. I think this might be the moment when Bending Spoons business model is exposed. I cannot imagine large enterprises using Harvest en masse and those are the only companies who will stick with BS (pun not intended) when they run their playbook. Small orgs will certainly take five minutes with Claude. This isn't a SaaS that is really sticky.

It's a shame. They tried to increase our prices about 800%. I actually liked the software and used it for over 15 years across a couple of services businesses. They do have this "usage based pricing" where you pay them for each user, but then also for each project and customer, and then a percentage of your billings. It's completely absurd to be paying them a percentage when you receive payments by wire and all they're doing is emailing an invoice. Anyhow caveat emptor. Looking at their website they do not show the actual pricing. They just say it's "starting from" $9/seat or $14/seat, but you only pay that if you don't use the software at all. As far as I can tell, what Bending Spoons has figured out is that they can raise prices 8-10x, but only maybe 60-80% of their customers will cancel because it's a big hassle, and so the profits go up massively for a while. They will probably even retain a hard core of say 10% of their original userbase for some time.

I have never once heard anything positive about Bending Spoons. Sheesh.

RIP Freshbooks which was a simple way to invoice for projects.

Used Harvest for many years. Was talking to an agency owner friend of mine who had been using it, but once Bending Spoons jacked up the price they jumped for their own self-cooked version. The "SaaS is dead" narrative is a bit overblown, but private equity needs to think twice before gouging folks these days.

• biglyburrito · 1 hour ago

  " private equity needs to think twice before gouging folks these days " And yet, they never do.

• MBCook · 39 minutes ago

  Maybe PE needs some serious ball and chains added to it by the law to prevent playing robber baron with no consequences.

  • cpursley · 6 minutes ago

    How is that supposed to work when they are often the same people writing the laws?

Should try PotatoTimer! It's a bit different, I haven't built a stripe integration into it, but it's something I've been building for myself for my freelance work and for tracking my time learning piano, working out, etc. I haven't shared it around much (not fully finished) so any feedback would be welcome. It works offline, no account needed.

We just got vimeoed (a handful of videos with hundreds to low thousands of views per year) and a 10x to our bill. We were able to build a replacement in an afternoon with the features we use. I know less technical orgs arent quite there yet, but the moat on this stuff is surely evaporating quickly.

time tracking, payroll and invoicing software is ridiculously overpriced really eats into margins for what should be simple that being said, I can also see that the market can bear it there is still room for undercutting competitors or rolling it out yourself much easier now, but only for your specific use case

Yeah, my consultancy had used Harvest for over a decade. It just worked; it did what we needed, and I didn't think twice about the $1500/yr or so bill... Until this year when they dumped us over into "enterprise" and charged my business card TWENTY THOUSAND DOLLARS. I found that charge two days after it posted when I went in to remove a seat for a contractor who had moved on. I absolutely lost my shit. I couldn't get ahold of anyone at harvest at all, and wound up executing a charge back. Finally someone reached out via email with a "deal" charging us "only" $5000/yr (for literally the same service we had used prior). I get an increase of maybe even 10% but this was pure nonsense. Ultimately harvest refunded us and the chargeback was reversed, which came with its own set of problems. In response we have moved every SaaS we still use over to monthly billing. For time tracking, we're now using an internal vibe coded app, and it does exactly everything I need it to do. These fools really screwed themselves, IMO. I'm pretty sure I'm not the only person who would have just kept paying for and using the software indefinitely.

I'm not quite sure SaaS is dead but I can't see the price not being depressed by AI clones. Maybe hard to see "Workday" being replaced but I also never thought AI could one-shot all the stuff it does today. Especially because you end up using half the features, building integrations, and working around all the limitations in off the shelf software - the cost benefit to just clone and get exactly what you need is hard to deny.

• boie0025 · 1 hour ago

  Yeah I think a few things are going to bite some people in the ass. Odoo, for example, is "open source" until you need "enterprise" things. Well, it turns out you can point a robot at it and just add those things on your own now. Bonus for releasing your fork under the same license as the community edition... Strange times.

Odoo CE (community edition) is open source and has built in stripe integration (along with a few other payment providers) that will invoice on time spent against projects/tasks. If you’re in a region that has l10n for accounting (via the OCA), you’d be able to handle all your tax compliance also. Only caveat to Odoo CE is you’d need to self host, which in itself is not that difficult. odoo.com github.com/odoo/odoo github.com/oca

I got annoyed of all the crappy invoicing tools, so we built something custom for https://m9rx.com/ I've also had to build a bunch of small tools since we don't want to be locked into crappy cloud services. We also funded a small improvement into hledger, because we are trying to use a % of our income to vote with our feet. If people really want to use our invoicing and other small tools, we'll consider open sourcing them or selling them for cheap. Not sure if this is useful or a distraction.

We also just encountered the new pricing structure when we received the notice for our annual renewal. I am able to limit the impact to a roughly $200/30% increase due to the prior year basis, but with no changes to the pricing structure, I’ll jump another $2k when I renew next year. The big impact will be due to us trying to serve a larger number of small clients. Since pricing tracks clients/projects/tasks, our 2028 pricing will rise substantially as we add clients (or decide to track projects more granularly for existing clients). The move feels like enough of a price gouge that I have also decided to move away from harvest. With a year to plan and execute, I will likely build a solution into our client portal. It won’t take much effort to replace what we got from harvest, and there’s an opportunity to create something that further automates the rest of our billing and tax workflow. I don’t tend to wish ill for others, but I’m happy to see similar anecdotes in this case. My gut is that bending spoons has done lasting damage to the brand. I wasn’t considering a change before I encountered the new pricing. But now I can confidently say that there’s not much they could offer to keep me as a customer beyond 2027. The trust is broken, and they have vastly overestimated their value proposition.

an invoice can be a small gob of digital (or if printed, ink on paper) text. it can be emailed. it can be generated from a template by a tiny shell script. feels like an easily solved problem for many decades now. its hard only if we want it to be. and no AI needed, youngsters

Incredible story from the ones who own their platform and how they evolved during time.

Bending Spoons tried to 2x my Harvest bill on the usage-based plan and 146x (yes, that's not a percentage) on the unlimited plan. I switched to Clockify [1] and it works well. 1: https://clockify.me/

Incredible unforced error by Bending Spoons to en*tificate Aphyr himself. As usual, his analysis is clear, calm and compelling.

If you're a sole trader in Australia, Rounded ( https://rounded.com.au/ ) is the best.

why not vibe coding one to replace this service?

They killed the free plan I used so i vibed a replacement and haven’t looked back.

> Any of y’all freelancers have a system you like? I do all of my company accounts, including invoicing, via FreeAgent, which these days I have a Claude Code instance running via API key.

> but I don’t think I need to spend 2% of my gross revenue to send a couple PDFs a month and wrap a Stripe web form. The question still is: why a SaaS that generates a couple PDFs a month and a Stripe wrapper needed (I suppose) so many employees and VC funding? BS is reminding us all of those follies. It’s the company that, after calling out that the emperor is naked, gets to choose what he wears next.