SaaS Needs a Single Point of Purchase | Hacker News Reader

SaaS Needs a Single Point of Purchase

landshark.io

170 pointsalangibson168 comments

Sounds more like certain purchasing departments need to modernize, not the other way around. IMHO.

Not being able to use the right tools for the job, because some bean counters don't like complexity, is silly and shortsighted.

Yeah, telling your customers that they should modernize is a great way to let someone else sell to them.

OTOH, telling an entire industry to reorganize to accommodate your internal processes is also a fools errand. Just imagine the dilemma this puts on the growth team: "Hmm, spend 1000x effort to land this one customer or fix an obscure bug in the signup page and get an extra 1000 self-serve customers this month. Golly, I don't know..."

Seriously, though, GP is correct. Small companies that can adapt and snag these contracts won't push for change because they have a "competitive" advantage, and the businesses they sell to won't change because it's too painful for them. The whole industry isn't being asked to change, small B2B software houses will just keep ticking along, navigating the maze of purchasing to live high on the hog.

The industry has the choice of changing or ever racing ahead of AWS, who can take their potential customers merely by already being in the purchasing database.

If the means don't justify the ends, let my competition sell to them... Then it will be adding to their failure, not mine.

I do agree with your sentiment though, asking the customer to change is an uphill battle. I have customers with lots of custom billing situations, I loathe then all.

This is true, but truly the customer's own benefits are being compromised here. They can obtain better effectiveness at lower prices and push their margins, but procurement policy is not something companies want to mess with too much. It's also how they apply the same procurement policies to equipment purchases and software subscriptions. Some of those forms I don't even have a clue how to answer since it's so irrelevant.

I've never heard a better plan for failure. "Hey $10B company, I'm gonna need you to modernize your purchasing department before we go any further."

I don't think it's a plan for failure at all. I think it's entirely within Atlassian's rights to consider these customers to be more trouble than they're worth and pursue customers who don't have such braindead internal processes.

I guess it's intentional failure. I agree you don't have to accept every customer.

Imagine that the company is paying 25% extra because of these policies.

As someone who works in the enterprise space, I fully agree with you.

In every enterprise I've ever worked at, the purchasing process is a nightmare - it invariably takes months at best to purchase anything, and the bureaucracy, meetings and wasted time will cost you your soul. When you add up all the wasted man-hours it takes to purchase a $100 app, it will easily run to an order of magnitude more - at best.

Once you've been through the purchasing grinder once, you will do just about anything to avoid purchasing anything every again.

So no dammit, the vendors don't need to change - the bureaucracy-ladel, rule-ridden, time-wasting purchasing departments need to!

I agree with you in principle, but they're not going to reform because reform is in they eye of the beholder. For them everything is fine. They don't care how many gray hairs it costs you.

Given that, it seems to me wise to look for ways to make purchasing less nightmarish without needing to reform several thousand large bureaucracies.

Even in the "happy path" where vendors accept POs, 30-day credit terms etc, the purchasing process is still a nightmare, even if it's less so than for card payments. And it's not "just" about stressing out those that interact with the purchasing process - big organisations waste a lot of otherwise productive hours, affecting their bottom line.

When I consider the vast sums of money wasted by the purchasing bureaucracy where I work, it boggles the mind.

Everyone knows it, but inertia and fear of change are rampant in large organisations.

This is why any smart Saas company is charging, at minimum, thousands of dollars per month for medium-large sized customers.

They know the purchasing process will be the same headache regardless of whether the product costs $100/month or $10,000/month.

If the company is large and the tool enables collaboration and time-saving, $10,000 is literally a rounding error for a company with 500+ employees in comparison to the potential value delivered.

When I see B2B Saas companies with pricing plans that top out at $50/month, I want to pull my hair out. They are leaving so much money at the table and simultaneously starving themselves of the revenue needed to grow and build a long-term business.

So true. I wouldn't even try to sell an enterprise to-do app for less than 2000/month. The headaches of selling are baked in no matter what you are building.

I agree. Incredibly though, the same organisations will spend hundreds of man-hours quibbling over whether a dev team really needs a $20 app.

Sometimes I really hate working for large companies!

I’ve come to the realization as a leader that I just do not have the skin for budget discussions of any form or variety beyond negotiating with potential hires what they would like to be paid.

And what you just said about $20 apps is 100% why.

I think it’s probably if not the entire reason, then part of a factor of having less power and influence in the former (Though being expected to be answerable in full for “why did cost x go up?”), and measurably more in the latter-personal history has born this out with a predictable consistency.

Ask me to do the math, I will happily and eagerly get to work. Ask me to sit on a budget call and deal with the politics of who gets how many beans and I want to throw myself out a top floor window.

Me: "John wants to use this $20 app"

Manager: "Ok, schedule a meeting with the 3 of us and my boss to determine if it makes sense"

Me: "The 4 of us talking about this for a half hour will cost the company $375."

Manager: "..."

[cut to inside the $375 meeting]

Manager's Manager: "Things are tight, we just don't have the budget this year."

In reality that would only be the first of several meetings. It might be funny if it wasn't true.

I recall a couple of years back I was trying to purchase a POEditor subscription, which is $14/m. This was for a large dev project where the customer was paying us around $10M. After several meetings encompassing a ridiculous number of people, many of whom weren't related to the project in any way, I added up the costs and showed the PM: it was well into 5 figures, not accounting for loss of productive time for our customer. The PM just shrugged, "that's the way it works, they won't change".

If that is $20/year app, even if they finally approve it, expect to have this debate over and over again each year whenever the management gets another $20 invoice.

Then there is another piece of software that costs $200,000 but that's okay because someone more important than John decided they need it. Probably some website where you can upload PowerPoint presentations.

This is exactly why sensible companies have discretionary budgets at every level, ideally including individual employees. Below a certain threshold, it costs more to review than to just trust that people are asking for a reason; as long as there are controls and records in place to catch potential abuse or mistakes (e.g. paying for the same thing multiple times), approval requirements should be proportional to amount. On the off chance someone is being frivolous, it's still not worth imposing painful review processes on everyone.

An individual employee at a technical company should be able to spend $100-200/year with discretion (mouse, keyboard, utility program, cheap service, etc). (That's not counting whatever the company might allocate for people to get their own development systems, which some companies do.) A first-line manager should be able to approve more than that (e.g. a monitor, a more substantive service).

Ofc you can buy it from the tiny discretionary budget. But to get to officially use it you will still have to get approval from IT, a process that will make the purchase shenanigans pale in comparison.

So every department builds up a hidden account and 'shadow IT'...

And atop each 'shadow IT' dept, is a 'shadow CIO'.

This is so true. I've seen some companies however who allow end user teams to spend a certain budget using their corporate credit cards. But nowadays IT security and privacy restrictions anyway make every damn purchase a 8 month project at least

I'm fortunate to work in an "enterprise" environment where managers have discretionary spending ability. A $50/month app doesn't require the huge purchasing process, just throw it on a company card and tell accounting what code to charge it to. Once you go over a certain threshold, the heavyweight purchasing process kicks in - but it's more justified when the dollar amounts are higher.

What's your company? I want to sell to you.

Unfortunately legal concerns about data privacy are pushing enterprises towards MORE of this stuff, not less.

Absolutely. And it is strictly security theater. We're delusional to think that checking the boxes on these arbitrary lists is making us "more secure." It's very frustrating to have a role in this game and find yourself completely unable to combat the bureaucracy.

This is so true. And big corporations push around smaller vendors to have so many compliances, insurance and other certifications in place, whereas larger vendors can get away with insane security and privacy lapses just because of legacy. Makes me mad.

Honestly, as someone also currently inside a v. large bureaucracy, I concur. Sometimes it almost feels as if half of the "value" or "work" in selling to big businesses that I hear of is in navigating the purchasing departments. Once you've done that, you're golden. It's like picking the lock on a safe.

That's pretty much not going to happen. That said, what you can do is:

- Provide a plan that fit within common purchasing card limits/tiers, to enable those who want to bypass the internal controls of their purchasing department and just slip it in on an expense report and be done with it.

- Provide a plan that adds several zeros, to make it worth the hassle when you get a customer that does try to go through the official purchasing department process. This will entail contract reviews, security/compliance certifications/audits, invoicing hassles, ongoing support requests (you will get lots of emails, even for things that could trivially be self-serviced by the client), etc. You'll be sinking lots of labor into this process regardless of the price tag, so make it worth it.

"common purchasing card limits/tiers"

Strongly agree with you although what "common" means here is tricky.

I imagine "common" in Local Government is different from tech giants ... and geography brings another whole dimension into it.

Less of a problem if your product is strongly focused on one country and one industry.

"Provide a plan that adds several zeros"

I think this is ideal but I've noticed that this isn't generally how it's done. The third tier of pricing is generally "Talk to us" and the first part of "Talk to us" is the supplier working out how much they can get and so requires the sort of sales management skills which many tech startups don't have.

In passing there's a lot of sense about pricing in https://www.joelonsoftware.com/2004/12/15/camels-and-rubber-...

> Strongly agree with you although what "common" means here is tricky.

Ha, true! I've never seen it done, but someone should really A/B test adding a "Expensing this product for work? Reach out for a custom payment plan that fits your needs!" message if their product is popular with a business crowd.

> is the supplier working out how much they can get and so requires the sort of sales management skills which many tech startups don't have.

So this is sometimes the case, but not always.

Is the standard plan enough for you, but you require us to invoice your AP department for payment? Or you a second login for IT with access to the login/audit logs, but otherwise don't need any "enterprise" features?

You can likely get it close to the listed rate, with a minor premium for the additional effort.

Will you be a high volume user, but capable of being self-sufficient with the self-service tooling? You can get tiered pricing based off of usage, no account manager, and a fixed rate for engineering time if you need any custom support.

Will you be a massive pain - requiring an in-depth security audit, require us to agree to potentially costly-to-comply-with contractual clauses, need an account person to do every little thing for you (plus constantly (re)train your staff/agency's staff), require a generous amount of engineering time baked into the contract, and split out the agreement between a base MSA and incremental SOWs for each actual usage internally? Your pricing is no longer based on the actual product itself, but on the uncertain but potentially massive amount of labor that's going to be required to support your account.

So "Talk to us" is as much the supplier trying to figure out specifically what sort of customer you are and what your needs are (and how much you'll ultimately going to cost them), as it is working out the pricing to go with that. Which is why you don't generally see an actual listed rate with several zeros added, because that is honestly only one of the potential outcomes and "Talk to us" leaves open capturing the other outcomes as well.

I've also had success using the "Talk to us" to get lower than list rates. Sometimes because my needs are such a tiny fraction of the intended usage of the plan with a list rate. Sometimes because I work for a company that does $10bn a year in revenue, and the company is hopeful that my usage of the tool will potentially domino into other users internally seeking it out. You're right in that many tech startups don't have a high level of sales management skills from the get go, but only the very last example above really requires true sales effort. The rest are just very low pressure sales needs that require minimal if any incremental effort to support, but just happen to not fit cleanly in the pre-defined plan templates you have listed.

Our app costs $18 per month and one of our customers said she could buy only using a PO. So, I told her to use it for free till she finds a way to pay by card. We are likely to gain more by having an enthusiastic user, rather than losing her for the stupid purchasing department.

Could you just set up a recurring invoice via Stripe so that she doesn't feel like a freeloader?

She said she needs a PO.

So sell her an enterprise subscription for $2500/year, with up to 10 logins, and take the PO. Now you have a year to figure out how to make that routine.

Eventually, we should be able to do that. But, that would get into a long approval process right now.

Right now, you're getting 0 from the sale. You can afford to let them take months to pay you.

We have 800+ small businesses as customers. This one is from a large company who don't come in as inbound lead. Don't you think it is better to understand how to work with them rather than hitting a wall with their purchasing department?

Or sell her a lifetime plan of size x that justifies the hassle of dealing with a PO.

i pay for a $20/mth saas from my own cc n claim afterwards because they dont issue POs and finance doesnt have a cc.

one day I'll be late with the payments a few times so that we can start the discussion with some other vendor that will charge >100x but can issue POs.

This is so true. The sheer administrative/legal effort and expense it takes smaller enterprise software vendors to execute a single sale is crazy. I'd happily offer lower prices if the process was easier.

I own a SaaS and 99% of our customers are huge international corporations. They only buy with PO, and every sale is a massive bureaucratic effort. I hired a freelance account manager to deal with incredible requirements like: - anti money laundering declarations - mandatory company stamp + hand written signatures - supplier enrollment forms (excel+macros) - notary apostilles for incorporation certificates sent across to Russia

Software resellers do this as a job, and I thought they are stupid middlemen nobody loves. Actually they are amazing.

Out of curiosity, what kind of SaaS is it?

It's not strictly a SaaS, but it's a subscription product based on an open core.

https://readonlyrest.com

PO2$$aaS

What does PO mean?

Purchase Order: https://en.wikipedia.org/wiki/Purchase_order

More important than the PO form is the entire rest of the corporate purchasing process it entails. You send them a quote, they try haggle you down, they might make their internal team send the bid out to multiple vendors for quotes and try to play you all off against each other (and how well you meet the requirements for the actual end user is not a a part of this process, just their checklist which is three degrees removed) they decide what quantities they want to buy, you confirm the price, send an invoice, they'll pay net 30 or net 90 or whatever their purchasing department decides via whatever payment method their department decides and you'll do it because with the bulk order amount and the markup you'll decide is justified for this hassle, the money is too much to refuse.

It’s called as master service agreement (MSA). Here is an example: https://www.contactually.com/platform/enterprise-master-serv...

The steps to complete a transaction at any Fortune 500 company are usually: NDA, proof of concept, MSA or short form MSA, PO.

This is a pretty established distribution model often called "bottom up" approach, and it's particularly relevant to enterprise sales where you allow the users to start using the software without friction [something that was proper to "consumer" applications] and then allow management to pay for that software, or users to pull their corporate credit card and buy what their budget allows.

One turn-off for us for machine learning platforms was going to the website, wanting to try the software, and then finding a button that says "Schedule a demo" or "Schedule a sales call". I want to try the software, not talk with your salespeople. I want to create an account, or request access to a demo account without speaking with someone. Then I want to be able to pay for it without speaking to someone. I have the proverbial, and sometimes literal, credit card in my hand: let me pay.

That weighed on our decision to make our own platform[0], because we simply don't like to buy things this way.

Again, there are differences with low-touch and high-touch software, and situations where the regulatory context or contract amounts require that at some point, people sit down together, but that's for particular cases and clients. We made custom machine learning products for large enterprise, so we sit a lot with clients, but the word "custom" is there.

Some content that goes into this is Slack's CEO discussing this with Yammer's former CEO[1]. Slack's CEO also discussed this in a Pando Monthly[2]: they didn't have a "sales" team, but sometimes needed to 'onboard' a large organization. He also addresses how raising $1B reassured many 'traditional' and 'sensitive' organizations into using Slack for internal communications.

- [0]: https://iko.ai

- [1]: https://www.youtube.com/watch?v=4NbXBGjUrA0

- [2]: https://www.youtube.com/watch?v=TbM1iOi33NY

>One turn-off for us for machine learning products was going to the website, wanting to try the software, and then finding a button that says "Schedule a demo" or "Schedule a sales call".

Share the sentiment but both of us are likely not the target audience for these kinds of funnels then. Decision-makers who go down the "talk to our sales team" route are more likely to sign big enterprise-y contracts in the end I guess

>Share the sentiment but both of us are likely not the target audience for these kinds of funnels then.

Except I'm an engineer first who just happens to be able to make that decision, and I made the decision not to buy.

Here's the thing: we've been making bespoke machine learning products for large enterprise for quite long. Not huge contracts, but say mid six-figures on average. We sit down with our clients but the key word here is "bespoke", which warrants these meetings and communications and calls. We build these solutions from scratch to solve their problems with their workflows. Anything that's not bespoke or where we don't have to discuss regulatory matters or where that conversation is not required by law is a roadblock for my purchase.

One reason we're making our platform is to reduce that necessity and be able to deliver that value to companies without ever speaking to us. Heck that "speaking" is also included in the cost of the product, and all those "engineer.days" add up pretty quickly, making our client list exclusively composed of behemoths who sometimes happen to have internal ML teams but just need more help, or because we happen to have expertise on a problem and they're inexperienced despite their size. One of our biggest frustrations is that we could not serve smaller organizations, and I'm not even talking mom & pop's or SMBs, but even some organizations that seem large shy away from this kind of projects even though they want to leverage data.

I find that these companies and products are targeting decision makers who do not consult with their teams and buy products nobody will use. They're targeting dysfunctional organizations which, I'll grant you that, may be a large enough segment to build a business around. Not my ideal to wake up and earn a living, though, but that's another matter.

The interesting thing here is that until I say "go" no one will "Schedule a demo" or "sales call" to them. While I will absolutely not do it for a fishy firm - and by putting something like that on the site they automatically get into "fishy firm position" - you would like to convince me with words instead of facts to buy your products and waste my time. No, thank you.

Show me the facts, and leave my mind to decide what it is, not lie to me what you think will make the sale. It is just a waste of my time.

This might be true but there are many engineers with decision making power. My pet hate (and a virtual guarantee of "no deal") is when I'm forced to talk to a sales person who is all hair-gel and swagger [1] who knows substantially less about their own product than I do.

[1]: https://gist.github.com/CumpsD/696599d1bd4cd472a056586967293... - this is about (UK) external recruitment companies, but the same applies.

In my experience the reason this is done is because customer acquisition is manual and the "schedule a demo" button is used for lead generation, and the reason it's all manual is because onboarding/support is high touch and you're not sure who the customers are yet.

Free demos are my nightmare. I really don't have that much time to split between fixing existing issues, adding critical features for paying customers, and supporting or troubleshooting for users that aren't paying anything. That's not to say at a point in the future the organization could have the capacity for a sick free trial, it's just that free trials cost money.

I dont like talking to sales people but I get where they are coming from. It might also be using the opportunity to do some discovery as to what people are using the platform for and who are those people.

But I agree that they should make provisions for people who just want to pay and get to business.

>It might also be using the opportunity to do some discovery as to what people are using the platform for and who are those people

The way we do that is giving people access to the platform, letting some time go by, then engaging the users and asking questions. We also have Slack channels in case they have problems.

Why we leave them alone: because I don't like it when someone immediately engages us asking us a bunch of questions when we start using something, or try and up-sell right when I'm discovering the product. I also don't like energetic chatbots popping up on different tabs of the same website, making all kinds of sounds wanting to get my attention to help me. Back off, I'll close that site and will not visit it again.

We give people time so we could at least observe what's broken. Did the person not start using the product immediately? Why? Was the user unsuccessful accomplishing a key task with a key feature of the product? Why? Is it a problem of discoverability, or is it just that it is not our target user? How can we improve that either way?

They might also want to massively price segment. Basically they don't want to quote you a price until they've got a feel for where your pain point is. If you're a whale, they want to reserve the right to bleed you dry by not preemptively quoting a lower sticker price way below your budget. But yeah, to me it still just screams "if you have to ask, you can't afford it".

>But yeah, to me it still just screams "if you have to ask, you can't afford it".

We're talking about buying product: I have at least to ensure the product does what I need it to do before even considering buying the product, and I can do that on my own. We're not talking about selling airplanes here, where there might be some setting up required for me to test the specs, neither are we talking about software targeted to governmental agencies or the intelligence community. We're talking about products that claim they do X for the knowledge worker, but fail to provide a way for a knowledge worker to test X.

This can be turned to: "If you have to talk to me before I could use or buy the product, something's fishy".

That's Theranos/Nikola hyping the product with claims it can do things but nobody ever saw it do that thing.

I get the price segmentation argument, though there may be better ways to extract more value by providing tailored services to large clients to solve problems only large clients have.

If it's an important service it's not a big deal at all to speak to someone, and frankly, even though they do this for price discrimination, the fact is they can incredibly helpful and save you a boat load of time and effort trying to understand the product in many cases.

Yes, some things should 'just be self serve' but many things, not so much.

>the fact is they can incredibly helpful and save you a boat load of time and effort trying to understand the product in many cases.

I take having to explain what our product does to users as a failure on our part. We have the same mentality when we have to personally explain what code does to our colleagues. We take the necessary measures of using better abstractions, writing documentation and Wikis, simplifying code, and making efforts so that "failure demand" does not arise.

In terms of product, it's either our documentation sucks, or a problem of "affordances and signifiers", or a problem of targeting the "wrong" users.

Well the job they have to do before I waste my time scheduling a meeting with your sales dude is 1) convince me that your product is valuable and 2) convince me that the ballpark pricing is workable for me.

You can really only know a car by test driving it. But that doesn't meant that Toyota doesn't publish photos of its interiors and asks you to schedule a test drive before finding out the price. I can quickly find out what cars might be suitable for my needs given their properties and their pricing, and THEN commit my time to test driving the best candidates. If you just say "Believe me. This is the best car. People are saying what a great car this is. This is a car that you'll come in and you'll test drive, and you'll say that you've never even seen a car this good." Well I'm just going to ignore you.

My main point is that this bottom up approach doesn't work as well as it could simply because of the difficulty of navigating purchasing processes. I know my team would have bought easily 5x what we did if we didn't have to invest 4 to 6 months in getting a vendor set up for each SaaS we wanted. It just wasn't worth the hassle.

>My main point is that this bottom up approach doesn't work as well as it could simply because of the difficulty of navigating purchasing processes.

I may have been unclear. The "bottom up" approach is precisely designed to circumvent and solve the case where the user is not necessarily the buyer. It emerged to solve the "shrink-wrap software", or when you did not have an army of salespeople.

One way to look at it is "bringing SaaS to enterprise", which was typically the distribution model for consumer applications.

Not to say this would have worked at your company, and not to say it's perfect, but I think the "HACK" of the bottom-up approach for software like JIRA or others is that it is priced in a way where you can get your team at a company using it and showing demonstrable value from using it on a free or very low-cost (AKA your personal credit card) tier.

Then, when you go to make it official, you have much more power in the negotiation.

If it's working well, making the "shadow IT" SaaS tool an official purchase typically goes:

Team Lead: Everyone is using this tool, we're about to go above the free tier and need to purchase it.

IT/Finance: You should have gone through the process, you need to fill out these forms.

VP/CEO (to IT/Finance, seeing value of tool): Just get it done

Exactly. Then taking it a step further. GP at VC firm makes the round at their "portfolio" companies and see your product on every screen, or hearing about your product from every team they go to, and then wanting to speak with you because clearly someone is doing something right.

Examples: YouTube, GitHub.

I think maybe I wasnt clear. No one tried to keep us from licensing anything. Pretty much any solution was fine with then assuming it wasn't a security nightmare.

> Just get it done

Getting it done is a process heavy, minimum 4 month process because of purchasing rules. That's the problem I'm calling out.

>I think maybe I wasnt clear. No one tried to keep us from licensing anything.

No need for someone to do that. The process and distribution method kept you from that.

Whether it's IT or Finance (purchasing rules), CEO/VP have the power to force them to make an exception or change their process if there's a clear business benefit.

I'm talking about 10K+ global employee companies. No VP wants to listen to my problem licensing Jira. And prices changes like that happen on the scale of years.

I don't really agree. I've sold software both independently and by hooking into the AWS marketplace to reuse existing billing channels. It really didn't make anything better even though using AWS as a single point of purchase sounds like a good idea. The enterprise software purchasing process is just a slow one and the actual billing mechanics aren't really the issue in my experience.

Plus delegating control of your payment relationship with your customers to a third party is more of a pain than a boon.

My CFO didn't care about the line items inside our $20,000/mo Heroku bill. What he did care about its size. When I told him we could save a lot of money by fostering a direct relationship with companies like Redis Labs, he was very receptive to making that happen.

My engineers could spend money on new pieces of third-party infrastructure pretty easily via Heroku plug-ins. If the spend got big we figured out how to address it.

For this dynamic to work for you as SaaS developer, you need enough runway to wait for your users to adopt your tech and have some fraction organically grow big enough to be real money.

This rhymes with how Intuit won the accounting market. They initially got destroyed in reviews when compared to their more mature yet complicated competition, but everyone starting a business wanted cheap and simple. You just need to be willing to wait for enough of your early garage-based startup customers to become S&P 500 companies. While you're waiting, you're also evolving your product so that it better meets your growing customers' needs.

Yeah, I've done enough fine dining in my professional career to know that "schedule a ..." means a low value product pumped up by dedicated sales.

Even if an actual product value exists after accounting for spin, the gated access means high training cost which in turn means vendor lock-in.

"Schedule a ..." ticks all the wrong boxes. Except for dinner, dinner is often quite good. In fact maybe they would get more traction with "schedule dinner."

An addition to all the comments in this thread, from a finance point of view, consolidating the tax treatment of multiple resold items would be nightmarish if you want to launch this service internationally. Additionally, the mismatch of billing structures would lead to the hypothetical reseller either keeping margins (big turn-off for onboarding I bet) or keeping wildly volatile short-term liabilities on their balance.

It just seems like the author is looking for a band-aid instead of the real problem; a heavily bureaucratic procurement system in their company.

It's not my company, it's every large company I've ever worked for.

I'm not looking for a band aid. I'm looking for a solution that helps the SaaS industry without demanding that the Fortune 500 get their act together, because they ain't changing for our benefit any time soon.

Can you expand on

> Additionally, the mismatch of billing structures would lead to the hypothetical reseller either keeping margins (big turn-off for onboarding I bet) or keeping wildly volatile short-term liabilities on their balance.

In Sweden we also have something similar to purchase orders which is common even with B2C and it always bothered me how often it is used compared to the US where credit card was more normal. I think the article brings up a good point, if a startup could figure out the taxes and bookkeeping, set some reasonable margins or costs for the service it could be a really good business.

Edit: Sorry, they would handle the POs so that is liability, I guess some kind of verification or legal framework is also needed here.

>In Sweden we also have something similar to purchase orders which is common even with B2C and it always bothered me how often it is used compared to the US where credit card was more normal.

Our clients send us purchase orders in Europe, but the context is custom, turn-key, machine learning products developer for large enterprise. We take them by the hand from "I have many problems" to a product their people want to use.

We're building our machine learning platform to make that product development faster (data, scheduling training notebooks, automatic tracking, etc.). We also adopted a plugin architecture for product so we can plug functionality and, more importantly, remove functionality to allow clients to discard what they don't need and avoid a product with 2000 buttons and menu items.

That's a shame, but it seems like it's enabling resellers to charge 25% margins. I don't see those margins being competed out.

Imagine you have services that the end-customer has subscribed for. Some of them are billed weekly, some monthly. Some of the monthly ones are billed the middle of the month based on usage, and some at the week following the month's end based on a mix of monthly subscription and overage.

When a specific invoice that has the potential to vary greatly from period-to-period, you might encounter cases where you have to pay for an invoice that was twice the amount of last month, but you will only be billing your end-customer 20+ days from now! This means that you have to either foot the bill, or wait a little bit and keep the amount in your balance sheet.

When you scale up, you are essentially providing a financing product. You can counter this by asking the users to deposit an amount equaling the base subscription or other service fee, but you can still run into cases where you might simply be cash-strapped to pay your clients' bills; this would put the hypothetical reseller into essentially what's a bank run. Not fun.

It just seems like the author is looking for a band-aid instead of the real problem; a heavily bureaucratic procurement system in their company.

Sure, but as someone who works at that sort of company, we happily give lots of real money to companies that solve that problem for us. As recently as a few month ago I wanted to spend 1-3 thousand dollars pr year on a SaaS product. The problem internally was never exactly how many thousands I was spending, but what the payment process looked like. And the sole reason I didn't give that money to the SaaS company was because they couldn't work with our procurement process. We're now investigating a much more expensive solution from a much bigger company.

I'm sure the SaaS company turning down our money was being rational and figure the cost of taking our money was higher than the profit they'd make from us, but there still seems to be money left on the table for someone to scoop up.

I bet they were rational, but somehow I can't imagine the lifetime value of you as their customer would overcome bending a little bit to suit your procurement needs.

Odd.

It's not a little bit. The process demands of a large company can suck the life out of a small one pretty easy. I've seen it happen.

> And the sole reason I didn't give that money to the SaaS company was because they couldn't work with our procurement process. We're now investigating a much more expensive solution from a much bigger company.

Bingo. My guess is that it may be efficient for many small companies to offload the burden of dealing with procurement processes to a single specialized entity.

I was CTO at a company that got bought by an enormous company with these sorts of procurement processes. I mused with my engineers multiple times about a company that would simply do billing process arbitrage.

You've made me wonder if there's some way to arbitrage the billing itself so that neither party needs to pay anything.

The biggest obstacle to this is not necessarily technology, but the vastly different ways SaaS companies opt to go to market via reseller channels - or don't. Google and Microsoft have both had established reseller channels for a long time, but there are still major vendors who only recently (or still haven't..) established this.

Slack is a great example - they only sell direct. And companies who use Slack have a tendency to "sprawl" into other SaaS apps that may not have a reseller channel, creating a bit of a snowball effect. Contrast this with companies "bought in" to Microsoft where it's much more common to see them adopt Microsoft Teams and other technologies to avoid that sprawl - and they typically buy via Microsoft partners who can bill them on a single invoice.

While we can't force SaaS vendors to acknowledge that customers want a single bill/distributor, we are working on the problem further upstream by creating a marketplace [1] backed by a network of connectivity agents who are expanding into the SaaS world. One point of purchase/bill for connectivity + SaaS (and IaaS) with additional benefits such as management, etc. Even more interestingly, we're covering the developer, distributor, reseller, and customer channels on a single platform.

[1] https://www.appsmart.com

Do you have a salesforce?

There’s merit here but one challenge is that such a service is just another revolving line of credit for the company. Another credit card. Finance would likely want it locked down as well.

Ive had a lot of success shaving hundreds of thousands off aws budgets and my first tactic is always to create mental awareness that every terraform apply is a virtual swipe of the company credit card.

In big companies, making it hard for non-executives to spend money is a feature not a bug.

That's not a tough but to crack. It's pretty straightforward to build a feature that sets limits by team, business unit, etc.

Budgets are important but mindful spending is even more important.

He complains about the reseller and their incompetence, but (incompetence aside) this is basically what he's looking for?

I know it's one of those things that tends to change over time, particularly if your account rep changes, but wouldn't simply finding a reseller that doesn't suck be the answer here? Perhaps one that lets you automate quotes and purchases...

Nope. Resellers are very hands on and high touch. I'm talking about something more like a pass-through payment processor than a reseller.

Quotes in this case are as trivial as a PDF generator that just inserts the fixed price into a commercial offer template. As it stands, you wouldn't believe how hard it is to pry something like that out of a typical SaaS vendor.

I literally wrote “I’m begging on my knees” to a LastPass sales person recently.

I’ll make sure to have another solution in place before next year’s renewal.

> “I’m begging on my knees”

I feel your pain

I'm a maker of satchel.com. This is one of the core problems we aimed to tackle. In doing so, we've actually build out _exactly_ this.

There are still some crucial obstacles to ensure robustness, but I'd love to get your feedback nonetheless. Could I ask some questions over email? I can't find yours, but I'm at andrew@satchel.com

Sure, I'll email you.

(And thanks for reminding me that I need to add my email to the site.)

I feel like the main problem here is an overly bureaucratic and suffocating organization. Say what you want about change being slow, but more nimble competitors will eventually eat their lunch.

So don't be the the subject of this story.

There are many, many huge corporations that aren't going away and aren't changing any time soon. In the mean time, the SaaS industry is missing out on billions if I'm right.

"Overly bureaucratic and suffocating organizations" will always find someone who wants their money. A place I used to work was happy to pay SaaS company A 4 times more than SaaS company B was charging for exactly the same service, purely because company A could send quarterly invoices and B couldn't.

"Overly bureaucratic and suffocating organizations" might be a pain to deal with, but they are also fairly price insensitive once they've decided that they want to buy your service, assuming you make it easy for them to give you their money.

Absolutely correct. If you are trying to sell to a Big Co, price should probably be about the 10th thing on your list to talk about.

As a director at successful startup I see how we went from being able to buy new software in minutes by pulling credit card to multi stage procurement process of public company.

All changes got good intention: 1. We need to be secure. Know what kind of data there.

2. We need check and balances for finance.

3. We need to be compliant.

Etc.

I would love if there would be open consortium that could provide a lot of services, without price overhead that also establish some quality/security/compliance bar for its members.

Companies as a broker model is flawed. Procurement won't accept paying 20%+ extra for same stuff. Terrible lock-in.

In healthcare this is referred to as the Group Purchasing Organization (GPO) model.

Instead of waiting for sales orgs to band together, purchasers band together to negotiate and vet products and services.

This is a much better model, in my opinion, because it keeps the incentives on the side of the buyers, not the sellers.

Keep in mind, certain products still have to be vetted by the individual purchasers due to statutory requirements. But, as a buyer you narrow the scope of offerings you have to consider and you increase the likelihood that your in-house audit will pass. So, there's still a benefit even if you absolutely must inspect their facilities yourself.

What also happens in healthcare is the manufacturer raises prices to account for margins of/lost to the GPOs. Individual buyers become required to buy through GPO as the price differences are stark. As GPO gains negotiating power, they want to keep more spread for themselves. Manufacturers adjust. The cycle continues, all the while the ecosystem becomes more expensive since you inserted yet another middle man.

You can't do a traditional quote, PO, invoice, receipt purchase process when you're selling $99/mo. It just doesn't work.

That process makes sense when you're buying and shipping pallets of physical inventory for 5-6 figures. It makes no sense in our SaaS world today.

True story: a previous employer would not bother issuing POs for less thank 2000 euro, so we couldn't even think about buying anything that was too cheap

Depending on how it’s implemented I could see this being a very good thing. My public sector clients esp. have a hard time buying any Saas as each one involves finding vendors, getting it approved, setting up payment, etc etc etc. obv they have a whole procurement department for that, but I could see their job being easier if all their aws, azure, and smaller saas came on one bill. The organization would still need to handle compliance and be sure POs were valid, fair pricing, no conflicts, etc. but do that against one vendor instead of 100. It would be difficult to get right though. Balancing buyer and seller interests while trying to get a cut. In fact the same org I’m thinking about was recently forced to do all purchasing through a 3rd party vendor and it slowed down and complicated purchasing even more...

> My public sector clients esp. have a hard time buying any Saas

Indeed. It's probably even harder for then than it was for us.

> Balancing buyer and seller interests while trying to get a cut.

This is what made me jump to the idea of a non-profit consortium set up by SaaS vendors with a mandate to grow the market by reducing friction in purchasing

> My public sector clients esp. have a hard time buying any Saas

I worked fulltime at my state college while I was also a fulltime student there. The amount of process to do anything was staggering. Buying something like a replacement hard drive required "quotes" from three different vendors (I printed pricing from 3 different websites). Forget about trying to buy 5 or 6 figure server equipment.

> Balancing buyer and seller interests while trying to get a cut.

You're essentially floating them a short term loan. You pay the provider immediately with your credit card and then sending the buyer a PO. Give procurement a portal to see the bills as they are from the sellers and your PO has an added line item charge (flat fee or preferably a percentage of original bill) for the service.

Depending on risk level of the institution, could probably require an upfront deposit to make sure you don't get burned.

Nice idea - but ...

His argument is "I could buy any SaaS service I liked by simply clicking 'Subscribe to Foo-SaaS for 99 / mth' on my AWS dashboard."

This is true - it would solve his needs immediately and would yes lead to thousands of sales.

For about a week. Then every purchasing dept in the Fortune 500 would collectively spray their coffee over their monitors, and suddenly one of two things would happen a) AWS would offer Enterprises a "turn off that Sht" option (at a cost) or b) every purchasing department would add even more hurdles* before you could pay out.

Yes, it is hard to buy stuff in a corporate - but it is not a bug its a feature.

I half disagree. I'm not trying to avoid financial controls or any of that. What most people that don't do this stuff don't realize is that just working the process to i.e. create vendors in your ERP are incredibly time consuming. This is even if you have full support of the business. We always had 100% support from everyone and it was still a nightmare.

I quarter agree :-)

I have been through (many) purchase processes and they have always been painful. Especially true the first time you try and make a purchase at any large org.

I call this "Director Hazing". Being promoted to Director (US company designation) usually involves a year of administrative hell as you have to learn all the horrific systems that have grown up in the company over time. There is a sort of "sink or swim" mentality and once a director gets through it they are selected for being the sort of person who can operate this system (ie responds to emails quickly) - and that tends to work against promoting people who will fix it.

Anyway, as I say I reckon 25% of the time management thinks it should fix the problem and 75% of the time "I can work the system and so can anyone else at a certain level - everyone else has not been promoted and so should not get to spend the cash"

I still remain convinced the system works as intended.

A side note - there was some study I read about a major supermarket chain that used cameras and POS data to rank their checkout workers - and a small revolt from store managers killed the project. The store managers knew who performed best and worst at that part of the job - but that ranking did not fit the ranking for the other intangible parts of the job. And they wanted the flexibility to decide who to fire / keep / promote. And a clear metric would have got in the way.

Similarly management can if they want make sure your vendor is set up, your licenses paid for. If it is high enough up their priority tree. And if it is not having a widely hated terrible system is a good excuse to avoid saying no to subordinates...

Oh come on! Big companies! and Atlassian! Being hard to do business with is so 1980s.

FWIW, Stripe.com supports ACH (Automated Clearing House electronic funds transfer in the US) as well as payment cards. So a SaaS can take that kind of payment too, without lots of hassle. I haven't investigated Stripe's worthy competitors' ACH offerings, but they surely have them too.

And, big company purchasing departments can definitely get their hands on suitable payment cards for this purpose. They can guard them closely, and don't have to hand them out to random people.

This problem can be solved. (Well, it might take a decade or two...)

It's not that can't set up a credit card or do ACH, it's that they won't. The reasons usually have to do with financial controlling rules that I can't begin to fathom.

It looks like Atlassian does take POs though. Is there something about the way they (or other SaaS) companies take POs?

How do I submit a Purchase Order (PO number) to Atlassian? (https://www.atlassian.com/licensing/purchase-licensing#order...)

To get started, complete our secure online Quote & Order Form. On the final confirmation page, check the PO number box to add your PO number. The PO number will appear on your quote or invoice (valid for 30-days). If you already have a quote or invoice, you can update the quote or invoice through my.atlassian.com via the Orders tab.

Atlassian is happy to reference a PO number on a quote or invoice for your internal tracking and record keeping. However, we do not accept purchase orders as a form of payment nor the terms and conditions commonly associated with purchase orders. We provide fully functional trial licenses for you to use while payment is being arranged. We are able to keep our prices low by offering standard product agreements to all our customers and do not offer commercial credit.

The blog post describes only the billing advantages of the "single vendor" approach like AWS or Microsoft. But the other half of it is much more important, and that's the security/compliance aspect. You know that anything you buy from AWS or Microsoft is engineered to a common set of security / privacy standards, and that whatever those standards are have been audited by governmental customers etc. Of course they're not perfect, but there is a definite increased security risk in having your corporate data sit inside 20 different companies vs. 1.

It’s not just a single point of purchase, big co need to ensure your subcontractors subcontractors subcontractors enforce our policies and control around protection of data, we need to be able to Inspect that whenever our auditors feel like it. The complex issue with SaaS for big co is getting them to demonstrate a good business continuity plan and actually execute tests (your primary building was destroyed ... you own another building with connectivity , hardware just sitting around on ice ready to go right ? ) You have key man dependencies please asses that every 6 months and hire 35 people to cover that 1 KMD and show us your working on how you assess this. If your SaaS is so killer / think it’s a big competitive advantage big co will help you set all this up and guide you to an extent. But in this case this is your first big co and they will make sure your time is totally consumed by our mountain of lawyers , procurement officers, security , it , operations.... but at the same time finance want to check your books to make sure we are not your single customer and that you have a history of making money , check the key peoples backgrounds for any history of fiscal malaise , conflicts of interest blah blah blah.

The worst is when. We deal with the vendor negotiate a price the procurement people who have no idea about what it is their purchasing have a kpi on saving big co x % off every piece of paper that passes their desk so they go in at the last mile after pretty much everything is set and pull your pants down (don’t forget to leave fat for this )

And good luck.

I would love to see someone start Stripe for POs. I work at a University and navigating the bureaucracy just to pay for Asana was an utter shit show. Especially once the university procurement team learned ANOTHER department was paying for it, and suggested we just use their license... smh. In a similar vein, I often opt for yearly billing even not discounted just to not have to submit receipts and proof of purchases every month for the things we are allowed to purchase.

Cannot agree with this more.

My team was able to deploy a highly sophisticated technology stack in less than 12 months that could add up to about 50 years of procurement (based on past experience acquiring software for large enterprises). There are 20x more disparate solutions/components, but it all rolls up to a single AWS Bill.

This approach has given us remarkable agility and easily outperform traditional vendor products like Mulesoft, IIB etc. because we are easily able to augment/extend/enhance

That's about right. I was working on a an AWS project a while ago while a colleague was working on a procurement project. What's when I had the realization that 1) AWS services are nothing special 2) there are usually superior SaaS options for each service 3) therefore the primary value I got out of AWS was the fact that I wasn't pulling my hair out like my colleague

So do news sites, so do streaming services, so do (in my opinion) charities. The manual overhead of setting up (and then managing) a bunch of micro-subscriptions is a significant source of friction when it comes to adoption. It raises the barrier of "is it worth dealing with?". Payment providers like PayPal and Apple Pay do a little bit to help you centralize things, but I think there's a ton of improvement left to be made in this area.

Actually, AWS Marketplace is trying to do exactly this.

GCP has a similar capability.

One thing that is kind of amusing was that I was able to find a price sheet in the AWS marketplace for GCP-owned Looker that isn't available anywhere else.

Apparently the AWS solution not only works as a way to simplify your purchasing, it also works to produce better transparency in the market.

I'm not sure the product mix in either case is necessarily what the OP was talking about, though. For example, I don't think, e.g. SalesForce or HubSpot are available on either.

I dug for an example on Google Cloud and found this "Zenoss" product [1].

"Managed Services are fully hosted, managed, and supported by the service providers. Although you register with the service provider to use the service, Google handles all billing."

Seems to be exactly what the author describes. No idea what share of the subscription fee Google takes.

[1] https://console.cloud.google.com/marketplace/details/zenoss-...

A significant problem with GCP is how they only handle credit cards, and with various limitations, in many areas.

I had outages purely because I could not just wire the money - a common way of handling payments in Poland (I have even used bank transfers to pay for pizza takeout).

I know. And given how hostile Amazon can be to sellers, I would love to see them fail to become the standard.

Amazon caters towards the buyers for sure. That makes them a fair-sized favorite to win. Golden rule: "He who has the gold, makes the rules."

I'm surprised I had to scroll this far down to see this. AWS, Google and Azure all have this to varying degrees. If the product you want isn't listed, it's frequently simply a case of asking the vendor - there are some processes in the back end that might have to happen for you to get a quote.

The cloud providers are trying to solve this problem - more and more of our customers are purchasing through the marketplace. It allows them to drive down spend commitments to AWS, GCP, etc - it's a pretty good experience all around and enables usage based pricing + easy collections. Which is something all startups hate dealing with.

I had a similar idea recently after hearing a CIO say that there was only 1 (one) corporate credit card in his 1000+ employees company, and it was the CEO's. For him, SaaS on a credit card was a no-go, it needed to be a monthly invoice.

I think it's an opportunity for a new kind of financial business, where SaaS subscriptions can be ordered (with some kind of approval workflow) by a selection of LargeCorp employees, and there is only 1 monthly grouped invoice to LargeCorp, with a known and pre-approved PO number.

Money could even be made by getting the money in from the LargeCorps earlier than it's being paid out to the SaaS services. Or a simple markup.

Y-Combinator backed company which does something along the lines of what you describe: https://onetool.co/home/

https://onetool.co/saas-vendors/

"Keep your competition out! onetool accepts a limited number of partners for each service and pricing-tier. Most of your competition will not be onboard if you are."

Read as, we'll only support a limited number of vendors who approach us first, regardless of which the best solutions are.

More likely, they will only recommend the vendors they get the biggest cuts from. As a user this is no use.

I'm very skeptical this would work as a for-profit business, especially a VC backed one, because their will always be pressure to take a bigger and bigger piece of the revenue.

Interesting that their social proof lists a bunch of silicon valley companies that are probably about the least likely to have the issues I described.

One of the issues of making all the services together is the discount. Implement a discount system across multiple SaaS has to do with sales strategy.

Standardize everything, including the costs, seems like the end of creativity. However, creating a massive checklist that makes the user choose between services and, step by step, simulate final costs (which will include computation, task management, monitoring, customer support, etc.) is a titanic job that could change within the next generation.

I don't think discounting bundles really needs to even be included in this. Just having one payment processor counterparty is already a huge win.

We've been waiting 90+ days for one of our clients to pay their invoices once we switched from CC to PO + ACH through Stripe.

I'm about to just see if we can charge the card for the full three months so we can manually flag the invoices as paid. Terms are net 18 usually.

The PO process is awful. We have to fill out forms as a vendor, submit tax information. It's completely unneeded.

If your company requires POs for SaaS you need to take a step out of the 20th century.

This could also enable bundling. For example you could have a startup bundle that includes anything from cloudflare's pro plan to tools for email marketing.

Sure. Once you solve the fundamental purchasing problem, you have all kinds of options up to an including building a sort of hodge podge AWS.

These types of marketplaces do exist but it is very difficult to justify using them. Why pay a significant margin (15-30%) so that we can use a pre-existing billing mechanism inside AppExchange or Azure or AWS. It makes more sense to maintain your own billing system and buyer journey. If the margins go down to 5% then it becomes interesting to give up that control.

I don't think marketplace is the metaphor I'm looking for. A single purchasing portal that takes maybe 5% to keep the lights on is more like it. A PayPal or Stripe set up by a SaaS consortium just to remove friction from the industry might be a good approach.

Such companies do exist. They are usually run by friends of the people working in procurement at various companies.

Sounds like my reseller buddies.

The entire SaaS industry already looks like a single vendor with many products for sale, accessible by search engine.

I wonder to what extent managers are defaulting to an AWS provided services just to reduce cognitive overhead (ie. thinking)?

'Putting things on corporate card as a no-no' is definitely a rule they should get over and literally have the CFO or Controller use a card for such purposes, that's entirely feasible and rational.

The companies I'm taking about have 10K+ employees around the globe. The CFO would to nothing but type in the credit card number all day.

Why would employees be buying services individually? With 10K employees there would be a site licence.

Just because your company has 10k employees doesn't mean they need 10k licenses for everything they buy. Often only 10 people in one department in one country need that one specific thing.

Indeed. The team should pick the right tools for them.

A 10K employee company will use a finite number SaaS services, they should absolutely be well equipped for their financial staff to 'pay' for stuff, including software, using a variety of means, including credit cards, it's what they do all day, this is not a new concept and neither is the scale of it.

They buy toilet paper, cleaning fluids, pens, envelopes that way.

A team of '10' needs 10 licences, they have their rep. from the finance do that, if it's over $X amount it needs approvals etc. etc. it's not rocket science, just like the special chairs, the screen for the window, the special hard drive.

They buy toilet paper, cleaning fluids, pens, envelopes that way.

I can assure you they do not (at least not in any company I have worked). They have pre-approved suppliers for all those things that are well integrated into our invoicing system. Adding a new supplier to that list is a major process.

If I need a new one off hard drive I can either either ask IT to order it from the one approved computer component company or I buy it out of pocket and go through the reimbursement process.

Yes, of course many things are purchased through contracts.

And many things are not.

Every company buys stuff from Staples/Costco/Wallmart, one-off parts and supplies.

The article mentioned having a 3rd party for the billing, indicating probably some kind of larger purchase, it's odd. Commercial Banking usually provide Amex/Visa linked accounts with limits, for a variety of things, it's not a new idea.

> Every company buys stuff from Staples/Costco/Wallmart, one-off parts and supplies.

They have vendor relationships with them. The amount of stuff they're putting on the credit card is a rounding error.

That's a 2 year process to get 10K site licenses, possibly to the point where the CFO gets involved. Have fun with that.

Extra credit for

- team accounts

- budgets and rules for new software

- password management with security audits and MFA enforcement

- automated purchasing during employee onboarding

- hardware and other purchase management like laptops and mobile device management

Isn't that just what https://www.appdirect.com/ offers?

https://www.appxite.com/ is one company doing exactly this

That’s a brilliant observation. Something we came across as well and we ended up starting Spendflo.com to solve it :)

I should be able to license any Saas tool without talking to a sales rep or doing some kind of "scoping" call.

This is a technique many companies employ when the product is inferior, they do not want the competition to evaluate their products/offerings without signing contracts.

It's even more lopsided. AWS is a short term lending institution, it isn't pay up front.

> Atlassian because they don’t take POs

To me that sounds like Atlassian is leaving a lot of money on the table

>lot of money

I believe lot of money is a relative term.

Fuck no, get your company to not be dumbasses.

doesn't Salesforce already has something like this ? Heroku addons ?

There's limited examples. What I'm saying is this model needs to be much more widely used so it's easier for SaaS companies to access big, slow companies.

Sounds like he's asking for a new PaaS (Procurement as a Service).

What is a PO?

Purchase Order, sort of a 'reverse invoice' used by large companies to manage purchasing from suppliers.

Reply on news.ycombinator.com