[haswell]

I worked for enterprise software companies for a number of years.

I’m not personally a fan of “call us” pricing, but it is indeed effective, and the companies that sell software this way do so because they prioritize establishing human relationships between a salesperson and the buyer.

In the long run, this leads to more sales/upsells.

[YetAnotherNick]

Also not only just establishing human relationships, sometimes software might not be available at all for plug and play and require client level tuning to be effective. Even in cursor, setting up a good .cursorrules seems to have quite a big effect.

I don't think "call us" is an effective inbound sales tactic though, but Mistral does have folks in sales who can do outbound sales.

[haswell]

This is a key point. A customer that buys $100K+ worth of software is automatically a future debook risk if they can't/don't implement it successfully and start seeing value right away.

The high touch approach ensures they're getting support at every step of the way and increases the chances that they'll use the software and renew later.

I'm not sure how many of these factors are in play for Mistral, but to your point, it's easy to imagine some scenarios.

[b112]

No, they specialize in knowing how to extract every single penny from you they can. Find out how big the company is, and bleed them as best possible.

If you're too small, they won't even talk to you.

That's what besoke pricing is.

Knowing your client happens regardless of showing pricing or not.

[haswell]

This is a fairly cynical way to frame this. There is absolutely bad behavior in the enterprise software space, but that is certainly not universal, and is not inherent to the "call us" approach.

> If you're too small, they won't even talk to you.

This is often true. Some of the software I worked on was extremely expensive to host, and there was indeed a minimum threshold that was many multiples of $10K.

It's not that the company didn't want smaller players to use the software, but that smaller players just weren't large enough to benefit from the minimum buy-in, and selling the software at a lower cost for those smaller customers would have just been pure loss. Over time, they were able to lower the minimum threshold due to improvements in the architecture and economies of scale, but the point here is just that these minimums often exist for good reasons.

> Knowing your client happens regardless of showing pricing or not.

It really does not. Many software companies have a minimal relationship (if any) with their customers. For some customers and some product types, this is perfectly fine. But when a company is buying software that will cost the company millions/year, having a direct line to a real person who in turn can arrange conversations with product management, customer success, etc. is table stakes, and is often not possible or available with smaller vendors.

You can dislike the model, but I'd suggest digging in to some of the why before dismissing it too reductively.

[GuinansEyebrows]

> This is a fairly cynical way to frame this.

dismissals of negative reactions as 'cynical' rarely acknowledge the fact that a 'cynical' response is often no more cynical than the cynicism of the target.

bespoke pricing is a cynical tactic, no matter how you dress it up. it provides a legal shroud, and i've no more patience in me to give the benefit of the doubt to any profit-motivated enterprise that can't at least be upfront about what they want to charge for their services.

[haswell]

I think you overestimate the degree to which some software can be standardized and sold in a set of well-defined SKUs, and the degree to which some buyers want to be intimately familiar with extremely granular pricing (e.g. something like AWS).

Again, speaking only for the places I worked, part of the reason pricing wasn't simple was that larger customer deployments were tuned to the customer based on a myriad of factors ranging from the specific software modules the customer purchased, use cases they intended to deploy and the load characteristics of those use cases, etc.

Setting aside for a moment any potential bad behavior, the bottom line is that for some kinds of software, bespoke pricing is a more accurate reflection of the reality of the deployment than trying to force some kind of standardized label on it. The places I worked also had pricing books they'd show customers, but due to their complexity, they would not publish these publicly.

> bespoke pricing is a cynical tactic, no matter how you dress it up

We'll have to agree to disagree. Having worked with quite a few large vendors over the years, there are clear and obvious differences between them, better and worse reasons for this type of pricing, and there's a reason that some companies have earned a negative reputation while others have not.

It's also not clear to me why you've concluded that this is all inherently cynical.

[GuinansEyebrows]

maybe i overestimate. maybe i just appreciate the difference between the product and support for the product. for support, i understand a more-bespoke pricing model, but it'd be nice if there was at least some upfront inkling of how much i have to open my organization's veins in exchange for a working implementation.

interacting with capital is a cynical act. capitalism is predatory but it's necessary to interact with it. if you're not cynical, you risk being taken for a worse ride than you have to be. this isn't me handwaving things; it's a fundamental aspect to how i see the world. cynicism doesn't have to be a simple doom-and-gloom "well, everything is bad, end of discussion" (regardless of my personal feelings about it) - it can be a tool to make sure you're able to interact with systems in ways that benefit you and others while retaining whatever modicum of control you can - because even if you aren't, your vendor is (or they're on their way to being out of business as a profit-motivated entity).

[haswell]

> maybe i just appreciate the difference between the product and support for the product. for support, i understand a more-bespoke pricing model

I'm trying to frame this sentence so it doesn't sound like a jab because that's not my intent, but based on what you've written, I don't think you understand the kind of software I'm describing.

In your mind, what is the distinction between the two, especially for an enterprise solution hosted by a vendor? In my experience, the two are often not so different at all, and the clean lines you imagine here are not lines that exist in practice. This will depend on the nature of the software, and the kind of support customers need (i.e. infrastructure vs. implementation).

> but it'd be nice if there was at least some upfront inkling of how much i have to open my organization's veins in exchange for a working implementation.

You are assuming this is not part of the model, but it is. Not publicly listing a pricing sheet does not in any way mean a customer doesn't have clarity about how much their deployment will cost before they sign a contract.

> interacting with capital is a cynical act.

Your use of "cynical" is in a different category than what I was describing above. If we go with your definition, there is no reason to differentiate between bespoke pricing and up-front published pricing.

[nickpsecurity]

I think your argument is supported by the fact that many companies, in and outside of AI, have pricing upfront. That's for cloud platforms, OS's, frameworks like Sciter, shrink-wrap, semi-custom with value adds ("starts at..."), per-token pricing on models, etc.

Then, some companies wont give the slightest hint of pricing unless we talk to their paid salespeople. It's definitely a game to extract more money out of you. While they may or may not, they often ignore low-volume customers who could easily buy the product if it was on an online store. Your skepticism is warranted.

[haswell]

> While they may or may not, they often ignore low-volume customers who could easily buy the product if it was on an online store. Your skepticism is warranted.

Setting aside any qualms about how the pricing is published, if a business chooses a strategy in which they focus on large customers and choose not to take on small customers, why is this an issue? Especially when the market is filled with alternatives?

The support model, predictability of yearly renewals, per-customer overhead, etc. look quite different when selling to larger customers vs. small/low-volume customers.

[nickpsecurity]

It depends on one's moral philosophy. One kind would see it as discrimination with a negative, long-term impact on people and markets. Others would say they can do whatever they want. Subjectivists and capitalists would encourage them to do so to maximize selfish gains, esp money and power.

If about selfish gain, then you should have no concern with people calling out their practice to warn others. They're doing what they want to do. They're also helping others with a warning to avoid harms, like lock-in and overcharging, that are more likely with "call us for a quote" type companies. The warnings are also market signals for buyers.

In my case, I also tell people to encourage good practices like having prices up. Posts like mine might also inspire regulations that force prices to be shared ahead of time. They might also inspire people to use or develop lock-in-free alternatives which some out there are doing.

[haswell]

> It depends on one's moral philosophy.

To make sure I understand your comment, are you saying that a company that sells a product with a $50K minimum buy-in — a number determined to be the threshold at which the company can recoup development costs and make a reasonable profit — is engaging in some kind of immoral behavior because it can only be purchased by larger companies?

> One kind would see it as discrimination with a negative, long-term impact on people and markets. Others would say they can do whatever they want.

This is quite the false dichotomy.

There are many valid reasons to sell to specific customer segments/markets that do not amount to “we’re doing it because we want to”.

In the early days of 3D printing, such hardware was prohibitively expensive and primarily sold to businesses. Even now, there are classes of 3D printer that cost many multiples of $10K. It would be strange to classify this as discrimination vs. acknowledging the realities of the market and the fact that it only makes sense target specific customer types depending on the product.

> They might also inspire people to use or develop lock-in-free alternatives which some out there are doing.

Lock-in is orthogonal to this pricing structure. It sounds like your primary issue is with companies that behave badly and use dark patterns. I share those concerns. But those issues are not inherent (or limited to) to “Call Us” pricing.

[nickpsecurity]

Your post has good points. The problem is you're assuming a group that invested X and believes they have to charge Y to recover it. That this is fair and makes sense. (It mostly is and does.)

In fact, most AI suppliers ripped off other people to the tune of TB (copyright infringement), spent a lot on the training, and want to recover their costs only while denying content creators theirs. So, let's start with this being about massive infringement of something they sell to shift all value to themselves.

From there, they can sell only top dollar to enterprises or tiered amounts based on buyer income. The latter being discounts applied to a high, but reasonable, default price. Also, they can be open about this or call us for a quote.

There are companies that do the first. They often bundle features into Plus, Pro, or Enterprise deals to make those sells more straightforward. So, the AI suppliers could definitely do that. They dont have to but it would be good. Hence, me bringing it up where their employees read.

The other point creates asymmetric information between the seller snd the buyer in a way that only benefits the seller in a sellers' market. High-performance AI is probably a seller's market right now. Most writing on that says the asymmetry is bad for buyers by default. Also, goes badly for them more often than transparent pricing upfront. So, I call it out by default.

The only thing worse is if the terms of deals are kept secret. That enables and drags out abuses for longer. I cant find any pricing information on a lot of the AI offerings. That suggests they're doing that, too. If any NDA's on price or performance, stay away unless absolutely necessary.