Ask HN: Who is the target audience for ~$1000 software? • Hacker News Board

Re: Ask HN: Who is the target audience for ~$1000 software?

#41

Post by vram22 » Sun, Oct 15, 2017, 9:28 PM UTC

troydavis wrote: post #34

jdietrich wrote: post #16

Earlier quoted context omitted.

$1000 is the purchasing limit in most large enterprises - anything over $1000 needs specific authorization. Buying a product at $999 is completely frictionless, while buying a product for $1001 involves multiple meetings and reams of paperwork. In enterprise purchasing logic, $999 is indistinguishable from free and $1001 might as well be $50,000. Clever software marketers have realized that a $999/mo SaaS product fit…

At least from having sold to a small subset of them, the “most” in “$1000 is the purchasing limit in most large enterprises” isn’t accurate. Yes, almost every large company has different purchase requisition approval processes based on the amount and often the type of expense, but they vary widely between companies and roles. There’s nothing magical about $1000 vs $500, $250, $0, or $2500. I’m not even sure $1000 is…

What does "Procure to pay" mean?

Re: Ask HN: Who is the target audience for ~$1000 software?

#42

Post by vram22 » Sun, Oct 15, 2017, 9:40 PM UTC

twobyfour wrote: post #36

If you make $50k/yr (as a freelance headshot photographer in a small town) using Photoshop, a $1000 license plus a $500 upgrade every two years is worth your while. (I'm not up to date on the current subscription pricing but assume it's analagous.) Many people make much more than that per year using such software. In my current job at a small business, there are things we spend $10k/mo on to make $1M/mo in revenue. E…

Great realistic comment.

To reinforce it, there's that saying that some people know the cost of everything and the value of nothing.

Re: Ask HN: Who is the target audience for ~$1000 software?

#43

Post by earenndil » Mon, Oct 16, 2017, 4:19 AM UTC

zxcmx wrote: post #4

People who use the software professionally to make money. There is a market segment between enthusiast and enterprise where people pay real money for software which supports their freelancing and consulting. Hex-rays and (and formerly photoshop) fit in this category. Similar things might be wireless site survey software, antenna/ EM modelling, CAD, all kinds of engineering software really. "Pro versions" of security…

That makes sense. But why that? It seems like it would make a lot more sense to market to hobbyists, of which there are a lot more of (as jetbrains and newly adobe have done), or to large companies which will pay a significant sum to give all their employees access to the software.

Re: Ask HN: Who is the target audience for ~$1000 software?

#44

Post by earenndil » Mon, Oct 16, 2017, 4:42 AM UTC

p0nce wrote: post #33

Hex Rays and ICC are unmatched. Once you have worked with ICC backend you'll miss it a lot when using say, LLVM. Same story for IDA, I don't think any competitor does something at this level. Photoshop is being displaced by nimbler competitors.

The one time I used icc (I got it for free, and wanted to play with it, and I had an application for which perf was relevant), it had significantly worse performance than clang, which trailed just behind gcc. In what way is it unmatched?

Re: Ask HN: Who is the target audience for ~$1000 software?

#45

Post by p0nce » Fri, Oct 27, 2017, 11:47 AM UTC

earenndil wrote: post #44

p0nce wrote: post #33

Hex Rays and ICC are unmatched. Once you have worked with ICC backend you'll miss it a lot when using say, LLVM. Same story for IDA, I don't think any competitor does something at this level. Photoshop is being displaced by nimbler competitors.

The one time I used icc (I got it for free, and wanted to play with it, and I had an application for which perf was relevant), it had significantly worse performance than clang, which trailed just behind gcc. In what way is it unmatched?

I used it for years. The auto-vectorizer is unmatched imho. It has a very systematic way to turn loops into incredibly tight code. I use LLVM 4.x and this is very far from ICC.