Ask HN: What problem in your business is worth paying 10k/y for?

66 points by cronjobma ↗ HN

116 comments

[ 2.4 ms ] story [ 182 ms ] thread
Slack. Apparently.
My current employer doesn't use Slack and instead uses XMPP and a web front-end called "Movim". I can tell you right away, it is shit. I'd gladly pay 10k/year for the privilege of using Slack, so please don't bash on it until you realise how bad the alternatives are.
Discord is free and in many ways better. If you don't mind the gamer vibe.
In what ways?
It doesn't have (as much) performance issues. Seriously, loading a couple hundred messages shouldn't take even a second, what are we, living in 1990?
Wait until you get Slack, then you can revise your statement.
I've used it for talking to third-party clients that hire us and I don't see why should I revise my statement. It just works, it does the job just fine. The only complaint is that the desktop client melts my battery but given that I run it in the browser it isn't that big of a deal.

Also, I forgot to mention - we did have an Slack clone called Mattermost running and the overall response from the dev team has been positive. The only reason it was shut down is because of some open-source fanatics who thought Mattermost wasn't open-source enough (apparently the fact that it's backed by a big company providing support is a sin) so we've switched to this XMPP abomination.

Well we already had a proven self-hosted Slack alternative called Mattermost that pretty much everyone enjoyed. The push for XMPP is because some of those fanatics live in a parallel world where everyone is using XMPP and they'd be able to talk to people on other servers with it. Of course reality is very different.

I'm leaving soon anyway so I wish them the best of luck with this (and I feel bad for some of the people who actually wanted to get work done instead of fighting with broken and outdated tools).

Mattermost team here, glad to hear we're making a difference!

Yes, we do have people wanting connectivity via XMPP, IRC, Discord, Slack, etc. There's "matterbridge" that offers those options for Mattermost: https://github.com/42wim/matterbridge

You didn't really use Slack if you only had it to talk to one client.

It needs to be company wide for you to understand the truckload of channels, the spam of notifications and gifs, the poor mobile app, the poor replacement of emails, etc.

Ever checked out Rocket Chat? Open source solution that is extremely similar to Slack and you can self host.

https://rocket.chat/

Slack reminds me of that saying that there are some software/programming languages that everyone hates and those that nobody uses.

Its almost as if Slack is the JS of the team chat world.

Everyone might hate it but they're definitely using it. Agile is too trendy.
On the fly image processing & optimization.
Interesting. You mean something more advanced than you can do with some command-line scripting?
Image processing tends to be quite application-specific. What's the application?
What context and what does that mean exactly? Image processing is not usually a difficult hurdle.
As pier25 suggested, I was thinking about imgix. We (a marketplace) stopped using our hand-rolled image processing service for Imgix, and now enjoy their performance and suite of clever tricks to deliver the least bytes the fastest to the user. The impact on SEO and conversion is non-negligible.
I work in this space. What specifically are you thinking about? As another person said, image analysis is very application specific.
databases, migration tools
I've written a sql to elasticsearch mapping tool I feel I should of released. Eleasticsearch is in desperate need of many tools we take for granted in sql land, like a command line client you can easily punch in queries on the fly because trying to write multi line json into the commandline curl is death so I'm forced to install kibana.
Can you give a few more examples ?

I am interested because writing these support tools IS effort+time+focus but not usually a technical problem.

Mostly emulating what phpMyAdmin can do, the time spent for a elasticsearch beginner trying to figure out how to clear a index without losing or having to rebuild their mapping is insane, basically the answer is just don't and just copy your mapping but if something could easily extract that and say flush index then awesome. Creating mappings is a nightmare if you don't remember what types they have and it would be insanely simple to do the phpMyAdmin style of create table where you type in a "column" name and then use the drop down for the type "string, date, geo point".
saving 20k/y (cloud bills, for example)
What does your tech stack look like?
I would also pay 10k / year to save 20k / year.
Depends on the company, really. But raising the effectiveness of a sales team by 5% if each sales person brings $100k in revenue each year makes sense with over 2 employees.

Bumping up the conversion rate within a sales funnel. A company is making $1,000,000 a year selling dog food. Improve the conversion rate by 1% and you have yourself 10k.

Anything that any salesman or customer service employee mentions that plugs into salesforce. A mail button for sales force ? A print button for salesforce ? A make my phone call customer button inside salesforce button? Oh sweet only $20 per month per salesman....so we have 50 people on salesforce so only 1,000 a month, only $12,000 a year? This should help get us out of dept so fast !!! FML.
I'm guessing you weren't pushing for Salesforce, but out of interest, did your company evaluate any other CRM platforms? If so, do you know why they chose Salesforce?
You would incur the same type of thing in Dynamics. It's a hit or miss trying to figure out if a feature is even possible and how to implement on your own. There really isn't an open-source collection like you would hope. Subscription or dev. consultant?
There's not an open-source CRM I'm aware of that's a strong enough competitor to Salesforce, however there are much cheaper options available. Are you in the market for one?
They bought into it when they were a tiny company over 7-8 years ago and barely used it for a few years, a few yeas ago me and some other devs joined and said why do you have this and why spend this much money for storing a customers email and some notes when we you also have 2 other systems doing the same thing. We tried to pull them away and build a small system for their small needs but that's when they dug the hole ten times deeper with numerous integration. No matter how big or small the integration the prices add up quickly. They were a company with out developers at one point and still never learned to ask if we can do something vs buying into mailchimp or salesforce plugin. No matter what the CEO wants he's afraid to take away the salesmans costly toys.
(comment deleted)
Interesting, I have no direct experience with salesforce but I have heard numerous people discuss similar things as either consultant centric services or outright business models. Is this that common?
Honestly leaving salesforce is a piece of cake if it weren't for the ecosystem of other tools and integrations. A few years back I did a simple export of all our salesforce data and made a demo of a salesforce clone and if that wasn't good enough it was easy to import into sugar crm. 1 out of 40 salesman complained it wasn't exactly the same so they stayed and went full force with plugins and now it's a web of plugins that's harder to escape. As soon as a salesman sees a ad for "this plugin will make your life easier and your job " it's done, they setup a demo with our CEO and beg for it like it's a Christmas pony.
I think it's a two step process. First, install SuiteCRM, second, replace all those salesforce plugins with a single SuiteCRM developer who can craft custom plugins.

This has the added benefit that you can then sell those SuiteCRM plugins to other businesses.

Tableau. We pay far more than $10k/year even after the pricing model change.
Think about running your own Superset instance. Could be the solution and IMO it's much better than Tableau.
Is Superset mature enough? We have been exploring a lot of BI tools so much so that we near scraping the bottom of the barrel. But Caravel/Superset never came up because it is still under incubation and that doesn't go well with the enterprise managers. It doesn't help that there have been constant name changes.
It's great and IMO pretty damn mature. They use it widely in prod at Airbnb so the only ways in which it's not "mature" are in customizability.
What kind of customization?
I think there are just certain scenarios it may not support, like at one point writing queries in the editor for JSONB fields in Postgres was either hard or impossible, things like that. But generally it's pretty great and ready to be used IMO.
What do you use it for?
Making fancy graphs for data analysis. Sorry, I can't get into it much more than that.
What do you use Tableau for? I'm curious we also pay quite a bit but it's a less than ideal solution for us.

[Update: just saw your response to another comment]

(comment deleted)
This is good thread for anyone wanting to figure out what B2B SaaS business to build.
I presumed that's why the question was asked; someone is looking for opportunities.
Problem is that pretty much everyone in here is already a software engineer so their problems might be quite hard to solve.
(comment deleted)
an AR headset that allows us to work on 50, 60 inch displays where ever we are, without having to mount large displays to walls etc.
What's your business that you guys like to work on large displays?
We design User Interfaces for Web and Mobile development teams (http://fairpixels.pro). Working on large displays seriously increases productivity.
Better EDA tools.

Oh my god, i would pay so much for that...

What does EDA stand for?
Electronic design automation, e.g. Chip design.

This is not low hanging fruit.

I'm interested in this field. What are the current pain points with EDA tools? And are we talking device level, process level, circuit level or all levels?
Pretty much everything.

There's a landmine of bugs in even the best EDA software. I'd just like for the software to run nicely once, like TensorFlow... sigh.

If you're thinking getting into this field, I might suggest a modular software architecture where the non-proprietary parts are open source and everything NDA-sensitive is kept closed source (pretty much anything the foundry touches/provides).

Super super interesting. So this is the case for the big expensive commercial products like those that come out of shops like Synopsys? It strikes me that helping to improve (device side) EDA is probably a pretty sound investment of my time, since I've got some phys chops and it's pretty important to the global economy.
Oh, believe me, in many ways it's worse with the big expensive tools because those are produced for only a few companies.
what are some or one of the largest offending bugs or class of bugs?
Compatibility is a huge one. Honestly, if you write a robust HSPICE to ___SPICE (or Xyce) translator, I'll be very grateful.

There are other problems too. Just in general, lots of bugs. For example, we'll run something and we'll get an obscure error, and then we'll ask the EDA company, and then two weeks later we'll get a generic "we'll look at it".

The hilarious thing is, sometimes the more expensive EDA companies aren't necessarily the better ones, our current partner isn't necessarily known as the most expensive tool suite, but we've actually had (so far at least >.<) had a much better experience with them than the bigger vendors.

Isn't that the problem with whole industry? Every vendor makes it really hard to work with tools from other vendors. Furthermore initiaives like openaccess/accellera acctually seems like they are a paywall to stop others from entering the market.

Is your company big? Usually bigger EDA vendors don't care much for smaller design houses.

Btw. how are your experiences using Xyce? Do you get much speed up using its parallel processing capabilities?

Our company isn't that big, and that's a fair point, but one of our circuit designers worked at Intel for over decade and reported the same experience.

Xyce is new and has some syntax differences, so it's not a mature software quite yet, so there are some troubles there. But as we've been starting to make it work, it's given us some amazing speedups on large clusters.

I spent a bunch of time with design teams. Some SOC, some standalone memory, some memory compilers.

I could definitely see the value in having a plug & play design flow.

Correctly identifying if the customers are going to pay their bills or run after two months without paying, leaving a bill of ~1000 (consider that the State justice system doesn't work and the customers are individuals).
Do you report them to any credit agencies?
No, because they can sue me if I do and even if I'm right I don't want the risk of having to pay an indemnification, Brazilian judges are crazy.
I maybe wrong but some people in my team were building a Customer HUB at a point. My understanding from the project was you could buy third party data to help you out with this.
Extracting data from PDF in a reliable way.
We do a decent amount of work in this, lots of scraping the web and extracting. Unfortunately we dont do it against pdfs that have any strict format or even a loose format at all, I wish it were government forms or any type of forms. Would you say the pdfs you guys are looking at have some type of format and the readers are just hit or miss?
Extracting at least text from PDFs is not always 100% perfect, due to inherent issues with the PDF format (partly because it is a graphic format, and does not have a one-to-one mapping to text, also maybe because of some weird decisions they made). I both read about this and was told about this by a key person at a PDF software product company, whose product I researched and then used in a project. The product was xpdf (a C library, it also had binaries or EXEs), from Glyph and Cog. I was contracted by a client to research PDF libraries for extraction of text from PDF; found and evaluated a few, then recommended xpdf to the client, and used it in the project. That is how I know this.

The only guaranteed way to get 100% accurate text from PDF is ... to not do it :) Instead, get the text from the same source that is used to generate the PDF. Obviously, that will not always be possible, but when it is, it is the better solution.

I am a freelancer doing exactly this. Drop me a line if interested.
Anything that increases team productivity significantly. If my team of developers is > 10 then I'm probably spending about 100k per year in supporting tools:

- source code management - server monitoring and metrics - payment processing - testing tools - continuous integration tools

Any of those could easily cost 10K/year, or more, depending on what the team is doing and how fast they are growing.

Beyond that, anything that ties back to revenue. Email marketing, PPC, SEO. Also possibly things that optimize finances, analytics or anti-fraud.

Would https://feram.io fall into this category? (Disclaimer: I'm a cofounder)
Interesting, working on a bitbucket version?
I'd like to hear how your pitch to upper management differs on spending for tools for a revenue team vs spending for tools for a development team. Customer service and sales gets whatever they want at my job but it's pulling teeth asking for anything for increasing programmer efficiency.
(comment deleted)
Great question, there are many opportunities here. Network with entrepreneurs, sole proprietors, private practice physicians/dentists/etc, and other small business owners and you'll often hear similar complaints or pain points.

As a general rule I'd say nearly any business would pay 10k/year for:

- Anything that would offer 2x that (or greater) in ROI

- Anything that would automate or handle routine tasks that otherwise an expensive person must perform (there are many of these if you think about it)

- Anything that saves 10k+ worth of someones time (this is not too hard to identify if you focus on the expensive persons)

- Anything that generates 12k+ of new revenue

- Anything that replaces 10k+ worth of annual contractor/consulting work

- Anything that undercuts competition charging more for the same service, without a degradation in service quality

You probably want specifics, so two areas I often see smaller businesses struggle with are email lists and security/software updates.

There is substantial opportunity in finding new ways to monetize the boring old email/customer list in nearly every industry. I frequently hear small, medium, and even some larger businesses struggle to monetize an email list in any meaningful way, and so instead the primary metric is always about 'engagement' and other buzzword centric statistics that rarely translate into revenue. Private practice doctors and dentists often struggle with this, as do everything from bakeries, app developers, web sites, authors, interest groups, etc. There are many businesses with large customer lists or customer prospects in the form of email lists that make $0 from those lists.

Another is anything related to security, particularly anything that takes a perceived complicated task and makes it easy or handles it outright without any downtime or complication to the end user(s). Whether it's routine client and server software updates, routine security audits, moving web apps/sites to https, maintaining and keeping up with web/email/crm/accounting software updates for small businesses, etc, there is huge opportunity out there for much of this to be a subscription type service rather than expensive periodic consultant or contract work that is often very distracting for the small business to implement.

Anyway, those are just two simple ideas that I see/hear frequent issues with. Usually a focus on something common that already exists but needs improvement will be much easier than trying to do something completely unique.

Extra good point for security. It's easy for us to say we are using 10k security product to make 60k customer happy but we can also brag to any customer from there on.
Curious. About 1 week ago I saw a post on reddit that asked "What product would you pay $20,000 for?" I thought that was odd, typically that's not how you plan a business, you find something or create something of value and then you sell it. You don't start from how much you want to earn and then look for ways to earn that much.

Now here's a similar post, what would you pay "$10,000/yr" for.

You can label me a skeptic, but I wouldn't buy anything from anyone who's primary motivation for starting their business is to make X amount of dollars in Y timeframe.

It's a standard way of finding business ideas. look for where there is a need. hackernews has these every week, I actually think they provide interest, even if the intention is business rather than discussion.
Fair enough, I just think that people should look for their career by either finding ways to monetize activities which they already enjoy, or by looking for ways to create new kinds of value for the world. If you do those things, revenue often comes.

By contrast, every disgusting corrupt snake who has ruined the world for everyone else has started out with the goal of making X dollars in Y timeframe. So while I'm not saying they're all going to be evil, all of the evil ones share that trait.

Someone with the right skillset can get some ideas from the discussion. Not randomly pick a good one, but the one that fits you.
I do fraud investigations for a bank. We spend an insane amount on labor costs because all of our transactional data resides in different systems. Want to review credit card data? Go pull system X. Want to review wire transfers? Go pull system Y. Etc etc. some of these things have to be separated out of necessity (e.g. card data is stored in a particular way due to regulations), but I would easily pay 10k / y to be able to get a complete view of customer transaction activity from a single place. We'd save millions in labor costs if someone could pull it off correctly.
how are these systems implemented? are they all proprietary systems? do you have access to the back-end data?
I'm in a bank as well but I reckon 10k won't be enough to solve the issue. We've built an enterprise data warehouse that pulls everything together and it still hasn't been perfect. What we do well though was extract the data at a higher level of abstraction - we came across a tool called Thoughtspot that also helped us join things together and drill down but from what I heard, it's pricey.
For a dev shop, anything that brings in more than (more or less) $20k /year.

Otherwise also fundraising (as someone else said).

Though for both of these, if the business is cash-strapped, the results would have to be pretty much assured.

Enterprise pays the most for ERP. If you can create a better SAP, then you will make millions. I am saying as an SAP user, it is good but really expensive.
There are already tons of ERPs and the space is crowded at this point.
Creating an standard ERP system is not the hardest part (it's still really hard tho).

Deploying ERP for enterprise customers means writing tons of custom modules & processes specific to the customer. That's why integrating an ERP for the first time or switching vendors is costly and risky. Most successful ERP systems for enterprise customers are basically domain-specific frameworks to create customer-specific solutions.

However, there is still a big market for business management software for SMBs. The market is so big and ever expanding that I doubt it can ever be saturated. Aside from english-speaking market, there are even bigger opportunities for localized solutions.

Acquisition of high-quality clients for my software agency is a problem for which solution I'd be willing to pay 100k/year.
those are some pretty high quality clients then. What type of projects do you do. We've implemented a pretty good model for our agency to attract quality clients without networking and doing all sorts of stuff that we don't like. Happy to help.
What's your contact info?
in my profile. Just replied above with an overview.
Would you be willing to share your model?

For me networking so far has been the only consistent way of attracting quality clients. I don't dislike networking at all but it takes up a considerable amount of my most finite resource: Time.

Therefore I'm always eager to hear about alternative approaches especially if those are more scalable than me personally having to put in time for networking efforts.

I'd be very happy if you sent me a message about your model, my address is marius@tolud.com.

We tend to describe ourselves as "the tech team for your (tech) startup", our goal is to be a one stop shop for tech-heavy startups and companies.

We work with Node.js, TypeScript and React/React Native. We've created a bunch of quite unique libraries and components that help us develop a lot faster and with less bugs (I can get into detail over e-mail).

The company is new (founded in March 2017) and so far I just used my contacts from the past to find us work (7 developers, 2 analysts/consultants), but this source is quickly drying up. Right now we're finishing our website and my plan is to go to some meetups, but I'm not entirely sure it's going to help and I'm afraid I'm going to lose time.

Our biggest problem is that we don't have much of a portfolio, only a bunch of internal applications we can't really show. Clients I found online don't trust us and/or want our services extremely cheaply, which is a problem because our developers are senior guys that charge a bit - but that was a deliberate choice, we want to make high-quality software.

You can find my contact info in my profile. Basically in short, what's worked for me is side project marketing:

- Create a tiny free product, that solves a tiny problem for your target customer - Promote your side project. Since it's free and frictionless, it's so much easier to have it spread on the internets. - On that side project, there should be an email subscription form somewhere and your main service as an upsell.

This exact model is pretty much all we do and it's resulting into at least 5+ figure each year worth of client work. We spend zero time networking or doing all sorts of stuff that is time consuming.

That's a good idea. I considered something similar, but different - I merely thought about having the SaaS app as something to point out as a reference project. Thank you for your time!
really like the idea of your company - hope I can use it one day!