Ask HN: Why are POS systems all POSes?

I worked at one of the bigger POS vendors in the very early 2000s. The primary attribute of these folks seems to be how cheap they are. The 2 groups here would spend 100K a year for a Java Developer and then supply NO tools. The project I was on was still using Borland tools from the early '90s, including the Borland database engine. Debugging was next to impossible, especially when you used the networked "features."

What was really funny was they kept buying other companies to get their POS systems, but they could not manage to get one that worked as well as the old Borland based "thing"...

Java, fail

C#, fail

Modern C++, fail

It was rather amusing, especially after they chased off all the decent, experienced in the field programmers they had that they wouldn't allow to write the new stuff. They kept trying to buy it and failing.

The only thing I've seen worse is all the "video store management" software out there in the days before Blockbuster got big. Mom and Pop video stores were REALLY unwilling to spend any decent money on good software. I'm kinda glad about that as that's what kept me alive in the very early '90s, fixing all those Clipper, Dbase, Paradox and Flex "applications".

So basically it's the same reason most software sucks: because most companies suck. Thanks for your insight, I appreciate it.

Years ago, when I was helping out my favorite bar get back on their feet, I had thought it would be a good idea to develop a POS system. I wasn't capable of it at the time, I was just out of college and knew very little about running a project. Nowadays, I've got quite a bit of experience in inventory and ERP systems, but I also understand the market a little better now. For one thing, the people who complain the loudest are often the least willing to pay more for better.

This all got started because my business partner and I were doing some on-the-ground research into what makes a small business owner tick. We have an investor who has a Great Idea that is completely unworkable without a ton of data that these sorts of places are either uncomfortable providing or unwilling to put in the effort of providing.

And one of the constant refrains we heard in our interviews was that these people hated their POS system, knew it was partly their fault for not using it correctly, but didn't have the time or inclination to get into it more. We're talking owners of walk-in wine stores and pet grooming shops. Places where the owner is behind the counter, running the place, or directly managing the staff. They all opened their businesses because the business focus was something they loved, but they ended up spending most of their time wrangling data systems and tax compliance. It's a major frustration point for them and we'd like to do something about it.

Do you have any insights into the types of support contracts that these sort of places are running? I'm considering trying to sell my services as "labor for hire" to help small business owners clean up and figure out how to best use their POS. It's something I could get bootstrapped on right away and I think it would build relationships that could lead to more business opportunities. And it seems to fill the need that most of these owners have, in that they know that their systems are a mess but they don't have the time to do anything about it. But I'm concerned that the typical POS company has their customers locked into a support contract that would not allow a 3rd party access to the system.

Check out Sturgeons law... 90% of everything is crap. Some think he was an optimist! :)

You've hit a few nails on the head, especially with the loudest complainers being the least likely to spend money and the unwillingness to expend any effort.

What I noticed with the few clients I visited for the POS vendor was that the owners weren't busy, they were lazy and cheap. If your idea man's idea involves getting people to do anything besides say yes or no on the phone, I wouldn't bother unless you meet face to face with a few people that are really enthusiastic and prove they're willing to make an effort. They will just frustrate you and make you hate life otherwise. This is especially true for any kind of banking or financial business such as mortgage brokers or credit reporting agencies. I have yet to see a group of people less willing to spend money or to try to cheat you out of it. Restaurant and bar owners would be second on my list.

I didn't see much of the support portion of the POS business because the majority of the people were in a different state, but, more than 1/3 of the employees were in the training and support department at this particular company. A support contract was required for any big business simply because the customers wanted a neck to choke when something went wrong. Developers could intuit a fair amount about what was going on with the customers due to the problem tickets and CRM entries we saw.

If not for the support/training revenue I suspect the POS corp I worked for would have been out of business years ago. However, they had some huge customers that just wouldn't buy anything without a support contract. The mom-and-pop customers paid per incident and screamed and hollered the whole time. The large corps are locked into the POS support dept. The mom-and-pops are not, but... I don't think you want to deal with them...

Oh. God. Don't remind me.

Source: Former Rentrak employee.

I got into developing POS / retail management solutions in 2004 and founded GoFrugal Technologies. When I started, it did seem like a low hanging fruit, ripe and just waiting to be consumed

Close to 10 years and after engaging with 200,000+ retailers across India and with 20,000+ customers in about 20 different types of retail businesses - I have started appreciating the complexities of retail

The simple reasons why POS solutions are struggling to catch up fast with the times are 1. All retailers have the freedom to innovate at the front-end & there are many others who push down innovation at the front-end to the retailers : either because their customers are demanding or because the banks, technology vendors, government and competitors are constantly pushing the retailers to introduce change at the front end 2. Retailers want to service demand - come what may. The manufacturers continue to innovate on packaging, pricing, promotions etc. and no one waits for the retailer's system to be ready to handle these changes 3. Retailers have to keep the cost per transaction as low as possible to earn a profit. This, combined with the fact that retail is different every 100 miles, makes it hard to build a universal retail solution

I hope this explains why POS in most cases continues to be pain point for both retailers and consumers

What do you mean by "retailers innovate at the front-end"? And what do you mean by "innovate on packing, pricing, and promotions"? Shouldn't the shape of the package or the specific price have no bearing on how the inventory and sales are recorded?

I may be able to help here since I worked as a data analyst in the grocery world. I believe "retailers innovate at the front end" refers to the store specific details that are created by giving individual stores some freedom. For example, a product may be required to be on display at the front door of a store but store and category managers are given the freedom to choose which item it is. Big football game coming up near a NFL city? Maybe chips and soda.

As for packing, price, and promotions, it does have a bearing on how inventory and sales are recorded. Product packages, prices, and promotions change all the time and while it may be easy to keep track of sales, units sold, etc., the actual storage of this information is much more complex. Retailers need to structure their data in a way that allows analysts to determine how the sales were generated. In addition to knowing Coca-Cola had an increase in sales last month, innovation on packing, price, and promotions lead to technical requirements to know which specific UPC of Coca-Cola, at what specific price point, under what promotion, and at what location in the store drove those sales.

My comments are primarily based on working for a supermarket chain with over 2000 stores and about 30,000 POS terminals. Subsequently, I've worked with some smaller retailers.

First of all, managers and business owners buy these systems. But they don't actually use them. They will sooner listen to the sales guy's pitch than the bitching from the employees. When you pay customer service staff the minimum wage that shows that you don't value them and as a boss, certainly know more than they could possibly know.

I think you comments are based on 2nd generation systems, i.e. ones that are more or less a PC with cash drawer and receipt printer attached. The big manufacturers have cornered the market by claiming that "only they are big enough to take care of your business." Typically the cost of customisation is so high that POS screens, functions are not updated as business changes.

The 3rd generation systems, based on iPads and other tablets are shaking up some segments. Of course, the struggle against the above two biases.

In case your wondering, 1st gen was stand-alone cash registers, whether mechanical or electronic.

In any large business, e.g. a chain, the POS integrates with the ERP systems to optimise the supply chain and distribution functions. This integration, alone, snookers many smaller, newer entrants.

Of course, I think you have hit the nail on the head, it is CRUD, unglamorous and once you confront the customisation and integration issues, a lot of hard work. As Paul Graham would suggest, the hallmarks of a niche to profit from.

There's nothing fundamentally simple when you realize that the POS system is just one of the many front-ends for the store's inventory management, sales and cost reporting, barcoding and label printing, purchasing and tax accounting software. If you ever have a chance to help open a new retail store and actually see all the moving parts, you might want to take it, as you'll learn why POS software isn't as simple as it looks, and why the affordable ones are limited to small single-store establishments in specific industries.

Just handling taxes in a piece of software meant to be used in multiple locations can be hideously complex, and that's just 1% of what the system is responsible for. Remember that not only do the tax rates differ from one half of a town to the next, but what items are taxable do too.

You make a good point. I guess in the general case it's "not simple". I've built a number of inventory and ERP systems, so it's kind of my wheelhouse now. But yes, such systems aren't necessarily trivial without a rock solid design. I can build a goodish ERP system now because I have experience, but to someone just entering the field it has a lot of good looking corners that turn out to be awful decisions.

Do check out Vend (1) who are looking to change this paradigm.

Here are some thoughts.

The POS business is littered with devices that were designed for purpose by giant companies like NCR and IBM. Their primary customers were very large corporate retailers, the cost per unit and for development was prohibitively high and because these were giant corporate IT projects the results were awful for the end users.

At the lower end the retailers got those really basic electronic cash registers (ECRs)with lots of buttons and LCD displays showing numbers only. These were out of Japan mostly and usability was not a priority.

The biggest issue with these ECRs was that they were trying to solve too many end-user requirements, so they all sucked for all but the most simple transactions.

With the PC a bunch of DOS and then windows based retail POs systems emerged - mostly awful usability due to Windows and DOS. The best did made it very efficient for a particular industry. E.g. for service stations their systems were integrated with the pumps in the forecourt and a back office system, all of which likely linked to the oil company. But it all stalled for years until the cloud and iOS based systems came along.

Right now we are in a moment of rapid change, with the end-user experience becoming hugely better for both the customer and the retail assistant. Still plenty to come so watch this space.

(Source 4 years in Mobil Oil, 2 doing payments and POS projects, 2 in the field, and I'm an investor in Vend)

(1) http://vendhq.com/

I notice that Vend is cloud based. What happens when the internet connection goes down?

Your inventory is synced to the browser's local storage when you open the register, and sales get recorded locally until you're back online. It works fine without continual internet access.

"Vend continues to work even if your internet connection doesn't. Your data will automatically synchronize the next time you connect." See http://www.vendhq.com/tour - Online and offline section.

Vend has an awful lot of stupid bugs, and pretty inept front-line support. Even with reproduction steps and screenshots, they don't seem to understand tickets reporting bugs at all. It can take multiple back-and-forths before they escalate it to someone in dev that might try to fix it.

An example they just fixed after I handed them the bug and the fix -- they handled discounts on a purchase (i.e. a coupon, or F&F discount) by adding a special "Discount" product to the sale. In their database, hidden to the store owner, this "Discount" product was flagged as a taxable item with negative value. If someone wanted to buy a $10 item and handed you a $10 coupon in a state with 7% tax, the register would tell the cashier to give the customer $0.70 in change, and take $0.70 out of the month's tax you owe to the state. There were similar glitches in basic, critical calculations in inventory tracking, cost of goods sold, etc. There were oddities like if you used a barcode scanner, and had mis-typed a barcode when adding a product to inventory and lost a digit, the scanner would still pull up the product via the partial match, and it'd also add sales tax to that item even if it was marked as a non-taxable product (i.e. groceries in a state that doesn't tax them). So someone doing inventory could mess up your sales tax collection on that product for the whole year and it might never be noticed.

It's not a solid product, and the only reason we still use it is its integration with our Shopify store to keep the website inventory updated as items sell out in the retail store.

That last line is a serious dealbreaker. We run Lightspeed Pro POS at our two retail locations, and while it is great it still does some things stupid. On the bright side, it does sync with Magento for our website.

Is it still written in RealBASIC?

I had a similar complaint about Venmo. Yes, it was pretty, but dammit, they should have known to scrub their CSV export for newline characters and commas. There was just a lot of apparent sloppiness to the actual working of things under the hood, even if the front of it was clean and smooth.

I think my partner is currently using Vend for his tourist-town restaurant. I'll have to check it out. I've got a TODO task point to setup a "shop simulator", a basic Linux VM to see what exactly it takes to work end-to-end. I've been involved with all of the constituent parts at some point in my life, but I've never had a chance to see the big picture all at once.

And yes, Users are the worst people from which to get requirements. You can get a rough idea of where their problems lay, but you won't get any good ideas for how to fix those problems from users.

It's well known that POS also means "piece of shit".

I'm not sure how open it is for disruption, nor if it really is enough of a pain point to be worth replacing. Google has tried with Google Wallet, but who uses that?

One thing I have seen is that Girl Scouts are using some attachment on an iPhone to take credit cards at cookie sales. I'm not so sure that this would be useful in a restaurant setting or in a checkout process at a department store. It seems to me that these two use cases cover a large fraction of credit card purchases.

I though Wallet was more of a payment system, to compete against the likes of PayPal, then a Point-of-Sale system?

Probably because the people who buy the systems aren't the ones using them. Price and feature checklists become the main factors, rather than how suitable the software is for actual use. It's the converse of the point about Apple from that article that's been going around (http://stratechery.com/2013/clayton-christensen-got-wrong/).

I've been working on a mobile POS for vendors at markets and fairs for the last year or so. I built it mainly so that my wife could keep track of things when selling her stuff at conventions.

I think one problem is that everyone wants something different. I've been contacted by people who run festivals that want very specific functionality. It is hard to build that while keeping a feature set that most other people would find useful.

it doesn't have to be a pos! :) http://www.shopify.com/pos

obviously biased, because I work at Shopify. but this is a sweet product, and it works with your online store for inventory tracking, reports, etc.

I used to do Tier 4 support for a now-defunct (3rd Largest) grocery store in the US, and NCR has a stranglehold on this market. Every Self-CheckOuT lane with the same voice is probably an unbranded NCR SCOT. I've seen them in Home Depot, Loblaw's/Superstore, and even in some surprising locations (a nursery/greenhouse). Everyone uses the same terminals. It's been ripe for disruption for nearly a decade now.

LemonPOS[1] has been around for ages, is Free Software, and works great. It'd take very little effort to undercut the big guys in this market with a tiny agile startup.

[1] http://lemonpos.org/

Not my company but check out Kounta http://www.kounta.com/

thanks, I will.