[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: pos hardware information


On 24 May 2008, at 17:50, Chris Travers wrote:

...
1)  Every POS system out there is subtly different from a workflow
perspective.
As it happens my former employer specialised in niches within the POS niche. The company was bought out by a firm of stock-takers shortly after inception, so its focus quickly became POS & stock-control for bars & restaurants, particularly for one national hotel chain.
They had an interface module that worked with the hospitality industry industry-standard Fidelio system, so that one could get a drink at the bar and have it already added to one's bill when one checked out 5 minutes later. I believe that most all hotels worldwide use Fidelio to manage their complex system of bookings, reservations, group-rates, discounts for different travel agents (so long as the agency supplies a certain number of customers), telephone & minibar billing &c &c. Fidelio manages things so that in some hotels the cleaners pick up the phone and dial a number when they've finished changing the sheets, and the system then logs the room as empty, so that it can immediately be allocated to new guests. Fidelio is a terminal-based application, and of course hotel-professionals love it because they know all the keyboard shortcuts, and Fidelio is a constant throughout their careers. I found the complexities of this system quite fascinating.
Having focussed on this niche, the salesmen subsequently went out with glossy brochures and a corporate profile (my former employer really thought the world of themselves) but took any sales they could get on the "we can make it do that" basis. And so the sales director insisted that development bodged the software to "also specialise" in cinema tickets and concession stands (popcorn, ices, drinks). They were branching into safari & theme parks when I joined them, which was another, entirely different, process. 


2) Every single instance I have ever seen for small businesses
(commerically or otherwise) is poorly engineered (LedgerSMB currently
included).    This includes every commercial product I have evaluated
for small businesses as well as every open source project I have
reviewed.  Sorry, all software in this area currently sucks.
This was very much my experience of our software. The stuff I saw, I wish TheDailyWTF was around back then, because we could have featured daily. We had reports that would crash the software completely if you entered the date in American format - in the UK 05/13/2006 is invalid, as there's no month after December. When I pointed this out to the lead developer he made sure the to-be-paid-for NG release of the software, due on site in a year's time, would check dates for validity before passing them to the 3rd-party reports generator, but he wouldn't add this to the current release tree, deployed at many sites.
I have to give the lead developer credit, though - the NG software did seem MUCH better that the predecessor that I worked with. The lead was competent and committed, and they had employed an architect that knew his stuff, too - although they were way behind with release, it did seem to come together. I don't know what you, Chris, would have thought of this product, but it was a VAST improvement on the suckage that I experienced.
Unfortunately I never really saw NG deployed, outside of a couple of sites, so I can't say what it was really like underneath the XP- themed UI. When the CEO retired, he was replaced by the CEO of the company's French subsidiary; naturally the company's flagship product, 3 years in development and only just out the door, was retired in favour of the subsidiary's flagship product. The consolation of this - and I really wish I'd been around to see it - was that several of the idiot pointy-hairs were removed, also. 


3)  This is business-critical financial data you can't afford to lose
(it is irreplaceable).  If you lose a day's worth of data, your cash
controls go out the window for that day too (cashiers could steal
money and you would never know)!  This is a *big* problem with most
commercial offerings.
We had a South-East London multiplex cinema that, for 3 nights on the trot, couldn't run the till-closure procedure. The cashiers said "here's my takings" but the site had no idea how much the staff *should* have taken. Friday & Saturday nights were affected, and the losses were *massive*.
We had another site which came to us in the middle of the busiest week of the year. "We're losing money!" they said, "we're selling stuff on the tills and it's not being charged to the customers' rooms!". At these kinds of sites it often takes a day or two for a problem to get reported, because the minimum-wage staff just don't give a monkey's proverbial. The first manager to be informed about the problem panics and gets stressed, but "doesn't know anything about the computer system" so he leaves a note for the next day's duty manager, who gets bullied into learning it. So the site doesn't have a clue how much money they're losing, only that people are getting their bill at reception and saying, "Oh, you haven't charged me for the £50 Ascot Week Happymeal Special." Obviously some customers who aren't being charged don't think to mention it.
A day or two of homework - during which time the site's still packed busy and losing money hand over fist - and it transpires that items sold in the restaurant have to be allocated an item group (PLU group?) when new items are added on the back office PC before downloading to the tills. Beers are group 1, spirits group 2, meals are group 3 and so on. Obviously it tells the operator (not the till- operator, but the supervisor operator) to do this in the manual, but this is an amateurish 2" thick ring-bound affair that no-one ever reads, even if the site can find it. Heck, I didn't know this, and I'm employed by the manufacturer! The group defaults to the invalid group 0, but anything allocated to group 0 just goes in the bit bucket when it's transferred from the till system to the hotel's (Fidelio, see above) billing system.
Since technical support could dial into any of the hotel sites (PC Anywhere at 33.6k, natch) the office joke quickly became that we were holding the department Christmas piss-up at one of them. We'd take a room, of course, and allocate the finest whiskey to PLU group 0. The ability to allocate an item to an invalid group - heck, that the system DEFAULTED to an invalid group! - was never fixed.
In this job I had to laugh, because it was better than crying. The company basically had two big customers who had 5 or 10 tills at each of 30 sites, and who accounted for maybe 70% of their business, but the sales director wasn't immune to the temptation of stitching up smaller operators, either. If you had 1/2 a dozen tills then I guess he'd be happy to take 20 grand off you, confidently reassuring you with the quality of the hardware our software ran on. We were an "IBM Preferred Business Partner", you see, and I only now know that anyone can be an "IBM Partner" just by filling out an online form at IBM.com. I'm an "IBM Partner", or was last time I checked, and IBM add the "Preferred" when you sell X units per year; I'd guess we sold 50 or 100 of their tills annually, which we bought for £1100 or £1200 a piece; the real money was in the support contracts, which were another £1000 per till annually. The tills were abused rotten by the cashiers, but somehow lasted years.
I don't have so much sympathy for these larger concerns, who could have afforded to buy a single till from each of half-a-dozen suppliers and tested each thoroughly before deploying them. Those nationwide retail chains fell for the corporate bullshit and look what they ended up with. But we had one or two customers who had only a handful of tills, and they (presumably) bought them thinking that they were buying the best. That was absolutely how my employer marketed themselves - a premium-grade leader-in-the-field. My former employer is beyond caricature - Dilbert regularly covered topics which were current office jokes - but I still see my own customers, on a slightly smaller scale, falling for the same thing, simply because they're ill-informed on the subject.
I mentioned in a post yesterday how one can't really know which POS systems are good & which crappy, because Amazon don't have any reader reviews for these products. I Googled yesterday for "pos review" and the second hit is http://www.possoftwareguide.com/ This website gives a very poor product impression to me, but some of the statements he makes are absolutely spot on: 

  I've talked to hundreds of retailers that went through 2, 3 or even
  4 different POS systems before they found one they're happy with.
  Those mistakes probably cost at least $10,000!
I can so totally believe that. You may read all of the above as the ranting of a dissatisfied employee, but I hope you can recognise that the following is good advice, whoever the bearer of it: if you're considering buying POS then do your homework carefully, take care when buying a product, and make sure it suits you. Chris is absolutely spot on when he talks about "workflow" and "business processes" - if you can't conceptualise what those are, or think they're just buzzwords, then you're at risk of making an expensive mistake. 

Stroller.