Which Shopping Cart to use?

hsj2011

Thanks paddy, I will be going in tomorrow afternoon for a meeting with the MD anyway so will raise these concerns.

I guess the best bet would be to find out what they do with current orders (i.e. how they process them). If they process them manually, then I guess manually accepting the orders and adding them to the regular workflow as you suggest. If they use sage or anything like that, to a certain extent, that'd make life simpler because I know of ways to integrate online stores to sage.

Thanks again...it's more complicated than I initially thought but glad you've mentioned it

If you have anymore advice for me to think about, please do post cuz it's mega useful stuff

paddyrg

No worries, I only know this stuff from sheer experience (back when I did it full time it used to pay for 6-month holidays and learning to fly! I do miss the money sometimes...) - integrating systems is hard work! The problems really kick in when people try to get clever without really knowing the limits of the technology and systems. Things that sound simple on the surface often have layers of complexity that are very hard to automate, and any form of duplication is always a massive red flag - things always go out of sync. Always.

Even using the shopping cart software to connect to their ERP/Sage system sounds simple - but a few things to consider here as well. You can have these for free ;-) ...

Sage contains the company's financial information and maybe personnel stuff too - it belongs deep behind firewalls and security layers, unlike your web app which belongs on the extranet/internet - so you have to start managing a secure data feed process between the security zones. Sounds simple, but how do you do it in practice? Periodic dump files? Message queueing via SMTP? What happens to the web orders if Sage goes down or the SMTP fails because of some security policy? If you do a daily brain dump (stock levels, etc) from Sage by some means or other, what happens when the person whose job it is to do/check that goes on maternity leave or sick for a week? And what if some of those web orders go unnoticed and fall between the cracks? Big reputation damage. Or the person who updates the website prices from sage loses a decimal point and some pillock buys 10000 teak doors for 20p each to sell on ebay for £10 each? What happens with cancelled orders? How are they handled and kept in sync? What if the webshop says cancelled, but the ERP still bills and dispatches the order (this seems fairly common from this site)? And what about the legal wording/terms on the site regarding when an order is deemed accepted?

What sounds simple "oh yes, we just attach the web shop to your Sage system" is in reality a considerable job for an experienced team of mixed disciplines (business analysts and techies). Beware of anything that sounds too simple!!

The advantage of just using the surface layer of a web shop to generate orders which get emailed to a human to deal with is that humans are really adaptable and can spot problems where computers simply obey. The teak door order would never get past any sensible human, but the ERP would just process it and send the confirmtion email, and maybe not even tell anyone what it was upto as it sent the packing order and dispatch trucks automagically, because the packers have no need for financial info... The human catches that and sorts it. Somebody posts here asking "the website said 20p but they wrote saying "sorry fella don't be so stupid", can I force them to fulfil the order?" (Again look in this site for plenty of examples), and get told that the company has not completed the order and does not have to do so.

War story - I worked for a VERY high-tech company once, and they had computer systems for delivering peoples payslips electronically. Sounds like a sensible idea to save paper/cost. But the way the system worked was a huge security risk, and would break every 3 months as group policies changed the password requirement, prompting 5pm Friday HR calls. To mitigate all this, complex systems were built up, support helpdesk processes were sponsored, and the server was (expensively) isolated as a security risk so lived in a weird corner of the corporate network. Then the server itself got old, and out of warranty, so alternatives were considered. One of the factors we discovered in research was that people *printed* their payslips anyway ( I know I would have if I was paid via that system ) - so all that happened was that we shifted the printing cost to a more expensive per-sheet printer network (colour printing instead of bulk mono) and had a failing security hole costing tens of thousands of pounds to support. I calculated we could reduce the cost by more than 75% by just printing payslips centrally then delivering them via internal mail. Fully automated solutions are not always the right answer.

Savvy_Sue

paddy the payslip story is very salutory. I find myself saying at regular intervals "just because it CAN doesn't mean we SHOULD!"

paddyrg

Savvy_Sue wrote: »

paddy the payslip story is very salutory. I find myself saying at regular intervals "just because it CAN doesn't mean we SHOULD!"

The horrific thing is that this kind of war story only comes with experience. When I was younger, grreener, I thought the world wad crazy for not being more automated. I would automate any task I could think of. But I slowly learned that because the world is organic, there are always exceptions to elegant rules, and elegant rules are what you automate... this means patch code to cope with this or that exception, and that is when you build in bugs.

Another war story, and a shameful one. Big publisher, small in-house facilities and fulfilment team who had to bill other cost centres in the company's "internal market". This was pre Y2K, and as the foundation of their billing system would not survive Y2K, I was assigned the task to upgrade it to Access 2(!). And I did an awesome job, simple GUI, Y2K-proof, and imported all of their old data for them too. I rocked. But I got a summons from the very important people somewhere in the company furious about my replacement system. They accused it of doing sums wrong. This seemed unlikely (there wasn't even VAT to consider when rebilling other cost centres, so the sums were VERY simple) so I had to go through piles of paper to work it out. And sure enough, my system showed differnt values for about 15% of 'invoices', and mine were always higher, by inconsistent amounts. Turned out their old system was wrong. They had underbilled by tens of thousands of pounds over the years - if they added an item to an invoice, the totals didn't recalculate, but of course on mine they did. But, importantly, it was my system that was wrong because "you can't change history". I had to rewrite it to bake in all the old errors, and it became another horrid dog of a system instead of the pure, automated thing I'd envisioned. They never apologised for my saving them from future underbilling, and I learnt a big lesson about making assumptions!

So now I am much more wary of taking what seems like a rule and enforcing it, there are always secret exceptions which need accommodating. Humans can do that well, but as soon as you try building all those business rules into a system you're chasing your tail! And if there's already a working system, piggy-back it!

raineshoe

I use Romancart and have done since I started my online shop 4 years ago. You can start with the base package and move up as you require more facilities which is what I did. I now have the top level cart.

Support is relatively good too.

pitkin2020

No offence (and I don't know your skill level) but your talking about shop carts etc like you don't really know the ins and outs of them yet the company who turned over £2m wants this to be a success. If you don't set something like this up correctly from the start you may aswell not bother. If the company is trying to expand into the online market then you should get a pro in to set it up exactly as they want it rather than a bedroom designer to get just something up and running. If the company was only turning over 30k I would say just get a half decent site up and build on it but your talking a different kettle of fish. You wouldn't buy a fleet of Ferraris then get your mate who has a toolbox and some spanners to service them would you ??