Archive for June, 2007

This week we hooked YuppieChef up to Jump Shopping. As a merchant you create an XML file with all of your product information (name, description, price, picture etc) and point Jump to it. The Jump server downloads the file once a day and keeps their comparitive shopping database up to date. Visitors search or browse the categories for stuff to buy, and if they click on the link to "view more on YuppieChef", we are charged R1.

We were impressed with the personal support that we received from the jump guys, and our products are displayed nicely in their listings. We are not expected hoards of traffic because our brands are not regularly searched for, but the lack of setup or monthly fees makes listing here a no-brainer. We have also created a banner that is displayed randomly on the site, again with a R1 cost per click. It is great to see a South African company offering CPC advertising options, unlike many who still only charge per impression, or worse, like Ananzi, a flat rate per month.

So true:

Thanks to XKCD.com

In the distant past of Web 1, the adage "Content is King" drove the internet. You couldn't just be an online shop - you had to also be an online encyclopedia. It was classic bait-and-switch tactics - bring the visitors in with tips on how to teach your dog to roll over, then at the last minute ask them to buy your new range of dog food. Hey, you're the expert on dogs, of course you're the expert on dog food too!

The Eisenberg Brothers have this to say in their book, "Call to Action":

Google and I are circling each other around the ring, and I’m trying to spot his weaknesses so that I can pounce.

In the last month Google has sent YuppieChef 665 visitors. That’s not a bad number in itself, but unfortunately our legacy of having recipes on the site means that a lot of visitors came looking for “mulva pudding” and “cooking game fillet”. There is a chance that those people might spend some money on cooking tools, but it’s more likely they’ll take their recipe and head back to the kitchen and use what they’ve got. We need more targetted traffic!

Currently (18 June 2007) on Google, if you limit the results to South African sites only and search for “KitchenAid”, we came up 5th, behind a semi-competitor and an auction site. Try the search here. There are many theories about how to improve search engine rankings, but “inbound links” is probably the leading one - if lots of people link to our site then we must be important, and if they link to us with a particularly word then we must be important for that topic. So to experiment a bit, and use this blog as a shameless promotor of our stuff, I’m going to write the next paragraph in a way that tells Google we’re important for the word “KitchenAid”, and some appliance names, and we have a great cafe category.

YuppieChef is the leading South African online store for buying the KitchenAid range of mixers, food processors, blenders and coffee machines.

Now we sit back and wait a few months to see what happens. This method is the same one that was used by bloggers to manipulate Google so that if you searched for “miserable failure” it would bring up George Bush’s page on the Whitehouse website at number one. They way they did it was to ask everyone to link to that page and use the words “miserable failure” as the text. Google rates the text in those links so highly that it didn’t matter that the Whitehouse page itself didn’t contain that phrase at all. If you do the same search now the page doesn’t appear any more, which is probably because the Whitehouse removed that page.

If you have a site or a blog and want to participate in a similar experiment, how about linking to some of the pages on YuppieChef like I did in the paragraph above, and we’ll see how well we do at beating the system?

And now, while I’m in the mood, here are a few more:

Buy Global Knives from YuppieChef and have them delivered to your door, anywhere in South Africa.

YuppieChef also stocks Eva Solo kitchenware, Trebimbi kids cutlery and Chef’n kitchen gadgets.

myadsl.co.za has an article about how difficult it is to build a profitable transactional website (ie an e-commerce site like YuppieChef).

Arthur Goldstuck is quoted in the article:

“Even the large retailers are battling locally because of the high cost of maintaining a transactional Web site, with issues ranging from secure and stable payment systems through to effective fulfilment,” he says. “In addition, site audits of online retail sites indicate tremendous strategic gaps in implementation, content management and transactional processes, with local online retailers showing very poor awareness of best practice in Web site strategy.”

Well, that's a relief. If the big guys are battling, then I suppose it's OK for us to feel like there is still a mountain of work left to do on our sites!

Filipe Serralheiro, application performance management consultant at Compuware SA, says transactional sites should form part of any retail business, much like shop fronts.

“Yet how often are IT managers asked to justify additional bandwidth, or marketing managers asked if the online site is feasible?” Serralheiro asks. “The tools used to measure performance, accessibility and - most importantly - user experience, are therefore vital to any business decision maker.”

In fact, I'd take it a step further. The IT manager, who spent his morning ordering new network routers and meeting with hardware companies, spends his afternoon meeting with a web designer, or worse, developing the site himself in Frontpage! It amuses me how senior management put websites into the same box as "computers", and therefore it lands on the IT guy's desk, when in fact a proper website presence encompasses everything from marketing to the supply-chain. If you don't treat your company's website holistically, it's not going to work.

“A high proportion of visitors to most sites begin a shopping process but then abandon their baskets during the checkout process – often due to the complexity of the transaction process,” he says.

You will notice that on our e-commerce sites we don't ask a user to register before making a purchase. We see this as yet another hurdle in the checkout process, particularly when they are required to come up with a unique username, a password that contains 9 characters and at least 4 digits, and then wait for an e-mail to confirm their registration, which often doesn't arrive. Having user registration does make the second purchase slightly easier, but that doesn't help if they never complete the first purchase!

On our sites, after the customer has added everything to their cart, we have a simple form to capture their name and contact details. At that point the order is entered in to our system and an e-mail sent to the customer, so even if they pull out then we can still make contact to push the transaction through. We use a cookie to store a reference to the user's information, so in most cases if they make a second purchase in the future they won't have to re-type information.

After the first transaction is succesfully completed we could ask the user if they want to make a username and password to speed up future purchases. This step will always be optional.

Half the respondents claimed to monitor and manage the performance of their transactional Web sites – yet only six percent have specific information about the end-user experience for each individual logging on. In fact, 61% of companies have no idea whether, or when, a client drops off a page - or how many times they might click the refresh button in a particular section.

Goldstuck says that local online retailers do have the data available to track the behaviour of their customers, but most have no idea how to monitor that information or make it meaningful in order to improve their performance.

I agree, anayltics is hard and time-consuming. We use Google Analytics and Crazy Egg on our sites, but we have only experienced a hint of their full potential. The massive online retailers can afford to pay a professional a few thousand dollars to tweak their site, because increasing conversions by half a percent will easily earn that back. For smaller sites like our's, a 1 percent improvement in the conversion rate would barely cover the cost of the phonecall to a professional analytics company.

Goldstuck predicts the market will not remain as forgiving as it has been until now. “Our research indicates the online shopper of 2008 and 2009 will be much more demanding, forcing online retailers to be far more strategic in managing their Web sites,” he says.

It's decision-time for South African companies - make a serious go at developing an online presence and be prepared to put a lot of money behind it, or don't bother. Sometimes a one-page site with your contact details can cause less damage than a poorly executed e-commerce site.

A bad day for productivity, a great day for fun at the office!

After just sneaking in to our targets (using a bit of Enron-style accounting practices), our foosball table arrived today. Shipped in from Brazil (reading the setup instructions was a challenge), this little guy weighs a ton, and looks like it should last for a while. We're not sure whether to capture it under "assets" or "liabilities".

Now that our office is taking shape, here's an open invitation to pop around for a cup of coffee and a game of foosball. Drop us a line.

Thank you everyone for your loving support, we really do appreciate it. Three cheers all round!

(Picture of Andrew on his second birthday, just after he received a gift he didn't really want!)