However Trade and Exchange were the Trade Me of their day. They seized an opportunity to grab and extend the classifieds market away from newspapers, and in doing so created a large business and a headache for the papers.

Then Trade Me then seized the opportunity to grab and extend T&E’s business, and finally Fairfax completed the circle when they purchased Trade Me to get their old line of business back.

On the way through T&E and Trade Me the classified’s business grew tremendously in size and usefulness, to become the e-commerce engine of New Zealand.

There are a number of lessons to be learned:

1. Don’t be afraid of destroying your own business by launching a new one. If you don’t then someone else will.
2. If you accept that you are not capable of destroying your own business, then pull out your check book when it looks like someone else will. Once you have purchased them then leave them alone and let them grow independently of your own systems and processes.
3. Know when to lose – exit a market when you stop making money, and not a moment before of after.\

(disclosure – I’ve consulted Trade Me, Fairfax and WAN’s Quokka)

I’ve mentioned it here before, and it bears repeating: Trade Me is different.

Trade Me has captured the Motors market, is well ahead of the incumbent in Property traffic and just behind the incumbent in Jobs. They are also in flats ( and have been for years) and rental property.

Meanwhile Trade Me’s free listings and simple localization meant that they have always been in the local classifieds game – things like sofas and so forth that require a face to face transaction.

The overall effect was a business that left no space for the well executed Zillion and Finda to get traction.

So well done Trade Me, but the threat will always be there – screw up and someone will take big chunks of the business. This will happen, for example, if Trade Me attempts to increase fees too much or if usability is lost in a host of features. If so then Craigslist, Gumtree or say Zillion will be right there to pick up the volume.

Sadly for the competitive threat Trade Me is still following the path to usability, and while I’d like to see lower fees in tough times they are probably at the right level for now.

It’s the same at the top end for eBay with higher quality product sellers moving across to Amazon.

Indeed while a substantial percentage (say 40%) of Trade Me’s listings are new, there does remain a bit of a gap in the market for direct selling of quality new products. It’s the one Ferrit was so hapless at finding, and I’ve mentioned how you can take advantage before. The gap is pretty tiny, but between New Zealand and Australia (where eBay is also screwing up) there is plenty of scope.

I’m saddened at just how eBay have managed to lose the plot, but respect Trade Me and how they are still showing the way after 10 years.

Meanwhile once you arrive at eBay.co.uk each day, why not check out the rest of the site – this should increase overall sales.

For once this is something that Trade Me could actually copy from eBay.

There are 3 key questions that remain to be resolved though.

1. How do you select the products? They have to be real bargains from sellers that will deliver fast. It’s also an administrative burden to manually select those products, so it there some sort of auction each day for the space?
2. How do you compensate Trade Me for the home page slot when  the margins are so tiny? The home page slot could be sold for cash, but Trade Me does clip the ticket on the increase sales of the bargain anyway. Do you offer lower fees to get the prices even lower?
3. How do you manage the fallout from smaller sellers who compete (poorly) with the products on offer. This is especially bad if the front page sellers get a cut on fees, but bad even if their fees are standard and they pay a home page placement fee. This could even cannibalize entire categories as members wait for the deal rather than shopping.

I believe that Ferrit failed for three main reasons:

1. Very poor usability, with the site proving far too difficult to use, especially at launch. This turned away customers, and in particular the early adopter web community.
2. Very high costs – showcased by an expensive advertising campaign, but also due to outsourcing other critical parts of the business and it seems the general lack of a cost culture. The advertising driven growth was unsustainably expensive and seemed (especially at first) to be wonderfully misguided.
3. Poor business case – the value proposition to customers and suppliers didn’t ever really make sense. Customers could just go directly to retailer’s websites, while retailers were paying Ferrit for a percentage of their sales, when again they could just use their own websites.

All this was done in the context of Trade Me’s vast presence in the online retail space in New Zealand. I’m still not sure that Ferrit understood that Trade Me was a competitor, with new products accounting for about 50% of their listings.

NZHerald’s Adam Bennett quotes me saying the same in Ferrit’s failure tipped by industry. Forsyth Barr telco analyst Guy Hallwright is quoted as saying that Ferrit was so small that it took time for Paul Reynolds to pay attention. I think that is pretty close – though Ferrit was given a chance to shine over the 2008 Christmas shopping period first. In the scheme of things Ferrit is tiny and a distraction for Telecom.

“Telecom yesterday refused to say how much it had sunk into the business overall”

I’d be pretty embarrassed as well.

Here are the litany of posts about Ferrit – it’s been an entertaining ride. They are in reverse chronological order.

[Dec 15, 2008] Organic Torpedo7 beats inept inorganic Ferrit why a great website and slow, steady word of mouth is better than a lousy website and  millions on advertising.

[Nov 3, 2008] Paul Reynolds on Ferrit Paul answers two questions from Geekzone about Ferrit, sidestepping the question about profitability. I also question the quoted foreign online percentage of sales statistics – the share of online sales in the USA, for example, is only 3.3%. Meanwhile Trade Me’s gross merchandise sales are probably being under-estimated in NZ estimates of online sales

[Jul 24, 2008] Ferrit beaten by who? Torpedo7, that’s who Where specialty store Torpedo7 first comes to my attention after matching Telecom’s Ferrit in some Netratings statistics. It was a poor showing for Ferrit.

[Nov 16, 2007] How to manage your company’s online reputation Which praises Ferrit for responding directly to previous critical posts.

[Jun 19, 2007] Ferrit – ‘quality not quantity’ and [Jun 21, 2007] More on Ferrit Quality responding to a “puff piece” on Ferrit which did release some quotes, such as “says the company spends one third of its budget on IT and software development, one third on marketing and the other third on business operations.” That was telling given the huge amount that was obviously being spent on advertising.

[Jun 14, 2007] Ferrit’s Sale discussing the 20% off sale, which I thought was a “pretty good promotion, as it drives people to actually do a transaction, rather than just to visit the website.” However sadly the execution was poor due to poor usability – there are screenshots of how poor Ferrit was versus Trade Me and Apple.co.nz.

The 20% off sale was tipped in [March 17, 2007] Ferrit – 2 unsubstantiated rumours…., which also tipped that Telecom staff had received Christmas shopping period discounts on Ferrit – great for traffic. “It is obviously better to slowly build based on word of mouth and minimal advertising – like Trade Me, or Plan HQ (congrats on the launch chaps). Of course word of mouth only works if your site and business model is actually half decent.”

[Jun 5, 2007] NZ Retail stats: How is Ferrit doing? Not so well – this stats-full post showed that Ferrit’s traffic growth driven by advertising was less than Trade Me’s organic growth – this when Trade Me was pretty mature as well.

[Mar 29, 2007] Random Photos: Ferrit, Sydney, Fairfax, boats has a photo of the outside of Ferrit’s plush office. Not exactly frugal. (This was taken from a Fairfax Media office in Auckland. I was acting Head of Fairfax Digital at the time)

[Mar 2, 2007] Who on earth is designerexposure? well they “beat Ferrit into second biggest retailer spot” back in February 2007.

[Feb 22, 2007] Ferrit as a price comparison engine – not good pointing out that Ferrit wasn’t doing well versus price scanning competitors. I also defend the number of Ferrit posts in this blog, saying “I guess I am just expressing the frustration that those of us in the business and the internet industries are feeling as we watch a partially Government owned company invest and reinvest in a seemingly pointless endeavor.”

[Feb 20, 2007] Ferrit – $36m flushed… Talked about the implications of this quote  “Mr Brayham says Telecom expects to spend about $12 million on Ferrit in the next financial year, matching its contribution for the past two years.” I summed it up as “there simply is no way to make the numbers work.”

[Feb 20, 2007] Ferrit – It’s worse than I thought This post tried to get behind quoted December 2006 sales numbers, which were pretty poor: “Firstly only 1.2% of the visitors in December bought anything, secondly Ferrit makes “between 4% and 8%” of each sale. (Let’s be nice and say an average of 6%).  Thirdly the average order size was “more than $100″. Let’s call it $110.”
Peter Wogan, head of Marketing at Ferrit, responded to this post. “The original business plan was (and still is) to provide an online shopping option for new goods for Kiwis, believing that New Zealand in time won’t be much different to the rest of the world who are spending more and more online. Telecom is committed to the business long term and the payback will be over several years – probably somewhere between three and five. Our costs are right on track. We were a few weeks late launching Phase II but we’re happy with sales progress so far. You bagged us about the revenue we have generated since we started as a business, but we have only been generating revenue for 14 weeks (also in the business plan). If we don’t meet our targets then we work out what we did wrong and do something about it. We’re as accountable as any business.”

[Feb 19, 2007] Online retail: Ferrit sinks Where I pushed the numbers farther: “Ferrit’s Unique Browsers have also dropped away post Christmas [2006]. Given that Ferrit is on record as saying that 2% of visitors buy something, that’s an estimated 1,600 buyers in the first half of Feb (to the 15th), or 108 per day. Call it 3,500 buyers per month, at an average of say $10 revenue per buyer (I made the $10 figure up, but the average sale price is most likely between $80 and $120 from the comments in the blog entry linked above). That’s $35,000 per month in income.
With 40-60(?) people to support and the huge advertising budget this dog still does not hunt.
When will Telecom wake up?“

[Dec 18, 2006] Ferrit. Incompetent. #3 Where I tried to work out the value of each customer required given the very high levels of advertising spend. It was not good. “If a customer costs $20,000 to acquire then they would have to spend $400,000 to $660,000 each on Ferrit (@5% take rate) for Ferrit to make their money back. Before discounting for time value.”

[Dec 18, 2006] Ferrit. Incompetent #2 where I am a bit scathing of an NZ Herald article that is pretty soft on Ferrit, and also a bit scathing of Ferrit itself, and (my apologies) Ralph Brayham.
“In another article from the Herald, we see a classic quote from hapless Ferrit boss Ralph Brayham, who says that he “estimates that currently only about 0.3 per cent of spending [in NZ] is done online…”

Actually, as later in the article shows, NZ has about $1.5bn worth of online retail sales each year and as the article also points out, $60bn of total retail sales. That’s 2.5% Ralph – you are out by about an order of magnitude.”

[Dec 18, 2008] Ferrit. Incompetent The post that started it all off. I quite liked the title so I used it two more times. The incompetence was fairly mild – wrong spelling in an online advertisement. I guess getting the basics right is everything.

I now believe I missed the most important one: 

0: Introduce auto-extend on auctions. 

Auto-extend makes the auction last for, say, 30 seconds longer each time a new bid is received. Trade Me does this. It will kill the practise of sniping and make the auction site a level  playing ground again. Right now the advantage is to those that use sniping tools and the auction experience is lousy.

My current next three are from the previous list. 

1: Sell PayPal (and allow Google Checkout)

–> greater velocity of sales, greater trust, money in the bank

2: Charge for Skype use

–> immediate revenue lift from a captured market

3: Simplify and reduce all marketplace fees

–>increase in Gross Merchandise Volume, particularly from smaller sellers

Here it is – 2008 – the year eBay lost its mojo. I’d also call it “How not to run an auction site” or “Compulsory reading if work at Trade Me or Fairfax corporate”

They would have been much better off to move all marketing, development and management staff to customer service, but.. oh well. 

What a disaster year. Here are a few random quotes:

1. Fees were raised and sales plummeted.
2. Shipping rates were set below actual costs.
3. Listings were hidden if they were indexed at all.
4. PayPal was made mandatory and checks and money orders banned.

Sellers were banned from leaving negative feedback.

Site instability. … went from bad to worse. 

Best Match was officially named the worst sort/search engine on the internet. Search results were linked to everything but the kitchen sink and it took divine intervention to find anything. Entire categories of search results are still missing and many sellers reported over a 70% drop in sales which explains the pitiful sell thru rates.

I’ve never seen an e-commerce site nearly destroyed in such a short period of time.

I’m strangely still long eBay – simply because I think they they do have a compelling business. It’s just that they are doing everything to destroy it that makes me sick. I’d love to buy them and turn them around, but I am short a few billion sadly. I’m long in the hope that someone else with deeper pockets will snap them up. How about it Yahoo? Microsoft? Google??

“How is Trade Me a productive NZ asset?

The answer requires just a little bit of economics, and it is really quite amazing.

First – Trade Me sold for $750m, and the money was paid to New Zealand shareholders. So you can easily say they created $750m of economic value for those shareholders. Many of those shareholders are now investing their time and wealth in new businesses or charities based in New Zealand, and so the economic value creation goes on.  (1)

Second, Trade Me has build a new marketplace, creating value for new buyers and sellers. Let’s look at the (estimated) 50% of goods that sell on Trade Me that are second hand. 

We have 540,000 more Listings each week  

Trade Me has 1.24 million listings right now, implying that they have about 600,000 2nd hand items for sale. (2) The default action period is 7 days, so let’s assume that there are 600,000 second hand items listed each week. 

Before Trade Me the market was found in the free classified listings newspapers – Trade & Exchange. I’m estimating that the free newspaper classifieds had, in their heyday, about 20,000 listings per big regional paper, and they had say 3 equivalent big regional papers.(3) That’s about 60,000 listings each week. 

We have $5m more sales each week

Let’s assume that the Trade Me and old time newspaper listings sell at the same sell through rate of 25%.(4)  Therefore under the newspaper regime we had 25% x 60,000 = 15,000 items sell each week, and with Trade Me we have 150,000 items selling each week.  That means New Zealanders are now buying and selling 135,000 more items between each other than we used to. If the average sale price is $50 (excluding cars etc.) then that’s $6.75m per week of additional 2nd hand trade.

That second hand trade was not happening before, so where was it in the past? I suggest that the items that are now being sold were then sold in garage sales, or given away or thrown away – either immediately, or eventually. Mostly I suspect they were thrown away or sold for almost nothing at a garage sale. Let’s say $1.75m worth was given away or sold far too cheaply in garage sales (5) each week, leaving $5m per week of items that would otherwise have been thrown away.  

That $5m per week is say $250m (6) each year of revenue that is going to the former owners of those items – money that they would not have received before Trade Me existed. 

Moreover that $250m is an increase in recycling of items in New Zealand. That’s pretty impressive.

$250m a year is about $2 to 3.5 billion of net present value to New Zealand, depending on how you value the future flows. Add that to the $750m (7) from the sale and Trade Me has added $2.75-$4.25 billion to the New Zealand economy – let’s call it “at least $3 billion” in net present value. What’s particularly great is that all of this money has gone to household rather than big businesses.

But there is more.

Third – Trade Me lessens deadweight loss

Trade Me offers a much clear marketplace than the old car magazines (auotrader) and newspapers. We dealt with the newspaper market above, but the car market is a different story.

With Trade Me the number of people viewing each car advertisement has risen significantly versus Autotrader’s magazines, and the quality of those advertisement has also improved. Also viewers of that advertisement are able to compare all other listings for that same model of car (including recent historical listings) and can therefore derive a much better market price.

The market price is much more informed (closer to “perfect”), and so buyers are less likely to over pay, and sellers are more likely to receive a fair price. Thus Trade Me is removing distortion in the marketplace, and that distortion was economic loss. I’m not going to attempt to value that economic loss difference (or that for property), but I suspect it would be a figure on par with the $3 billion above.  

We have also not considered the new goods selling without rental and other overheads, the reduction in advertising costs and the ability for small businesses to form and start selling quickly and profitably.

What do you think? What is missing?  

Notes under the fold

The point of this exercise is to get the approximate  size of the economic value added, not to get it absolutely right. While I have access to some historical Trade Me numbers I have deliberately not referred to them for this post, nor for the same reason have I tried to make the numbers too precise. If someone wants to do so then it could make a nice journal article.

(1) You could argue, as I did at the time, that Trade Me’s shareholders could have got more money if they had waited. Indeed – they may have been able to sell for say $1 billion a year or so later. But then again we didn’t know whether a 2001-type market crash would come during that year or two (e.g. housing prices were already pretty high) and so the shareholders made the right decision and took the deal on the table. As things turned out they may have left a little behind, but then the financial events of 2008 could well have happened in 2007, and they’d be looking at a much lower valuation right now.  Meanwhile purchaser Fairfax acquired debt to purchase Trade Me, but Trade Me is one of the few well-performing parts of the business.

(2) It may be higher, but a lower sell through rate would compensate.

(3) Actually I have no idea how many separate geographic editions of Trade and Exchange there were, and what the overlap was of listings between editions. I am assuming 4-6 editions, but that 3-4 of them were fairly slim, and that 3 larger regional ones would also take account of listing overlap. If someone knows the actual numbers then I’ll update everything above.

(4) This is a big assumption. 2nd hand sales should have higher sell through rates on Trade Me as new sales would more often be for buynow=reserve and they have a cost price to beat before they will sell. A percentage of free newspaper ads are relisted each week, and of the ones that do not relist many will not sell. I am 90% confident that the free listing sales % was in the rather wide 15-40% range. Again – the point of this exercise is to get a rough figure, so we will go with what we have. If anyone has a fact based number, then let me know.

(5) A garage sale is a poor market, as there is only one seller, and buyers come in one at a time. This means that prices were often too low (too high and they didn’t sell), especially at the end of the day. A working motor-mower sold by my Dad for $1 springs to mind.  

(6) We’ll give the traders Christmas and New Year’s weeks off.

(7) actually it should be 750m X 2.5 years of return to bring it to present day values. We’ll stick with $750m given recent financial events.

They are not.

Over the years the site has increased in complexity and decreased in humanity, until the community feels obliterated and the only sellers are the larger professional ones.

In New Zealand Trade Me’s strategy is to keep things simple and focus on the person to person transactions. That is working.

Auction and classified sites make money by attracting visitors, turning those visitors into members, turning thse members into sellers that list items, selling those items and collecting a percentage of the value of thise items listed and sold. eBay is losing visitors and is lagging in items for sale.

Let’s look at two data points to back up the decline of eBay.

1: Pageviews: This disturbing chart came via Auctionbytes, and shows a 30% drop in US page impressions from last October to this October. That’s a massive sign of an unhealthy marketplace.

2: Listings Today eBay Australia has 1,113,647 listings located in Australia, eBay.com has 25.2m listings located in the USA while Trade Me has 1,255,082 listings in NZ. (That’s right, there are more actual listings on Trade Me than on eBay Australia.)

That’s 55 listings per 1000 Australians, 84 per 1000 US citizens and 305 listings per 1000 New Zealanders. That’s astonishing - 5.6 times difference between Nz and Australia, and 3.6 times between NZ and USA.

Now we do need to adjust for the critical difference – eBay charges listing fees while Trade Me does not. We used to say (and observe) that free listings sites attract 2 time the listings of non-free ones. That would still leave Trade Me at 152 listings per person – 2.8 times more than eBay Australia and 1.7 times more than eBay.com USA.

Go Trade Me. What’s going on?

To me the differences are the business versus personal focus, the advertising mentality and usability.

Business versus Personal

eBay’s listings skew strongly to the biggest sellers, who essentially run giant shops on eBay’s site. Over the years eBay has increasingly pandered to these sellers and the result is a giant shopping mall rather than a person to person market.

That has meant they are vulnerable to increasingly competent competition from Amazon & Yahoo stores, and from sellers’ own online stores. Traffic to those stores is increasingly via Google, and so the need for a single marketplace is slowly fading away.

The focus on businesses has also left eBay vulnerable to the economic downturn. This is a travesty as the auction model should weather storms better than almost any other retail. (e.g. beer and flour will do well). In tough times people should increasingly turn to auction sites to sell their unwanted goods and to buy second hand goods rather than new.

As eBay has increasingly disenfranchised the individual sellers in favour of the giant sellers they are being perceived as a store rather than a marketplace. Thus they are treated like all the other stores when ties are tough, and sales suffer.

Meanwhile the giant sellers are finding that other stores, including their own, offer far cheaper alternatives to eBay for selling. eBay has steadily increased fees over the years for both the site and PayPal, creeping up their take rate to unsustainable levels. The big sellers are not that upset to see eBay die as they have other alternatives away from the former dominant marketplace.

Advertising mentality

eBay is desperately trying to capture some of that disappearing traffic with a vast array of Google ads, and traditional media ads. Actually they spend about a third of their gross profit on marketing, while Trade Me spends nothing – relying on organic growth.

As a result people go to eBay via advertisements to buy things from the big sellers, and go to Trade Me because it is the marketplace where they buy and sell things. eBay’s formula is doubly vulnerable as other shops can simply place their own advertisements and steal the customers.

Usability

eBay’s sites are enormous, and they are complicated and difficult to navigate and use. Meanwhile Trade Me has managed to maintain the essential simpleness and usability, with a the cartoon-like nature.

eBay.com’s increasing push for “fraud protection” to the critically flawed Paypal payment system has made it harder to buy, and harder for sellers to guarantee they will get their money. With Trade Me payments are via bank to bank deposit, or, increasingly, direct credit card transaction.

Meanwhile eBay has, it seems, wildly over optimised the site for Google, and lost something along the way. e.g. Look at the page title of the eBay.com.au page displaying the results of a search for all items based in Australia:

I have no idea what this mens, but I didn’t come to eBay to buy women or men.

I have no idea what my phone number from 1999 (or was it earlier?) is, and I certainly have no way of answering a call to that number.

So I started entering my NZ number.

Firstly, the page  should automatically select “Provide new number” when you start typing text into the new number box. It didn’t.

Secondly when I eventually selected “Provide new number” another box rolled out, demanding the “answer to my secret question”. (I hate secret questions. Why should I give anyone else the keys to my life, when they already have a password selected by me?) My question was “fish”, and so I plugged in the answer “chips”, which I still am not sure was right,

Thirdly, the new number was required to be in USA format – so my NZ number was never going to work. I suppose I could have used a Skype-in number, but enough was enough, and I was locked out.

They lost me. I couldn’t login, and eBay was telling me that my business was not important.

I did manage to login just now on my own computer, (and change that secret queston), but my confidence in eBay has plummeted to new lows. I use more than one computer, so I can see a future where I am locked out, and cannot get in without this sort of debacle.

eBay has 3 great business lines – Telephony/video, payments and auctions. They appear to be doing the best job that they possibly can to destroy all of them.

But regardless, here are my top 10 suggestions for change at eBay:

1: allow google checkout and integrate it. It will provide a clean payment system (unlike paypal) and will force paypal to get better.

2: Simplify and reduce all fees-the penny pinching has killed the market and is driving people to amazon and elewhere.

3: cut the marketing budget by 90% – we all know where eBay is, so take a short term revenue hit but increase profits

4: reduce staff by 50%, especially marketing, product management and even development…

5: Stop feature creep and keep it simple, light and fast.

6: keep copying trade me’s stuff from new Zealand.

7: charge for Skype use, but trivial amounts and only for conference calls

8: increase customer service quality and staff quantity – put those marketing mbas in there perhaps.

9: Sell paypal. Someone else will make it work and grow, even if you can’t.

10: allow sellers to give negative feedback to non performing buyers-it is a basic building block of community policing.

Is there a chance of any of these happening?