Building a new Marketplace – Part 1 of 4

If I had a dollar for every time I saw someone trying to build or clone Flippa I’d be a rich man. (Unless it was an Australian dollar, but even then I’d still be happy  :) )

** Reading Time for this post – Approx 10 mins

Since January this year, I’ve counted over 200 different requests on Odesk and Elance alone; there’s likely to be a fair amount of jobs scraped and posted on other sites bumping up those search numbers, but without a doubt the topic is hot one, especially on forums where ambitious entrepreneurs debate the task.

In May I touched on the basics of what features a new marketplace offering websites for sale would need in order to compete, in the article ‘We need a Hero‘. Since going live that article has been one of the most viewed on this blog and I’ve had countless hypothetical discussions with people on what it would take to get the job done. In the interest of avoiding an anticlimax, there is no solution guaranteed to work – well not here at least, but I am going to write on what I felt has been the conclusion of what I’ve discussed, with many people, ‘serious players’  and novices alike.

I’ll have to break the post down into four parts comprising of

1. Defining Strategy, two Marketplace Models and Site Policy

2. Marketing a Marketplace and achieving Critical Mass

3. Business Model, Financials and Due Diligence

4. A Big Resource List

The resource list will cover every potential resource someone would need to get started; it’s still being completed so please include anything you think is crucial in the comments box. I’ll also put everything into a pdf at the end, to making offline reading easier. If you haven’t already subscribed to the RSS, please do, to ensure that you don’t miss a post should a section get buried.
I know the majority of people who read this blog have no intention of ever competing with Flippa, but I hope it’s interesting reading for you and gives a good round up of using strategy to compete in a crowded segment.

Disclaimer – Please remember that all the ideas here are simply my opinion – not proven fact!

We don’t need another Flippa. We need another Marketplace

Flippa isn’t broken – don’t fix it

Trying to create another Flippa is pointless right now. There are occasional flaws with the software, and some issues with fraud and non paying bidders, but in the grand scheme of the amount of people who use the site, it’s minor and probably below average for a marketplace of their size.

Personally, I really like Flippa; they do an incredible job of serving a niche. Picture this – no Flippa, (and no Flipfilter) would mean anyone wanting to buy a site having to trawl through pages of forum listings on Digital Point or the Warrior Forum to find the (rare) diamond in the dirt. And if you think you’ve got fraud problems now, try buying a $15K site through a forum and see what happens to your bank details!

The satisfied majority are never the ones that speak up on forums about their dislike for a service, but potential competitors take this as a sign that the existing marketplace is broken and another (identical) one needs to take its place. This isn’t the case. Flippa is mass market because they are the largest in the market, and got this way by time served; all the hard work put into building a community in the form of Webmaster Forums and then Sitepoint earnt them the position they enjoy. By choosing to be mass market too, you lose any competitive advantage you can gain by being small and able to appeal to a niche.

Define your strategy before you begin

Rather than compete head on, two viable models have surfaced built around the main two needs of people who build and sell sites

Specialise in Starter Sites (Model A)

For anyone unfamiliar with starter sites, these are new sites with no traffic or revenue and are arguably ‘ready to work’ templates.  Many refuse to acknowledge starter sites as a business for sale, and those who sold them had recently come under attack. Needless to say the market does exist, and is likely to grow at a much quicker rate than selling established sites will, simply due to all the new people looking for alternative ways to supplement their income in 2011 and 2012.

Although the bottom fell out of the market in May / June this year seeing average starter sites prices drop from $349 to $127 to $whatever-you’ll-give-me, there is still a market for a cheap and quick solution to people wanting to start in internet marketing, but lacking the tech skills.

Specialise in Established Sites (Model B)

Aiming for the opposite end of the spectrum to model A, specialising in established sites would mean having an entry criteria for newly listed sites. From the stats we’ve collated across all four marketplaces, activity seems to naturally band at the following levels:

The sweet spot seems to be sites with a price tag between $500 and $7,000 which from the data we have are typically:

• Earning average revenues (gross) of between $60 and $820 per month
• Attracting more than 400 unique visitors each month
• Do not necessarily have any Page Rank and do not show any correlation in terms of links in Google or Yahoo
• Are over three months old (which I guess rules out a lot of click fraud, self purchase inflation, WSO flips etc)
• Are top level domains like .com, .co.uk, .net  and .org (although there are anomalies like .info .ly and .fm)

This is not the definition of an established site as one doesn’t really exist, but in a situation where you would have to ‘pick a side’, the criteria above would be a good place to start.
The drawback with having a criteria, is that someone would have to check the sites being listed and verify they meet it, which raises some issues, but this is covered in more details later on in the section entitled ‘Pareto’s Law’.

Policy and Security

The two models will need different policies but with either sitting on the fence and allowing the marketplace to self govern will not be an option. There will be some similarities and shared policies between the two marketplaces:

Open Viewing, Closed Buying

From a search and seo point of view, having your listings indexed by the search engines is crucial so viewing listings has to be an open process that doesn’t require people to register first. Bidding, buying and messaging should however require user registration in order to force the user through a security funnel. As a developer you’ll have to straddle the fine line between having the right amount of info to guarantee security in your marketplace but not scaring off potential users by demanding too much information up front. Studying what seems to work for other successful marketplaces, the following structure seems practical

Look for patterns, not pieces

Paypal originally began life as a security implementation for handheld devices but grew into the company they are when they became the most sophisticated company in detecting small volume online fraud. As an individual, I hate Paypal (but still frequently use them!) but I have a lot admiration for the company as a business. They learnt quickly that detecting fraud wasn’t about looking at individual incidents but overall patterns, for example one or two small purchases followed by one very big purchase of a fast moving electronic item that was easy to ship.

The bad news … no off the shelf software is likely to be sophisticated enough to work out patterns and detect fraud, but a policy can be developed with a little common sense. For example:

• Check for the obvious like different logins from the same IP address. Set up a script to alert someone when this happens for further investigation
• Setup keyword filters in private messages  to alert you to people potentially spamming members or soliciting services away from the site. There are also several ‘sentiment engine’ plugins that can be trained to parse text and identify patterns in the language, flagging it as suspicious based on your criteria.
• If you process payments through the site (not advised) ensure you verify every new payment method added, and set your gateway to the highest possible security setting (AVS, 3d Secure, Post / Zip Code and address matching) to reduce the possibility of chargebacks.
• Offer stepped security to avoid annoying genuine users. For example, a new user with a hotmail address and postal address in Uzbekistan (apologies to all my Uzbekistani readers) should not be treated the same as a new user with a US address and an email from an academic domain (alternatively, you can download George Bush’s "if it ends in ‘stan’ then nuke it" WordPress plugin)

Realistically, fraud will happen; with a little basic knowledge, you can create several Paypal accounts in a couple hours so clearly no one is immune. Your wisest strategy is simply to make things as difficult as possible for potential fraudsters and reduce the chances of them trying.

Mediation

Failing to establish a process for mediation BEFORE you begin is professional suicide, as you will need to react quickly should problems occur to avoid your users losing faith in the service.

Your policy should cover what you will and what you won’t intervene in, for example you may decide to solve a problem where the buyer has made a purchase but the seller not responded by refunding the buyers money and cancelling the transaction, but a problem with a site not having the revenue it claimed it would is a more complex issue. By knowing what you will and wont deal with, you can inform users before they use your site in addition to having a framework should you outsource the task in future. A good example to study is Elance.com.

Exclusivity

It’s a fairly common condition of most marketplaces that you should only list your site with them and not advertise it anywhere else. Logically, this makes sense especially if you’re paid on a completed sale, as a sale elsewhere will mean no revenue.

Potential start-ups have suggested not placing an exclusivity clause on their listings as a way to tempt people away from the larger marketplaces without having to take a risk (so the user would be able to try them out and still list with Flippa for example), but this will not work. Whilst you may be happy sharing the listing on your site, Flippa / Digital Point won’t be, and the users listing will be cancelled on their site. Long Story short – demand exclusivity from day one!

Adult Content

The debate over whether to include adult sites has been an ongoing one. Allowing adult content requires a lot of human policing and potentially a lot of opportunity for missing unsavoury images unless you happen to be very good at looking at porn. This opens up several legal pitfalls as well as moral ones (in the case of illegal porn, not porn on the whole) and is probably best left alone.
The adult internet sector tends to be incredibly savvy in internet marketing and SEO (they were the first to really utilise video marketing, CPA to unlock content and pay for access content streaming almost a decade ago) and so a marketplace for adult websites would be well received. However, I doubt the potential revenue would be worth the risk overall.

Specifics Model A

As a market for starter sites your transaction amounts will be lower and so security will be less of a concern than with Model B.

Policy should be relatively relaxed, as many of your potential users will be new to internet marketing and may require some coaxing in order to gain their participation. You will probably have to develop the concept of acceptable risk; many of the protocols you relax will open up the opportunity for small scale fraud, but the numbers are likely to be low. (Apparently, even Fiverr has fraud – go figure!)

Specifics Model B

As the primary user of this marketplace will be more experienced buyers and sellers looking for mid ticket investments, stricter enforcement will need to take to ensure the security of your marketplace. This will include:

Development of a trust and Feedback System
Flippa got this correct in theory but the idea still lacks finesse ( beyond the scope of this article). Ideally bidders should earn points for

• Linking a Paypal account
• Confirm a bank direct deposit
• Postal pin verification (phone verification is not optional)
• Completed transactions
• Length of time the account has been active (active = listings viewed / posted, questions asked etc)

Additional ID verification through passport / photo driving license and utility bill

This will require some thought as it will inevitably need ‘human power’, making scaling the business more difficult. This isn’t recommend for every user, but not all registrations should be judged equally. For anyone who has ever used Ebay, Paypal or Google and has consequently been asked to fax (it’s 2010 – who still uses fax??!!) their verification details you’ll know how annoying this is. Asking users to manually verify should only be done as an extreme measure and probably one best left for special cases and mediation.

Next:

Marketing a Marketplace and achieving Critical Mass

(This link will go live once the post is published. If you haven’t already subscribed to the RSS,, please do, to ensure that you don’t miss a post should a section get buried.)

Popularity: 14% [?]

Related Articles:

• An Evolution in Selling Internet Businesses
• Building a new Marketplace – Part 2 of 4 – Marketing and Critical Mass
• Feb 2012 – Latest Website Acquisition Opportunities from the A-List
• Flippas Price Increase and the numbers behind Brokered Sites
• How to find Websites for Sale (35 Sites Ranked and Rated)

Tagged as: Flippa, website marketplace, websites for sale