I guess I really should have called this “How to spot a potential dud that will take your time and money and eventually give you little in return before the other buyers do” but it wouldn’t have made that great a title.
This article is a guest post in two parts, the first dealing with a simple system of spotting potential problems with a site you’re considering buying, and the second part deals with what you can do if you’ve bought an asset and you’re beginning to see traffic drop. We’ve also included a live example so you can see it all at work.
If you buy and sell Internet Businesses either long term, as your primary source of income or just occasionally, trading up to get you closer to where you need to be, at some point you’re likely to get burnt. That experience should prevent you from ever making the same mistake twice, whether it’s buying a business model with no longevity or missing some type of inflated income scam put in place by the site’s seller.
You don’t always have to learn from direct experience though; furthermore, when the amounts you spend starts to creep past $5K, then learning by mistakes simply isn’t an option. This is where going beyond basic due diligence will sometimes save the day.
This is a guest post from the illusive Steve.
As a fellow data nerd, Steve is the man behind a service to pinpoint potential problems in a site’s link profile with the goal of preventing it being hit with penalties, or to help recover from them if it’s already too late. For this article, he’s agreed to give us an insight into how they spot potential problems with a site through forensic link analysis.
This is something you should do before committing any serious amount of money or time on a site; it’s quite common to buy a site and spend time creating new content and doing all the right things but never seeing any major results, simply because of mistakes made by the previous owner.
Over to you Steve …
Looking for penalties and problems using Google.
The history of a domain is crucial if you’re thinking of buying an established website, and it’s the bit many are often guilty of overlooking.
There’s several potential problem areas, but ultimately you want to ensure:
- The domain hasn’t been subject to any penalties from Google
- The traffic claimed by the owner is correct
- The site isn’t likely to drop positions in the future, as a result of any historical SEO problems.
In this article I’m going to do a case study based on a set of domains that have just sold at auction for over $17K and ended with 38 bids. You can see the result of the auction here – https://flippa.com/auctions/2841319/
First, a disclaimer: these results are opinion based and I do not claim that there is any manual Google penalties specifically applied to the domains in question. Instead these observations can highlight issues on any domain in general. In other words, should it all hit the fan, please sue Justin and not myself.
There are some basic ways you can look for penalties or problems with a site’s domain.
In this case study, the sites we’re looking at all have three keywords in the domain name and are part of a set of 5 websites. In this climate following Google’s Exact Match Domain (EMD) filter in October, it’s much more dangerous to have only keywords in the domain, since it’s possible that these types of domains could be subject to the filter and removed from Google.
You can check the keywords used in the domains (without spaces), and if the target domain does NOT rank at the top of the first page then it’s likely the domain in question was recently compromised by this filter.
The EMD filter only targets domains with 100% keywords; if there’s esoteric brand type characters (like aaakeyword.com) in the domain name those domains will usually be safe from this filter.
Arguably, the two word search terms like “Hunting Equipment” are where the real traffic is.
If the domain fails to rank for the competitive two word terms, then it’s crucial to discover if there’s any significant traffic for longer tail terms. You can do using Google’s traffic estimation tool.
Digging deeper, a quick check of one of the product pages listed on the homepage, brings up a url optimised for Bushnell 7 Day SKI FX Weather Alarm.
When we search Google for that term, the target site is nowhere in the top 100.
Disturbingly I can see many others competing for this long tail term and this is indicative of two potential problems:
1) The site is offering generic products which are ubiquitous across the web, and most likely is full of duplicate content with many people using the same manufacturer’s descriptions.
2) The obvious problem here is that the site is not ranking for one of its most prominent long tail terms. This is essentially enough information to tell us this domain is compromised for some reason or another.
Digging deeper to spot potential problems
I’ll continue to dig in the hope we can see some of the issues that are lurking in the background, and possibly causing future headaches for the new owner.
We’ll run a check for indexed on-page content, so I’ll take a full paragraph of text on the page and run it through a Google Search.
In this case, we see that our search produces thousands of matches including Amazon, which indicates the site is using spun and / or duplicate content. Worse still, the site doesn’t rank for that search phrase ANYWHERE; instead we have many other domains that are using similar content (which is where the content probably originated from in the first place).
We can do another search for the same paragraph, but this time I’m adding speech marks around the block of text, to check the exact match term. In most cases this would bring up the site, but here again it doesn’t, showing the gravity of the problem with the domain and the content.
The domain was auctioned as part of a package of five domains, and disturbingly they share the same IP address, meaning that if Google found one of the sites and applied a manual penalty for a reason common to all domains, the other connected sites would be clearly visible to the Google spam team who would probably apply penalties on all the sites.
Another consideration for this type of problem, is where sites use the same Adsense accounts or share Google webmaster tools accounts connecting all the domains.
Google can Sandbox new sites for competitive terms for up to four years for single word search terms, whereas 2 word search terms could be sandboxed for up to two years, and competitive 3 word phrases can take more than 9 months to rank. Here’s the revelation – it’s often not worth buying an ‘established online business’ which is less than about 2 years old and relies on, or will have to rely on organic traffic as its primary source of traffic.
Armed with this basic knowledge, any potential bidder should be able to see enough information to decide whether an auction is worth bidding on. It’s likely that the winner of this auction purchased on the basis of its current net profit, and without knowing their intentions for the site, it’s difficult to say whether or not this was a bad purchase for them personally. Arguably though, with no Adwords traffic reported by SEM Rush, this site relies on organic traffic for new customers and so going forward, a penalty or drop in rankings is likely to see revenue and profit decline.
There’s no crystal ball that can predict or anticipate every problem you may have further down the line, but knowing what to look for and how to check for it will significantly reduce the likelihood that your new investment will slip through your fingers.
A few tips if it’s already too late
If you’re considering an investment of over a few hundred dollars, then your due diligence should be extended to understand the asset’s links in forensic detail. We ran analysis on around 80,000 Flippa auctions over the last few years. Using Alexa API, Compete and Quantcast data around 65% of sold sites seemed to have suffered drops in natural search positions due to different algorithm filters and penalties.
Unfortunately, we don’t always have the benefit of hindsight, and sometime we can find that we’ve purchased a business which is only starting to show cause for concern.
We frequently work with SEO Agencies, Acquiring Companies and Brokers, to forensically examine a domain’s back links to discover which links can cause penalties or problems in Google.
We’ll look for things such as
- Links on banned domains
- Links on low trust domains
- Links on banned pages
- Links on the domains sharing the same IP
- Site wide links
- Over optimisation of the site’s main anchor text (Penguin)
- Pages not indexed on the target site
- Structural problems with website architecture which results in lost juice from links pointing but not captured
- Duplicate content issues
- Content quality issues (Panda)
- Historical issues no longer present but which could have caused the domains trust to be damaged in the past.
If you’ve already bought a site and you’ve spotted problems, it’s usually never too late to fix them.
Here’s a brief rundown of some of your options, based on what we do for our own clients:
- Link removal – Following a link audit, our first step is to start to remove the bad links. We have a bespoke CMS to manage the link removal process which takes most of the heavy lifting out of the equation, but you can also manually send take down notices to domains linking to your site. You can see an example relating to one of the site in the above auction here
- Bad neighbourhood links. In general a healthy domain would have about 20% of its links from a ‘bad neighbourhood‘ and 80% would be natural and healthy. Typically, we see search ranking problems when you have more than 30% of your links in a ‘bad neighbourhood’. This is important to know as it’s very rarely possible to remove all your bad links (site’s ignoring the request, incorrect WHOIS details etc), but you can still see results by bringing the total percentage below 30%.
- Reconsideration request risks. If you’ve been the victim of a penalty or a drop and you’re considering filling a reconsideration request, you should give this some serious thought. If the target domain is still getting decent traffic then there is a risk the spam team could apply even harsher penalties if the link profile is highly engineered which is surprisingly common. The other concern is if the domain owner has other sites which are still ranking well but that are connected together in Google Webmaster tools or are on the same IP, Google often look for networks and since domain owners are often guilty of similar practises on all their domains they sometimes hit those sites too if they have not already done so.
- The Disavow tool. Google have launched a disavow tool but we’re not yet hearing or seeing any success stories from using this method. It’s possible the reason this tool was offered is solely for companies who are suffering with genuine attempts of negative SEO. Google was getting a reputation management issue following the revelation many had when Google made it clear that bad links will hurt your organic search results, and companies were subject to penalties which could be engineered by others trying to harm their ranks. Since there is no evidence so far that this tool will help resolve any ranking issues, every effort should be made to actually remove all links which are likely to be causing issues instead.
Beware of the DIY Approach
Even SEO agencies often get this wrong, but one of the common DIY methods used to compile lists of bad links is to simply filter all the PR N/A URLS into a list, or select links based on a suspicion, then stick those into the disavowal tool.
This will most likely result in a total FAIL!
This type of approach will do more harm than good as simply filtering links by PR is a 100% guaranteed way to miss the real issues, and to compound the problems further by potentially disavowing the wrong links. PR N/A urls are ubiquitous across the web, and account for the majority of URLS out there!
Thanks for reading
- A step by step guide to spotting fraud and scams
- Analyse potential purchases and save costly mistakes
- Spot potential problems before they happen