I'll go first today and begin by saying that it proably depends on the type of site involved, the industry and other factors. But ... generally speaking ... I get alarmed when I see a website getting 70% or more of its traffic from natural search. Sure, it might be a testament to the SEO's skills, but what if the algorithms change next week? Where's all that traffic gonna come from them?

Several years ago, Andy Hagans wrote a great article about the idea of "defensible traffic" -- i.e., it's good to have a variety of traffic sources so you can withstand the loss of any one of them. I'm a big believer in that, so I don't like to see too much traffic coming just from search. You?

(Shame that Andy's original article is offline now. I wrote about it a little bit in this blog post of mine.)

I wouldn't necessarily say that there is such a thing as "too much search traffic," but you certainly want to diversify your traffic sources.  A ton of websites learned this the hard way when the Panda update crippled their traffic and made what were once extremely successful websites into virtual non-factors.

To me, there are two important takeaways:

• Have a diverse traffic acquisition strategy that targets not just organic search, but paid search, direct traffic, and referral traffic.
• Build backlinks that would give you value even if they didn't help your SEO efforts.  Some of my most successful links were valuable not because of the link juice they passed, but because of the clickthrough traffic and branding.

I'm not sure that there is an ideal breakdown.  So long as you are working all the channels in an optimum fashion, the breakdown shouldn't matter.

I've only had one site to get banned in Google (back in 2005, for cloaking, so hush) and it didn't affect it a bit, trafficwise, because it was a fundraising site whose traffic almost totally came from direct hits. I think their Google traffic was maybe 1% back then, if that, and they got nothing from anyone else. That is the only instance in my career where I have thought that search traffic didn't matter and because of it, I'd say that yes, you can definitely have too much. If you're dependent upon Google and they change the algorithm, you can get royally screwed as everyone knows. If you're not getting traffic from your links and getting direct hits, you need to fix that asap I think.

Something to consider these days is that direct referrals may appear to be lower with search referrals higher when it comes to brand terms these days. This is due to people using Google to easily get to sites they know about rather than bookmark or type them directly. Plus sometimes you just type the domain into the wrong browser box.

In other words, don't worry too much if your search traffic seems awful high, at least where branded terms are concerned.

Too much organic search traffic might be a warning sign that a site is vulnerable to a Panda slap. I've explored this in detail elsewhere (search 'panda quality sites') but in brief:  a high % of search traffic might often go with the following:

• few brand searches and low brand presence. (This will mean the site has poor brand data). Because site with weak brands get few mentions, links and type-ins = less traffic from non-searches site sources.

• pre-Panda SEO so good that it does well for searches it hasn't got the content to satisfy the searchers. Content frams did this in the long tail. But a quality specialist site can do this by performing well for populist searches. (This will give the site poor toolbar data). A specialist site can gets lots of great natural links (good for SEO) that don't bring much direct traffic and so search is a high % of traffic.

Poor brand and toolbar data is not too clever to have regardless of Panda. But they might be Panda signals too. We certainly know nothing easy-to-do can get a site out of Panda slap quickly. And there's little quick about turning around poor brand and toolbar data.

I look for strong internal referral traffic, either intrasite or intranetwork.  Web search services like Google, Bing, and others should not be sending a brand-value site more than 50% of their traffic.

In addition to strong internal referrals there should be good good referral traffic coming from other Websites AND from email.

A site's true value and quality is reflected by the extent of its traffic-producing visibility.  That cannot be measured through search and it sure cannot be measured solely through Website links.  There are other links that are just as important as links on Webpages (sometimes they are more important).

When i take on a new client or project and i see 80% search traffic is say "great, lets not focus on search then". I like to mitigate the exposure to massive algo changes in search and thankfully a handful of other sources CAN be just as beneficial for a website as a good ranking in SE's.

After their 'recovery' from Panda I decided to make my first post on DaniWeb about this very subject in hopes to educate some of the frustrated webmasters going their to figure out how to recover:

http://www.daniweb.com/internet-marketing/promotion-and-marketing-plans/threads/374629

There are also industries and segments where search traffic is lower quality than direct or socially driven traffic. One such industry seems to be video game media, the higher quality users that drive many of those sites find the site through 'starpower' not through search. it's typically recommended by someone highly respected in the niche that drives the high quality traffic, not a good ranking which brings mostly browsers.

Search traffic cannot be steady, so no matter how high the percentage of traffic is from the search engines, it will never be enough to rest assure from there and after. New search traffic and new ways to get targeted traffic that will convert must be constantly pursued, not only for the maintanance of the current traffic but the constant increase of it.