Search engine traffic, how much is OK?

[rencke]

A long time ago, I read somewhere that 30-50% search engine traffic is considered good. Anything above that very or very very good. And anything below 30% indicative of something being wrong with the site. Is this wisdom generally accepted?

[joshie76]

Do you mean visits to your site from crawlers or referrals from SEs?

If crawler visits then arrgh!! I only get about 20%.

[drbill]

Reneke,

If you are talking about referals. Then I would say if you are getting 30-50% of your referals a day from the SE"s I say you are doing really well. anything over 50% is good.

[rencke]

I was talking referrals. Thing is, I told this guy how to build and submit a state of the art 450+ page site and just got to see his log statistics - 80% of the traffic is from search engines and directories, topped by AV, MSN and Google. Just wondering how many beers he owes me.

[joshie76]

In fairness it could just represent a poor approach to non-internet marketing. Our referral Percentile goes up and down in rhythm with our external marketing operations (ie. Mailers, Conferences etc).

It's numbers that count for me, not fractions.

I'm still pretty sure he owes you a good few beers though :)

[rencke]

I see your point. It's 80% out of 4,4K referrals for March and for a very exclusive, very competitive and very very expensive global business service where the deals typically close between $10K and $100K, sometimes even a lot higher. The web is their only marketing vehicle internationally. I figured maybe free beer for life, what do you think?

[2_much]

wow I would think we're talking blue label johnny walker for life ;)

[bigjohnt]

Don't let me confuse this, but.. actually... going by a direct percentage means very little.
If you get two visits, and one came from a search engine - this is good? 50%? I read that one of the "demigods" of SEO was quoted that 15 - 30% of your traffic coming from SE's is about the most you should rely on. If suddenly SE's changed en masse, you're in trouble.
One of our sites gets over 90% of its traffic from searches (10% no referrer)- we do no other promotion for this site, this site lives and dies by the algo winds...I would not suggest this approach to a client.
The metric we like to use is what was the total SE traffic before we started the project, and what was it after - for measuring the traffic. But, the true metric of a commercial site is - what were the sales? Driving traffic itself is actually pretty easy. Driving BUYERS is the trick - not even PROSPECTS, or VISITORS count as much as the bottom line. All IMHO of course. ;)
[edit]
In retrospect, re-reading the thread, if you are driving that high of a percentage of the traffic yourself, you're doing great - or the cleint is doing poorly in other efforts. I guess I was taking a macro view, versus measuring an SEO's performance. As a measure of SEO performance, over 30% entitles you to bragging rights. Especially if the cost of doing what you do divided by that traffic gives a lower CPC than their other efforts. In my world, its all dollars and clicks.

[mivox]

We get almost half and half bookmark visits and se referrals, but very few people convert into online sales (almost all of our sales from the website have ended up with people calling us and ordering over the phone).

So is 50% SE referrals incredibly stupendous? Not really... I'm more impressed with the bookmark referrals. Means people like the site enough to come back. Eventually, maybe it will all start turning into some online sales...

[littleman]

Good point Mivox, and I am glad Bob bumped this thread [webmasterworld.com]. Because there is an inverse relationship between domain quality and SE referral ratio. My content-free promotional domains get about 98% of their traffic from the SE's.

[rencke]

A lot of wisdom has been poured into this discussion. Thank you. Let me see:
- Bookmarking and bookmark referrals should be measured. Of course! Great idea. Not done today, will tell him to install favicons to measure bookmarking. But how do you measure bookmark referrals?
- Dollars and clicks. Indeed! Cost barely measurable in this case. Site built by inhouse by staff member during slack hours and hosted with dirt cheap ISP. Biggest deal to come out of it so far: A $10M quotation which seems likely to be accepted. The nature of the business is such that even a 1 to 1000 conversion rate will pay off handsomely.

My thought was that a high SE traffic rate indicates that the site is easy to find for targeted keywords. Assuming visitors like the contents, the site will be bookmarked and there should also be a steady growth in linkage, which will push it even higher towards the top, making it even easier to find. And so on.

[chiyo]

Ive seen a few people refer to a low % of search engine referrals as being good.

And apart from the great point on numbers rather than %'s they have a point. As already pointed out, a SE suddenly drops your pages, or closes down (eg: go) and your page views may decrease dramatically. They are, when you think of it, capricious beasts.

I would not pay a lot for a site domain if most hits were from major Se's but i would pay a lot if it had hundreds of relevant external links in. That is more substantial I think for the long term, and we could then use our SEO skills to improve SERP positioning later.

I still think SEO is 10% inspiration (and knowledge) and 90% perspiration and even more so these days where short cuts are slim.

[drbill]

I have to agree with BigJohnT, I have domains that only survive with SE traffic and I would not suggest that anybody rely on only this as sometimes when all listings are dumped you land on your a** fast with no traffic or sales.

Been there Done That. Always keep something else as a back up.

I keep other traffic in reserve to keep the sales and traffic alive when the SE's mess around with a good thing.

[DaveAtIFG]

I read somewhere that the average surfer visits a site five times before making a purchase. Those repeat visitors (presumably using a bookmark) are where the money is.

[rcjordan]

Here are the extreme-dm stats re uniques for a single page I've been tracking for almost a year now.

Links= 103,979 - 56.70%
SE's == 79,171 - 43.17%

But, this stat is deceiving because I cross-link heavily from other domains and subpages I own in the same (themed) network of sites which act as doorways for this particular hub. I'd conservatively estimate that slightly more than half of those generated from links are from my own sites and they in turn pulled in the traffic from the SEs. The adjusted figures would then be

Links = 53,113 - 29%
SE's = 130,037 - 71%

[mivox]

- Bookmarking and bookmark referrals should be measured. Of course! Great idea. Not done today, will tell him to install favicons to measure bookmarking. But how do you measure bookmark referrals?

A lot of stat programs I've seen have a referrer category called either "bookmarks" or "direct requests"... basically, I think it's visits that don't include an outside referrer. Either they typed the URL directly into the browser, they used a bookmark, or their browser is screwy. However, if your number of "bookmark/direct request" visits vastly outnumber the amount of "odd" browsers you see in the logs, you can safely assume (I think) that you have a lot of bookmark/typed URL visitors.

You can also track favicon requests through your 404 logs if you don't have a favicon.