One of the most frequent questions clients ask me regarding Twitter is "how many clicks will I get?" Despite the numerous SEO (and other marketing) benefits of Twitter, most of my small business clients just want traffic in return for their time spent tweeting.

More than how many clicks to expect, clients want to know if there are factors that can lead to more clicks. For example, does the URL placement in the tweet influence click activity? Is there some secret to the length of a tweet? How much do you have to post on Twitter to get clicks? In addition, there are a lot of small business owners I've spoken to who seem largely discouraged by Twitter (and other social media). I hear the argument that only the big brands with lots of followers can actually get any kind of click activity.

Most data out there regarding Twitter click activity suggests that Twitter is a great place for getting clicks, certainly more than Facebook. There is also little doubt about Twitter's reach – there are lots of people using Twitter and lots of people clicking on tweets. Unfortunately, none of the data I've found helps answer the questions from my clients which largely amount to: what should a small business expect from Twitter in terms of clicks?

Because I wanted to give my clients a somewhat better answer on these questions, I decided to pull some numbers and start figuring out the answers myself. I found the data interesting and thought I would share with the SEOmoz community as some of you might be getting these same types of questions.

Getting the Answer

Using Twitter's API and Bit.ly's API, I pulled in the numbers on a small sample set of 61,557 tweets, 120 Twitter accounts and around 47 million clicks. The accounts were of a variety of sizes (from just 15 followers on one account all the way up to 3.5 million followers at the high end). The accounts were selected from different industries – technology, media/entertainment, news, travel, recreation/sports, education, health and cultural accounts.

The main condition for selecting the Twitter accounts to use in the sample set was that they had to use bit.ly to generate most of the URLs in their tweets, they had to share their bit.ly data publicly and they had to tweet on a semi-regular basis.

It is a small sample set (given the vastness of Twitter) but I think there are some interesting insights to be gained and a lot of the numbers and trends are very consistent over these different accounts.

How many clicks should you expect? Do only large accounts get clicks?

The best way to measure this, that I found, is to look at clicks per follower. That is by no means the same as a click through rate, but it does give a good relative number of activity. That makes it easier to compare accounts at different follower levels.

The average click per follower for this sample set was 1.107%. In general, the clicks per follower got worse as the account size got larger. For example, accounts with less than 500 followers had a 3.28% click per follower while accounts with more than 100,000 followers had a click per follower percentage of .32%. Accounts in the 50,000 to 100,000 follower range did get above a 1% click per follower, largely due to a few URLs that received a lot of clicks.

Interestingly, only 23,714 tweets with URLs that were analyzed had no clicks. The majority of the tweets with URLs that were analyzed had at least one click. Also, 21.2% of the tweets analyzed,13,091 tweets, had more than 100 clicks. However, only 12.9% or 7,992 tweets had more than 500 clicks. The cool thing is that there were accounts with just 100 followers that were able to get over 500 clicks on a tweet.

Do more tweets with URLs automatically mean more clicks?

Some argue that to gets more clicks, you should post more URLs. However, I have heard social marketers suggest that this assumption isn't true. If you post too many URLs, they argue, you appear spammier and get fewer clicks.

To begin measuring this, we first need to know how many tweets had URLs. To arrive at that number, this data study took the number of total tweets posted over a certain period of time and the total number of those tweets that contained (trackable) URLs. This percentage of tweets with URLs was compared to the number of clicks on those tweets. To keep measurements relative, the comparison was done in clicks per tweet - that is, the number of clicks received by each individual tweet at different percentages.

It turns out, at least in this sample set, that neither theory was correct. Posting lots of links, and posting virtually no links, is not the best way to achieve a lot of clicks per tweet. Accounts that posted URLs in about half of their tweets got the most clicks per tweet. Keep in mind that accounts with URLs in nearly every tweet did receive the largest share of clicks. Nearly half, 53.20%, of the clicks tracked were from accounts that posted URLs 80% or more of the time and just over 20% of clicks were from accounts that posted tweets 90% of the time. However, accounts that posted URLs just 40-59% of the time were able to get more clicks off those URLs posted than any other account level.

Here is the graph reflecting all data collected.

In reviewing this data, the 40-59% range looked almost too big in comparison to the other percentage levels. In filtering out the accounts with significantly above average clicks, the same trends can be seen as shown in this graph. This data, no matter how it is filtered, seems to suggest you will be more clicks for your tweeting efforts posting URLs around half the time.

Does the length of the tweet affect clicks?

This one is important to measure in two different ways: words and characters. For each tweet analyzed, the length in words and the length in characters was tracked. The average character length within this sample set was 100.8 characters and 16.9 words.

Like with the previous graph, to measure this accurately, instead of looking at raw click counts this graph looks at a percentage of clicks per tweet at different character and word lengths. That way, the numbers are not skewed by a high number of tweets at a certain length and instead look only at the relative amount of clicks gained from a certain number of tweets.

When it comes to character length, short tweets tend to get more clicks per tweet than long tweets. In other words, it took only 82 tweets with less than 30 characters to get 384,957 clicks but it took 16,990 tweets to get 7.1 million clicks for tweets with 130-139 characters.

Interestingly, when measured in words there are more clicks for longer tweets. Still though, that upward trend for longer tweets is not as great as the amount of clicks per tweet for shorter tweets, even when measured in words.

Does the URL position affect clicks?

Related to length, how does URL position affect the number of clicks? Does placing the URL at the start or end of the tweet get more clicks? Like with length, this was measured in relative terms of clicks per tweet to avoid skewing the data due to a high number of tweets with URLs at certain position.

Within this sample set, tweets that put URLs at the start received more clicks per tweet than tweets with URLs closer to the end of the tweet. There is a slight upward trend for tweets that put the URL between positions 80 and 110 as well. However, that trend does not produce as many clicks per tweet at tweets that put the URL nearer the start of the tweet.

In Conclusion

It is worth remembering that while the data is interesting and the patterns within the data are fairly consistent, all of this data is from a relatively small sample set — just 120 accounts and just 61,557 tweets. If you are interested, you can see the complete numbers for the click data over on QW Consulting's (my company's) website.