Last September I was on a panel with Guy Kawasaki talking about Twitter. He said at the time that he Tweeted 4 times for every story that he wrote. FOUR TIMES! The exact same Tweet. I couldn’t believe it. His rationale was that he found that his audience was tuning into Twitter at several different times during the day and he found that four was the optimal number to convert enough of the people reading his posts into traffic back to his website.
I asked him whether he was worried that he was turning off potential followers who didn’t want their streams flooded every day with Guy Kawasaki Tweets. He argued that anybody who followed enough people wouldn’t really notice much of a difference and if they followed so few people that they were significantly flooded then they were the wrong followers for him [on this point I’ll never agree – I strive not to overwhelm any followers]. At the time I had a small enough group of people I followed that if anybody was in overdrive on posting for a day I always noticed (as I pointed out in Point 2 in this post).
I left the panel thinking that Guy was off base but realizing he had somewhat of a point. I’ve argued previously that Twitter is a new form of curated RSS and in many ways it is. But it is a transient RSS reader. If you’re not logged in for a few hours and stuff passes through the pipes then it’s gone. It’s true that there are ways to make sure you don’t miss stuff (like lists or segmenting traffic in TweetDeck) but most people don’t employ these techniques. They just consume Twitter when they’re hungry for a conversation or some news right now.
So I started experimenting with multiple Tweets. In particular I would schedule some Tweets (using CoTweet, which lets you schedule Tweets) to go out around 5:40am (in time for East Coast 8:40am consumption) and then again at 8:40am for West Coast time. In fact, that is what I plan to do for this post. I’ll finish writing around 1am and that’s a dumb time to Tweet because few people in the US are online. Sometimes I would send a Tweet at 7pm and then again at 7.30am the next morning. I wanted to see two things:
- Would the second (or sometimes even third Tweet) convert enough people to my blog to make it worth potentially annoying some people on Twitter?
- would I get a reaction from the Twitter community telling me it was too much?
I’d like to share my conclusions with you but then also ask you for feedback. Many people reading today’s blog post would have seen it by clicking through on Twitter. My questions for you:
- how often do you notice my second Tweet? I’m going on the premise that on most days most users don’t notice. Some will notice it all the time (either because you follow 70 or less people or because you’re often on Twitter)
- how badly does it bother you when you do see a second Tweet? Do you think to yourself, “I can understand why he’d send it twice because many people might not see the first one” or “man, is that annoying. I wish Mark wouldn’t do that.” (I promise not to be offended by your answers – I’m trying to get a feel for the norm myself).
- If your goal is to send a Tweet that converts people to a blog post, sending more than one Tweet is recommended. I would assert that people following you by definition are more likely to want to see content from you and therefore you’re better off sending 2 versus 1 Tweets (we’ll see from feedback on this site whether others feel the same way). As an example you can see from my awe.sm logs a recent morning that 399 people clicked on my link on Twitter the night before at 7pm. I send out a second Tweet at 7am and by 8:30am I already had 224 clicks. This number often passes the first number by the end of the day. If I sent out a third Tweet later (I didn’t) it likely would get about 50% as much as the morning one. This means that there are still many people who haven’t seen it and would like to.
Note that these numbers only measure people who clicked on that exact link. Many people swap out my short URL code and put in their own so I don’t capture 100% of the total clicks with the codes but if you look at the overall traffic from that morning on my blog you’ll see that my Twitter link accounts for about 15% of the morning traffic to my blog (this percentage will drop by the end of the day as more people arrive via RSS readers or referrals) and last night’s link accounted for about 10% of the daily traffic.
So my conclusion is that the second Tweet is generally worth it. The third probably is also but I usually resist the temptation in the desire to balance “reach” with “frequency” so as not to piss people (you!) off.
2. I try not to “double Tweet” every day and I vary the time of day just to shake it up a bit. You’ll see from this morning’s logs that my post today was featured prominently on Hacker News. This always leads to a spike in traffic. I could already see this by early AM so in this case I didn’t think it was worth RT’ing to get another 200 viewers through the door.
3. The smartest strategy I’ve seen is implemented by Babak Nivi over at VentureHacks. He’ll send out multiple Tweets linking to the same story but with totally different text. What he does is pull out specific quotes from the story and then Tweets those but linking back to the same story. I find that this is more palatable for me than seeing the same Tweet 4 times (but has the downside of potentially driving people to your blog post that they may have already seen).
4. I also try to mix up my Tweets with a combo of Tweets linking to my blog, Tweets making general comments like where I’m going that night and some Tweets where I ask a question to engage the audience (obviously where I generally want to know something). I think this is important – otherwise your Twitter feed just becomes, literally, an RSS reader.
So, whaddaya think?