Well, maybe that’s too harsh, sweeping and broad a statement to always be true – but then so is another that’s gathered traction widely… “Twitter is a waste of time!”
So, look on it as a hypothesis – one to be validated or disproved by data gathered in a more or less systematic fashion.
On February 2nd, 2009, over a 12 hour period, I conducted a Twitter Experiment.
The data generated suggests that Twitter is worthless as a ‘direct response’ vehicle – especially if you’re working towards a specific time-bound deadline, and when comparing it to email marketing with opt-in lists as an alternative.
Why email? Because both Twitter and your email list are “permission assets”. People have opted-in to both because they want to – not because you forced them to do it, or harvested their contact details.
First, the data.
TWITTER:
On my Twitter stream, I have 2,369 followers. Starting at 07:30 a.m. on 02/02/2009, I posted a series of tweets at intervals of between one and three hours. Each requested a reader to respond by tweeting a reply @drmani
In the first 6 hours, in response to 5 such tweets, there were a total of 34 responses. That’s a 1.4% response rate.
Interestingly, almost all responses came in within 30 minutes of the tweet, with very few coming later.
Over the second 6 hours of the test, the results were almost identical. There were 36 responses.
In all, there were 11 ‘repeats’ (tweets from people who had earlier responded to the experiment).
The actual response rate then drops to 59 people out of 2,369 followers – or a miserable 2.4%
EMAIL LIST:
Contrast this against data from some of my recent email marketing, across various sub-lists I manage on Aweber.com
Response rates (as measured by clicks on links in the email) ranges from 4.9% to 15.8% (with ‘open’ rates, that some use as a measurement of email efficiency, going even higher upto 21.1%).
That’s why I conclude that Twitter is WORTHLESS for direct response marketing, as compared to email marketing.
To achieve even half-way as much response with Twitter, one needs to keep re-posting the same tweet at different times of the day, so that you make sure you will reach segments of your audience that catch up on their tweet-stream at their convenience, and hit them within that “golden half-hour” window.
(Also, I would imagine that if I re-tweeted the same message over and over in a 12 hour time-frame, many followers would get annoyed and leave!)
On the other hand, email broadcasts are ’set and forget’, so you can have the message scheduled within a few minutes – and go about your other tasks.
Earlier, I thought Twitter would be a nice alternative to email marketing, as it gets past filters. But with the over-crowded morass that the Twitterverse has now grown into, and where everyone feels compelled to ‘follow’ everyone who follows them, this advantage has been replaced by a bigger drawback.
Now, you need to cut through the clutter on Twitter if you want to get heard or seen!
Don’t agree?
Ok, but don’t just argue or state fanciful opinion. Run your own experiment on Twitter, share your data, and prove me wrong.
Believe me, I WANT you to!
And here’s another challenge.
Can you come up with a way to test the “relationship building” component of Twitter in a (pseudo-)scientific way too? What variables will you use to measure? How will you do it?
That should make for another interesting Twitter experiment. For the record, I still believe that Twitter is an EXCELLENT medium to connect and network, build relationships, and leverage them over time into positive advantages in whatever area you operate.
Twitter is WORTHLESS only for “direct response” style communication.
What do YOU think? 
Review the “Twitter Experiment Data”:
Here are the data points you can verify for yourself. I used unique hashtags for this experiment so that you can check and see if the numbers I’m sharing are correct.
Tweet #1 (07:45) – see response – click here. At 07:59 – 6 replies. At 08:30 – 2 more.
Tweet #2 (08:30) – see response – click here. At 09:05 – 11 replies (one repeat)
Tweet #3 (09:30) – see response – click here. At 09:36 – 4 replies. At 10:01 – no more. At 11:45 – 1 reply
Tweet #4 (10:30) – see response – click here. At 12:15 – 6 replies. At 13:15 – 1 reply.
Tweet #5 (13:15) – see response – click here. At 13:40 – 4 replies (two repeats) 15:00 – 2 more (1 repeat)
Tweet #6 (15:00) – see response – click here. At 15:15 – 2 replies. At 17:00 – 1 more reply.
Tweet #7 (17:00) – see response – click here. At 17:36 – 11 replies. At 17:55 – 1 more reply
Tweet #8 (18:00) – see response – click here. At 18:13 – 8 replies (3 repeats). 18:50 – 3 more.
Tweet #9 (19:00) – see response – click here. At 19:02 – 4 replies (2 repeats). At 19:27 – 4 replies (2 repeats)
{ 1 trackback }
{ 13 comments }
An excellent job on the test comparison on twitter. I’ll probably run the same type of test with less people, but still it should be enough.
One thing that differs and I don’t know how to factor in on yours or mine when I do it is the time factor.
There’s a lower percentage of email that gets opened up in the time (real time life of the tweet). Yet email could get opened up 24 or more hours later and go into that statistic. The only way I can figure to do that is to send out a newsletter at the same time as the tweet and get track the responses from that mail at the same time you close out the tweets activity. ie last tweet.
What do you think?
So does this mean salesmen and marketers will go away and stop trying to monetize this particular social network du jour?
Dr. Mani -
I’ve found the same. But I never thought Twitter would be a great direct marketing platform. Hard to keep your Tweets in front of your followers. Right now I have only 11 minutes of Tweets showing on my Twitter “home” page.
Social aspects and the buzz factor are Twitter’s strengths. the capability for two way communication is something people seem to not get there, especially those who limit their followers.
Keep up the great work,
Lowell
Hi Dr. Mani,
interesting experiment – but I’m not so sure if it’s a fair one for Twitter.
You said that with Twitter “Each (Tweet) requested a reader to respond by tweeting a reply”.
If I understood correctly the email statistics only tracks opens and clicks. And a click is easier received then a written response.
If you’d asked your email readers to respond by sending a reply, you might have seen lower figures.
Or if you’d measure click through for a link in a Tweet, you might get a much higher click through rate then reply rate.
BTW: I use http://brightkit.com to measure my clicks.
I find Twitter valuable for getting immediate traffic to any website of my choice. It’s not big traffic, but sending out a Tweet gets more people to click my link than for example posting to a forum and having the link in the sig.
Let’s continue the testing.
Ralf
Would you allow me to get back to yesterday (Talking Stick)- I was plannining a reply on “Intention & Action”.
If I did, mine would have been the third reply. Since, there seemed no urgency (my perception)- I took it easy.
Yes Doc – agree – across the board – Twitter is WORTHLESS only for “direct response” style communication.
Yet, wonder how it may have been – if there was a OTO of great value?????????
Intention – is a beautiful word with meaning so subtle – it is not a quality, that a proffessional so crooked would dare to, or be able to – possess.
I am here to learn more of cleansed thoughts – from a master achiever, so accomplished.
That’s not all — —- —-
Going further, all your readers & followers would agree, that all who are here, are people who have realized that they can become – better.
Who knows me better than my own self – an epitome of Inaction – NOT – “IN-ACTION”.
In the midst of so many options – they (Including self) have made a choice – TO BE HERE.
So, once again – agree – Twitter is WORTHLESS only for “direct response” style communication.
You are only fine tuning your Crystal Ball – predictions for 2008.
Salute & Namaste!
Hey Dr Mani,
For this test to be closer to relevant shouldn’t you just be comparing click through rates for links within twitter updates vs click through rates for email broadcasts?
I don’t understand how an @reply is relevant since that’s not the response you want. You want them to click through from the twitter link and then take an action at the page you send them to, the same as you do with email.
In my link tracking texts with Twitter the percentage of clicks is similar to the percentage I get with email.
By the looks of the examples you show from AWeber, it appears you are using targeted email lists or list segmentations, which will result in better click through rates since Twitter isn’t something you can segment at the moment – you just broadcast to everyone listening at that time. This again makes the comparison not as relevant.
Correct me if I’m wrong of course.
Cheers,
Yaro
@James Stratford – Yes, that’s a valid argument. The only thing I’m basing this upon is the fact I seldom scroll back over tweets from many days ago to study them – but often review what’s in my email inbox, at weekly or monthly intervals. So the ‘30 minute window’ seems far more critical for Twitter based communication that the more ‘ageless’ email
Just my thoughts. Would need another test like you describe to confirm it.
@Dominic – Y’know, if this succeeds in achieving that end, it would be worth the effort in GOLD! But give it time, they will… when they realize just how futile direct pitching gets on Twitter, and other avenues are more ‘profitable’.
@Lowell – thanks for sharing your thoughts. Nice points about social aspects and buzz being Twitter’s strengths, I agree.
@Ralf – I agree completely that this is neither full and perfect comparison, nor should the ‘conclusions’ be taken as cast in stone. Look on them as more like ‘decision assisting data points’, and you can’t go wrong!
You said “And a click is easier received then a written response.”
But consider this. My email has to get past filters into their inbox. Subscribers must first open the email, then read the message, and only then click on the link.
On Twitter, the delivery and open phases are pre-done – which is why I was initially excited about the potential for it to rival and surpass email in direct response effectiveness.
Sadly, the same ‘ease of communication’ advantage has been drowned out by the sheer deluge of tweets that hit the average Twitter user’s screen. I randomly sampled my followers once, and saw they were following FOUR HUNDRED people each… on AVERAGE! Hence my ‘10 Twitter Maverick Rules‘
@Dileepa, thanks for the gracious comment, I appreciate you. You’re right, I do seem to be refining my ‘Crystal Ball’ thoughts
@yarostarak – Of course, you’re right. This isn’t a ‘fair’ or ‘equal’ comparison in the sense it is measuring exactly identical actions.
Extending the point you make even further out, what matters is the ultimate bottom-line (sales, or opt-ins, or anything else you’re measuring). Jay Abraham drilled that into my head, the quality of your lead matters more than quantity, as 3 responses that each result in a sale is superior to 3,000 that makes only one or none!
That said, I’ve been trying to use Twitter to synergize with my non-profit work, which involves raising awareness about congenital heart defects. Feb.14th is “A DAY FOR HEARTS”, one of two of the big annual CHD awareness events, the other being my Heart Kids blogathon (or tweet-a-thon) that happens in September.
Both require the first simple step of clicking through to the CHD awareness website to learn about the problem facing millions of children and young adults around the world.
On email, every year, the response rate to my invitation has hovered between 10% and 35% for this first step. That’s my benchmark, so to speak.
And in that ‘direct response’ sense, it’s infinitely harder to get similar rates to happen on Twitter without behaving in a way that is abusive, and stressing to breaking point this nice ‘permission based’ service.
This informal ‘Twitter Experiment’ confirmed my intuitive assessment (based on experiences around the tweet-a-thon, and later with few other efforts at awareness building) that email and forum based communication works better for specific, time-bound awareness building events like my CHD awareness day event – which is why you don’t see me trying very hard to do it on Twitter right now.
After 2 days of tweeting for support to the Congenital Heart Defects FACTS page, I’ve got just 175 visitors to it. When I open up other communication channels, including email, with my ‘Heart Kids Tribe‘, this number will ramp up far more rapidly.
That’s the point of this experiment. Twitter is still great to ‘Connect. Engage. Respond’.
Very interesting experiment.
I’ve always been puzzled as to how people can keep up with thousands of followers and build relationships with them – I know some manage to do it! Interestingly enough, I haven’t seen many of the well-known online marketers using twitter to pump out direct response messages or to sell overtly, it seems to be used more as a way of networking and building relationships which result in JVs. Of course this is already going on behind the scenes and twitter allows us to be voyeurs in their interactions!
Personally, I find constant promotion on twitter a turn-off, well, as much as a turn-off as knowing when someone has broadcast having had a bowel function! I prefer to keep in touch with fewer people and get to know them as well as I can.
Comments on this entry are closed.