It was almost New Year's Eve and I wanted to do something special on Twitter. I had 69,800 followers and because I admittedly am an imperfect and superficial human addicted to vanity metrics, I wanted to get to 70,000 followers before midnight and it becoming 2020.
I knew Gumroad had done giveaways before. I had thought about it. Half of my friends said it was sleazy and attracted the wrong audience, the other half of my friends said "just do it and see what happens", and then there was Cam who said it was "both sleazy and attracted the wrong audience and I should just do it" 😂
I'm not a big fan of ads
I'm not a big fan of ads, nor consuming ads: I have multiple levels of ad blockers like uBlock and HOSTS files blocking any ad and tracker, nor paying for ads: I have tried buying Facebook and Twitter ads and they never really compare to the conversion/clicks/sales I get from writing viral tweets/blogs. Now doing a giveaway is somewhat of an artificial way of getting traction for a tweet. But so are ads.
So I wanted to try it and I wrote this tweet. Pressing the TWEET button on this I got nerves all over thinking this might go complete wrong, but let's try:
As you probably expected to happen, I've never had a tweet take off so fast:
When it came down to actually picking a winner I started Googling. How do you pick a random person from the people you RT? Luckily there was already lots of sites specializing in that:
My friend Marc was quick to point out that it would be impossible to pick a random person from the RTs due to Twitter's API limits:
Of course me being me, I didn't believe that at first. But the API that lets you list RTs has disabled pagination. So you can get the last 100 RTs, then if you paginate, you get the same last 100 RTs. It must be possible to get a list of all RTs? How about without the API but just through Twitter's web interface?
Nope, it doesn't scroll further than 100 last RTs. So that didn't work. Of course because Twitter's web interface also uses the Twitter API. So who can help me get that RT list?
Nobody except Twitter employees can.
So now how was I going to randomly pick a winner then?
My friend Marc again to the rescue. He suggested that since there was 10,000+ people RT'ing and following, I could just pick a random follower from my current total follower list (78,000 at this point), then go to their profile to check if they RT'd it and see. If they didn't, get another random follower and repeat, until you find someone. With 78,000 followers this should take about ~8 tries.
Where would I get my followers though? I could use the API but that'd takes ages with pagination I think. So I exported my data.
When you extract the archive ZIP file, you get this folder with all your data:
One file called follower.js has a list of all your followers. Beware it's more followers than you actually have. I had about 92,000, of which 14,000 were suspended accounts.
I then made a PHP script (I put it on Github here) to pick a list of random followers, so I could open their profile in my browser and see if they had RT'd the giveaway tweet:
And then I did it, and screencapped it for verification:
The vulnerability of this 100 RT limit
If I'm right (and I think I am, but let me know on Twitter if I'm not), and:
1) we know we can only get the last 100 RTs from Twitter
2) and every random RT picker service I've seen online uses the Twitter API to pick through
3) it means that it's relatively easy to hack giveaways to win them by being one of the, or all of the, last 100 retweeters
Not saying you should by the way, because it's very not nice.
But I don't think most people who do giveaways are technical and will go through the trouble I went of exporting their data, writing their own random follower script and opening people's profiles until finding the person who actually RT'd them.
If there's a service that downloads all your followers, goes to their profile to find the RT (which takes getting sometimes hundreds of tweets to get back to the date of the original giveaway RT), then yes they'll give an accurately random result. But I checked lots of these RT pickers and I don't think most do this.
That means if you can get 100 fake Twitter accounts, which again I don't recommend, and you have all of RT just at the end of the giveaway (after everyone else, and before they pick a winner). You can have a 100% guarantee you win it.
And that's a big vulnerability of Twitter giveaways.
Was it good for business?
Is this a suitable marketing strategy if you don't want to spend ads? Yes, kinda, maybe.
The tweet got 1,459,193 views, 80,368 engagements, 54,061 profile clicks and 11,723 retweets. If you'd use ad metrics for that that means:
CPM (cost per thousand impressions): $3,000/(1,459,193/1000) = $2.05
I could only find Q3 2018 data when Twitter's CPM for ads was $5.93:
So it's almost 3x cheaper than buying ads. But Twitter ads you can actually target, with a giveaway it's quite random who will RT it.
I added about ~10,000 followers, at a cost per follower: $3,000/10,000=$0.30 per follower.
I checked a buy fake Twitter followers site and it $269 for 10,000 followers, so $0.02 per follower.
So a giveaway is about 10x as expensive as buying fake followers. But then again, the fake followers are robot accounts, the giveaway followers are real people.
Even then, the followers you add from a giveaway probably aren't the people who actually want to read what you have to say, or use your products/business. They want a free thing. And I expect a share of the 10k+ followers I added to drop off once the winner is announced.
(Update: At the peak I had 78,700 followers, after announcing the winner it's dropped to 78,200 followers, so that's a loss of 500 followers.)
All that is nice for your vanity metrics but doesn't do much for your business. In fact, both sales and volume on my main project Nomad List (which I link to from my Twitter profile) were lower on Dec 31 - Jan 1 than the same weekdays a week before. That could be NYE of course. I didn't see much effect on revenue. That's logical though since again people just want a free MacBook 16". I'd not click through either, haha.
However, let's say I sell 30x Nomad List memberships for $99/year because of the exposure this giveaway gave me. That's ~$3,000 and would make me break even. It's hard to measure this because I don't know which of the new followers or people who saw the tweet will stick and sign up now or tomorrow or in a few months. It could very well be >30 people. Not sure.
And finally...the winner
For everyone scrolling down to find the winner: The random winner of the giveaway is: https://twitter.com/JSXRDX.
See the video above for verification of the randomness.
Houston, we have contact 👾 pic.twitter.com/653xkva1vJ— (@levelsio) January 11, 2020
And here I ordered the high spec one for him:
🎁 Ordered the 🍏MacBook Pro 16" for the giveaway winner, final cost $3,518.90:— (@levelsio) January 14, 2020
- 2.3GHz 8‑core 9th‑generation Intel Core i9
- AMD Radeon Pro 5500M with 4GB of GDDR6 memory
- 32GB 2666MHz DDR4 memory
- 1TB SSD storage
- 16-inch Retina display with True Tone
Video below pic.twitter.com/mrA22bUKH7
This was a fun experiment.
Would I do it again? Not sure: there'd be a risk in my Twitter becoming a superficial giveaway account when I like it more as a place where I tweet about what I do and make every day.
Still a giveaway helps if you'd like to quickly build up the vanity metric of followers, which I think Gumroad is trying to do with their Twitter account, and having more followers might help your business to get actual real involved followers who may become customers later!
Using revenue to support other people
It does feel really good to give away if you can afford it. And in the giveaway tweet I wrote "👩🎨 Make something cool with it" and I do hope the person who won will do that.
Last year I already started supporting makers on Patreon:
Admittedly, that's relatively small amounts though like $10/mo per maker. The bigger goal (if I make enough revenue) is to make it into a fund which can bankroll indie makers and creatives for a certain amount of time. Like 6 to 12 months at $2,000/mo to focus on building projects. That means $12,000 to $24,000 for one maker. Which is a lot to give away, but would be the plan if I can afford it with revenue.
The giveaway was exciting though, and humble bragging here but we/I hardly checked the RT/follower counter because we were too busy drinking IPAs and chatting with friends during NYE:
Happy new year everyone! <3