How To Host A Giveaway And Not Crush It – Our Experience


A few weeks ago, we were throwing around ideas about how we could build awareness and an email list for KyLeads ahead of our beta launch. We’ve been blogging and creating content on other platforms so that was bringing in a few visitors every day.

We needed something that was relatively cheap because we had no data about our core metrics like LTV, ARPU, MRR (well, we knew that, it was zero), value of a subscriber, trial to customer conversion rate, etc etc. The only thing we knew was we wanted something that would give us outlandish results – fast.

That means content marketing and website optimization were out. It could give us outlandish results – I can’t say it would be fast.

Advertising was out. We weren’t ready to spend a lot of money when we didn’t know our numbers.

Outreach for promotion wasn’t ideal because we didn’t have anything compelling to share at the time. After a bit of soul-searching, I stumbled on a few posts talking about how to host a giveaway.

After reading up a bit, I thought to myself – this can’t be that hard. Once I pitched it to the team and a few friends, everyone was on board.  We locked it in for the beginning of June and got to work.

Image of crush it gif

This post outlines what we did, what worked, what didn’t, and our key learnings.

How we settled on a prize for our giveaway.

If you look around the internet for information on how to choose the best prize for a giveaway, the advice is consistent. Don’t use an iPhone or any other general consumer good. The reasoning is you’ll get a lot of interest but those people won’t be right for your core offer.

We sell software that allows website owners to make popups and quizzes to generate more email subscribers and personalize their messages.

At the very least, the giveaway should appeal to people who own websites. The best case scenario would be bloggers, digital entrepreneurs, authors, etc who’re actively building their online presence. With the right prize, conversion rates can be as high as 85%.

That meant general prizes like tickets to x, phones, computers, etc were out of it.

We settled on a set of books related to business and marketing. In all, there were half a dozen books I’ve read personally and people have recommended to me. It was a good prize and it was cheap (less than $100).

We could’ve gone live like that but I thought we could do better. Why not give away some relevant software too? KyLeads may be unknown but there are a lot of software companies which would complement our offering.

The more I thought about it, the better it sounded. It took me all of 30 minutes to come up with the idea and adapt a quick template to reach out to folks with.

It took another 2 hours to compile a list of potential partners and get the email address of the decision makers. In all, I ended up with a list of 30 companies who I thought would be a good fit.

I personally reached out to everyone on my list. Within minutes of sending the first email, Ryan Kulp responded and agreed to a lifetime account.

image of reply from Ryan about partnering

I was stoked. That was much easier than I thought it’d be. Over the course of the next week, the replies started to trickle in.

Some people weren’t able to participate but they still replied and let me down easy. They were cool about the whole thing.

Susan from Buzzsumo even thanked me for following up with her. She said no in the end but her response was encouraging. Respect to people like Alex from Groove and Francois from Hunter who are polite even when they don’t have to be.

Side note: The worst thing that can happen in any of these situations is someone telling you no or ignoring your email. No one is going to hunt you down and f*ck up your life because you emailed them. Just go for it.

Once the dust settled, we had seven partners on board and a few more expressing interest. That’s a 20% success rate on a cold email. The reply rate was about 50%.

Gif of boom baby

Our partners included:

Inputkit.io

Tettra.co

fomo.com

Wpfusionplugin.com

Complice.co

Pixelme.me

Ninja Outreach

Big thanks to everyone who participated.

Our prize pool had swelled to over $100,000. Each company was giving away at least one lifetime account. The $100,000 is a conservative estimate.

Needless to say, with a prize pool like this, I was sure we were about to crush it.

Getting ready to host a giveaway

We knew creating the giveaway wasn’t enough. We had to promote it as far and wide as we could. After a bit of deliberation, we decided to attack as many avenues as possible.

That would be organic, ads, and other people’s networks. In preparation for that, I set up a Facebook ads campaign, created custom images, and compiled a list of possible influencers who’d help us spread the message.

Here are a few of the images we used:

image used to host giveaway

To find influencers, I searched for Podcasts related to marketing and entrepreneurship. A few at the top of the list were:

After settling on the podcasts, I ran through past guests and listened to shows that seemed relevant and took notes.

EOF podcast episode list

There were over 2,000 guests to choose from.

I’d then go to the guest website, find their email address, and save it in an excel file. In the end, I had about 80 names and email addresses from doing this.

The other strategy I used to find influencers was by typing “top marketing blogs” “top small business blogs” and “top anything else relevant blogs.”

I was looking for compilations of influencers that would find the giveaway interesting and possibly share it with their audience.

image of google search of best small business blog

This tactic yielded another 75 relevant contacts. In total, we had over two hundred contacts ready to email once we went live.

Apart from that, we set up a Facebook ads campaign. I’ll give you more details about that in the strategy section.

Setting up the giveaway for maximum results

We had a number of goals for the giveaway.

  • Build awareness about KyLeads by getting it in front of as many eyeballs as possible.
  • Get more email subscribers.
  • Grow our social media following on Facebook and Twitter.

We needed software that would allow us to accomplish those goals without having to do most of the work manually. We also wanted a piece of software that had virality built in.

We looked at Gleam, kingsumo, and Vyper.

We eventually went with Vyper. The setup process is pretty straightforward. I liked Vyper so much because they have a great landing page and a leaderboard to encourage additional actions. You can view a screenshot of the landing page here.

You can set it up with tons of custom actions that give participants extra entries and increase their chances of winning the grand prize. We took advantage of those features.

We decided to go with three prize tiers. If you enter the giveaway you get a prize, two winners based on points, and three random draw winners.

Anyone could win and everyone was guaranteed to get a prize.

image of entry page and custom actions

Comment on a blog post, share articles, follow us on Facebook and Twitter, share on social media platforms, etc. Once everything was set up on the technical side, we ran through the giveaway to make sure everything was working as expected.

The giveaway went live on May 28.

Strategies we used to promote the sweepstakes.

No matter how awesome you think your giveaway prize is, it won’t do anything if you don’t promote it far and wide. We did a lot of preparation leading up to the actual giveaway and now was the moment of truth.

We had fourteen days to go big or go home.

Harassed friends

The first thing I did was tap my own network to get the ball rolling. If the people who’re required to support you don’t then something is wrong. Luckily, I’ve got some cool peeps.

Since we already communicate on a regular basis, I didn’t stress myself about how to word the email I sent them. It was basically me telling them to share it on social and a rundown of what’s involved.

Adam from Blogging Wizard came through and tweeted it to thousands of followers:


Your network is only so large. At least, our network is relatively small so it didn’t take long to exhaust that avenue. We moved right into emailing our existing list.

Emailed our list

The very next step was to get my existing audience on board. This is an audience that’s separate from KyLeads but still relevant. The Friday before the giveaway went live, I sent out an email letting them know I had a surprise in store.

I didn’t let them know exactly what it was but I did make it clear that it was epic.

Image of the giveaway email I sent

On the following Monday, I revealed the details of the giveaway with a longer email. The goal was to let them know what was happening, how it was relevant to them, and get them to click to find out all the details.

This is how we got our very first sign ups for the giveaway.

Over the course of the giveaway, we sent out three more emails. One was to let them know it was halfway over and strategies to get more entries (or enter in the first place).

The second one was to let them know about the current leaderboard and keep the giveaway top of mind. The third email was to announce just a few hours left in the giveaway.

I thought about incentivizing the laggards on my email list, but when time rolled around for the final email, I decided against it.

Facebook ads

Facebook is a fickle beast. It showed us its true nature when we were preparing for this giveaway.

I’m an old hand when it comes to Facebook ads. Nothing is written in stone so I follow a simple strategy. Come up with your epic ad angle, get a few creatives, use small budgets, tweak until you have a winner, then scale until people get tired.

It’s straightforward. Until it’s not.

I made an ad campaign for the KyLeads giveaway and followed the normal steps I’d take.

I made three ad sets targeting different groups of people. I set a low daily budget of $15 for each ad set just to get a feel for what was working and what wasn’t.

Facebook ad sets

I won’t bore you with the details because that’s not the focus of this post. After I created the ad sets, I drilled down and created three different ads for each ad set. In the end, I had nine ads primed and ready to go.

Image of facebook ad for giveaway.

Side note: If you’re writing ads, test out long copy as well as short copy. I know there are schools of thought which champion short copy. I don’t know why. I’ve seen long copy perform better in head to head tests more often than not. The thing is, if someone reads all that copy and clicks through then you’re more likely to get a conversion.

I digress.

After everything was set up, I decided to go and pay my outstanding ad balance early so I wouldn’t have any issues during the contest. So, I click pay and boom, this is what I see.

Image of Facebook error

I was pissed. I wasn’t so annoyed with the time I spent or the fact that my account was blocked temporarily. It was because Facebook has the worst customer service I’ve ever seen. I would be lucky if they sorted this issue out in the same month let alone within the timeframe we were working with.

We tried to reach out to them but they didn’t reply with anything useful for the first week. The chunk of change I wanted to spend would’ve made no difference to their billions.

I had to move on – fast.

StumbleUpon ads

The next choice for us was StumbleUpon ads. I’ve used them in the past but I’m by no means an expert with the platform. When I landed on their ads login page, I saw this:

image of stumbleupon ads home screen

At this point, it felt like the world was conspiring against us and this contest. I had a fleeting moment of weakness where I thought about throwing in the towel. It passed.

I logged into my StumbleUpon ad account to see if I could salvage anything. It was a good day. I had a few bucks left from an old campaign. I tried to load some money into my account and they just refunded it instantly. After three times I gave up.

I set up a campaign and let it run. This ended up bringing in a few hundred hits.

Side note: StumbleUpon itself is also going through a transition. They’re moving to Mix.com at the end of this month (June 2018). It may be a good opportunity to be an early adopter for a platform that promises to have millions of users.

Twitter ads

I’ve advertised on Twitter once before. That was way back in 2016. The results I got then made me decide to write it off.

I was running out of options. Facebook was shot. Stumbleupon was shot. Pinterest was acting like a dick.

I turned back to Twitter and decided to give it another shot. For the most part, it was similar to other ad platforms. What I disliked was how to implement the tracking. Instead of putting it in the head tag, it goes in the body.

I couldn’t get it to work properly with the software I was using. In the end, I had to do some magic in Google analytics.

We created four Twitter ads and let them loose on the world.

Image of twitter ad examples

The absolute results weren’t bad.

Image of twitter ad results

Our targeting seemed to be off because even though we got thousands of clicks, they converted poorly. I switched up the targeting again and this time the cost per click shot up by 100x.

Needless to say, I turned that one off and turned Twitter off as well.

Side note: People at Twitter. It’s hard to find data on your platform.

Quora ads

When Quora rolled out ads, I requested access and didn’t hear back from them. At the time, I didn’t think anything of it and went on with my life.

It wasn’t until I saw ads from a company I knew was in the early stages and couldn’t afford $5,000+ a month that I got curious again.

During the contest, we decided to give it a shot. The interface is similar to Facebook in that it has campaigns, ad sets, and ads.

What I didn’t like was how limited the text space was and their punctuation rules. The kept disapproving my ads because their software didn’t like the way I constructed my sentences. I’m a fan of long copy which helps pre-qualify the people who click to my website. Over time, I’m sure their real customers – the advertisers – will change this.

What am I supposed to do with that?

The CPC seems a bit high in comparison to other platforms I’ve used. Usually, your CPC goes down when your CTR goes up. With one of the ads I sent live, the CTR was about 1.5% (decent) but the CPC was over $2. It was odd to me. I kept the ad running to see if it would balance itself out.

In the end, the CTR and the CPC dropped to .06% and $1.50 respectively.

Though it didn’t drive much traffic, I’ll revisit Quora ads at a later date and really dive in to see how to make it work for us.

At this point, we’d run through our paid advertising options and didn’t even consider Google because the Keywords we’d be bidding on were too expensive. We’d burned a few days and didn’t have the time or motivation to tweak a Display ad campaign.

That left us with a few options.

Email outreach

We knew we were going to do a lot of email outreach from the beginning. We just thought we’d have more engagement before starting the process.

Whatever.

We pulled out our list of hundreds of emails and got to work.

Our results weren’t too bad considering the context. These were cold emails we sent out to people in the hopes that they’d promote what we were offering with no visible kickback for them. Our conversion rate from this was ~ 15%.

Maybe they’re betting on our future and our ability to return the favor later. Not likely. Looking back, I think it was because the contest was epic and they were generous people.

A few people were really cool about it and shared on Twitter and even emailed us back. In a few instances it led to interesting conversations and possible opportunities in the future.

People like Sam Hurley and Nick Loper did us a solid.

Nick Loper Twet Sam Hurley Tweet

Tapping into existing traffic

We didn’t have a lot of traffic to KyLeads at the time so there wasn’t much traffic to tap into. That’s not to say I couldn’t tap into traffic from other websites I own.

I put a single page takeover on one of my websites announcing the giveaway.

Image of full page takeover

I tested two versions of the CTA and it seems people liked the phrasing “learn more” over “enter to win.”

It converted over 53% more visitors.

Through all these strategies, it was difficult to monitor the conversion rates via specific channels. I was busy slicing and dicing Google analytics data. After a while, I just monitored conversion rates in aggregate.

If KyLeads had more traffic, there were a number of things we could’ve done.

  1. Use a floating bar on all pages to increase awareness.
  2. Added a call to action in the sidebar to capture blog readers
  3. Add a site wide pop up to build more awareness.

Posted on social media

Of course we published the giveaway on all of our social channels. We added it to multiple group boards on Pinterest, used this tool to post in multiple LinkedIn groups, posted on Twitter multiple times, made a Facebook post about it, and hit the other major social networks.

It was surprising to me that the post was almost completely ignored on Reddit. Not even the troll who told me to get a life. It just got buried.

There are a lot of guides out there about how to use specific social media channels so I won’t get into that.

Things we could’ve done better.

We’re a small team by design. This is one of those times when more could’ve been better. Even if it was just a temporary increase in headcount by way of virtual assistants.

That’s neither here nor there. Looking back, there are specific things we could’ve done to improve the results of our giveaway.

Prepared for setbacks.

I didn’t think Facebook would kick us in the ass like it did. We spent a lot of time and budgeted considerable resources betting on it. When it fell through, we were left scrambling to find alternatives.

Though Twitter drove a lot of traffic, it didn’t convert well for us. At the same time, other avenues closed before we had a chance to explore them.

We could’ve cut down on the effect this had on us by preparing a few contingency plans. That may have looked like preparing ads on different networks ahead of time and testing interest through other avenues.

The best laid plans can and will go south. That’s what happened to us and we ended up being reactionary instead of proactive dealing with situations as they occurred.

I’m sure we could’ve had a different outcome if we’d been able to recollect ourselves midway and go in a different direction.

Lesson learned. Always be prepared for when your plans go wrong – they will.

Written a blog post

I debated writing a blog post because, as I’ve mentioned earlier, KyLeads doesn’t have much traffic. That being said, it wouldn’t have hurt us if we’d written the post and taken a unique angle.

It would’ve been another opportunity to promote the brand itself in addition to hosting a giveaway.

At the very least, it would’ve made for an interesting read. As it was, most of our efforts focused on the things we did offsite.

Going forward, we’re using our blog as much more than a way to contribute to customer success. Yes, posts about how to create compelling lead magnets are indispensable. They’re not the only thing we can produce.

The fact that I’m writing this post is evidence of that commitment. The old me would’ve kept quiet about it and gone on with my day. Now, I’m much more inclined to share.

Coordinated with partners better

The only thing we did to get our partners on board with promotion was to send them a few emails before, during, and after the giveaway. Now, don’t confuse this with not communicating with them.

We kept them up to date during every step of the process. What I mean is we didn’t push promotion on them. I felt like they were already doing us a solid so it should be our responsibility to get the word out.

With that line of thinking, I shared a few graphics we made for social media and encouraged them to share to their following. There were no other dedicated messages asking them to participate.

In hindsight, I could’ve chosen strategic opportunities to get them to involve their respective audiences.

  1. At the launch
  2. The midway point
  3. With just a few hours left.

Not everyone would’ve taken me up on it, but the ones who did could’ve made a big difference. Going forward, we’ll create collateral for all stakeholders when we’re cross promoting:

  1. Social sharing images
  2. Templates for posts to different social media networks.
  3. Email templates

It’s their choice whether or not they use it. People are more likely to participate if you make it easy for them to say less by reducing their cognitive load and risk.

Better Identified influencers

Another area where we fell short was communicating with the right influencers through outreach. We used a generic approach and the emails to top bloggers fell on deaf ears for the most part.

Not surprisingly, the emails to podcast guests that mentioned specific parts of their podcast episodes got the highest response. In some cases, that was just to tell us it wouldn’t be a good fit.

Think about that, not only did they open the email we sent, they read it, and felt compelled to reply to a complete stranger.

That shows the merit and power of a tailored email.

Going forward, we’re going to focus on the quality of our outreach as opposed to the size of our outreach pool. It’s obviously easier said than done but I think we’re equal to the task.

Breakdown of our giveaway stats

Finally, let’s look at the results we got from the giveaway.

Total impressions: 150,000 (win)

Facebook: 10,000

Twitter: 100,000 (paid and organic)

Quora: 25,000 (paid and organic – I’m an active member of the community)

Other (hacker news, Reddit, Stumbleupon, etc): 15,000

Click through rate: 3.0%

Total visitors – 4,500 unique visitors

Page views 5,000

Conversion rate (aggregate): 3.62%

Total new email subscribers: 163

Fake/invalid or catchall email addresses: 9 or 5% (average is roughly 15-20%)

Average email open rate during the giveaway 42%

email open rate image

New Facebook likes: 27

New Twitter followers: 8

Blog comments: 14

Virality score: 42

 

Virality score is a metric from Vyper.io that lets you know how well referrals are working in your contest. It’s from 1-100 and the higher the number the better.

As you can see from our stats. We got a lot of raw exposure but in the end, that exposure didn’t bring the result we’re looking for.

Conclusion

We did a lot of things right and we did a lot of things wrong during our giveaway. I’m happy about the learning experience and the connections we were able to build.

Do I wish we got 10,000 email subscribers? Of course.

Am I worried this is part of a larger trend? Not at all. You win some and you lose some.

I want to say thanks to all of our partners once again, you’re the real MVP (the teammate kind, not software 🙂 )  and for everyone who participated in the giveaway.

If there’s anything you think I left out let me know in the comments and don’t forget to share.

We Wrote on Medium Every Day for 30 Days – This Is What Happened.


Medium has been around for a while. Some people sing its praises. Others are vocal about how Medium continues to change their business model to their detriment

Whatever.

That’s not why I’m on Medium. Though I’ve been republishing articles on the platform for a while, I’ve never focused on it. It’s never been and will never be our base of operations. It’s more like a loosely defended outpost.

It was always a small piece to an overall content marketing strategy. One of the reasons I never went too deep into it was because when you publish on Medium, you’re just another Medium writer.

I cease to be Daniel Ndukwu or KyLeads. You lose your identity unless you’ve already brought a huge audience with you. I didn’t. Apart from that, the reporting features leave a lot to be desired.

I’ve never adopted a consistent strategy specific to Medium. If I wanted to write something that didn’t fit into the topic of our blog then I’d just publish a Medium article. That was all the thought I gave to it.

Some of them were successful. Most weren’t.

In April 2018 that changed. I decided to give it 30 days of consistent effort and measure the results. This article sums up my learnings.

Where I started my Medium experiment

On April 12, 2018, I started my Medium experiment. The premise was simple. Write an article every day for thirty days. I also put a few limitations on myself:

  • I couldn’t email my list
  • I wouldn’t promote it (apart from Twitter) across the internet
  • The only extra push would be from Medium publications
  • I wouldn’t tell anyone about the experiment so they wouldn’t visit my Medium profile more often.

Apart from that, I was free to write about any and everything. The only other thing I tried to do was link back to KyLeads on relevant articles. This wasn’t a big deal; it was just something I kept in mind.

I started the experiment with 731 followers.

first day profile page

I had 420 article views and 134 reads in the thirty days before I started the experiment.

Medium writing stats

It’s important to note I wasn’t exactly starting from scratch. I’m by no means the most popular writer on Medium but a few people follow me and clap for my articles regularly. In addition to a few followers, I’m also a writer for a half a dozen publications.

During the course of the experiment, I didn’t hustle for my articles to be featured in any new publications. If you’re brand new to Medium, that may be one of the most effective things you can do to get more eyeballs on your content.

Now, let’s look at what actually happened during my Medium experiment.

The Medium Writing Experience

As I mentioned before, I’m a writer for many publications on Medium. The plan was to write an article and submit to them on a daily basis.

I quickly ran into a snag. The editors are busy people. They get submissions all day every day. It takes time for them to review and accept a submission. I didn’t take that into account.

I’d submit one article a day and sometime three of them would get published at the same time. If you look at my timeline, you’ll see I would publish up to four articles at the same time.

That wouldn’t work for what I was trying to accomplish.

I stopped submitting first. Instead, I’d publish directly then submit to publications. They could accept at their leisure and I’d get the benefit of adding my work to a publication while keeping a consistent schedule.

It worked out pretty well. Not all my articles were accepted. Either it wasn’t a good fit or I removed it from one publication and added it to another because I felt like I’d get more exposure that way.

In the end, out of 30 articles, 19 were accepted to publications. Not bad.

Engagement increased

One thing I noticed was an increase in engagement across the board. I was expecting a few shares and claps. I got more than that. People were commenting on the articles and highlighting parts that stood out to them.

medium writing highlights

I think this was in part due to the way I was writing. Usually, I write about topics related to entrepreneurship and marketing. During the thirty day period, I touched on a wider range of topics. With many of them, I was shooting from the hip.

image of medium engagement and comments

What I mean by that is my articles weren’t as narrow in scope. I wrote about a variety of subjects like how I felt like giving up at times, not going for broke, and overwork. It appealed to a wider audience.

There was just a larger amount to consume. Gary V refers to this as content, on content, on content. His process is a bit more sophisticated.

Many of those people may not become subscribers or customers. It’s all good. They may still introduce me to the person that will. Apart from that, it’s interesting to see which parts of an article resonate with people the most.

The results of my Medium writing

The part you’ve been waiting for. What actually happens when you write on Medium for a full month?

For me, the metric I wanted to improve the most was my Medium follower count. While I got a little boost, it was negligible. This is probably due to the fact that I didn’t promote the articles much.

I started with 731 followers and ended with 759 followers.

30 day profile results

That’s an increase of 28 followers or roughly 4%. My target was to hit a thousand followers or a 36% increase. I failed woefully in that regard.

Looking back, I could have utilized more calls to action. For the most part, I just wrote and published. Only a handful of the articles had any sort of call to action to speak of.

Let’s look at my traffic stats.

After thirty days of writing on Medium, I got 1421 views on my articles and 488 reads.

medium writing stats after 30 days

I started with 420 views which is a 338% increase. Not bad. As far as actual reads, I started with 134 and increased it by 364%. Also not bad.

Even though I started with pretty low numbers, I’m happy with the increase. But these are stats for my Medium activity. More important to me is how many people left medium and went to my websites.

Google seemed to have trouble picking up the referrals sent from Medium. Luckily, I was also using a short link via bit.ly.

That one picked up 27 clicks to our website.

Out of the 488 people that read my articles a little over 5% clicked through to my website. Ehn, it’s not great and it’s not horrible. If it was an advertisement I would be jumping for joy.

It could’ve been improved a lot if I was actively promoting something. What I was doing was linking in the body of my content. There was no call to action or anything of that nature. So, I’m ok with the results.

Conclusion

I still believe in the power of Medium. Even though the results of my case study appear to be lackluster, you have to consider a few things. I basically published and prayed. I tweeted each article once, maybe twice, and did absolutely no more promotion.

If I were to do that on a normal blog, my views would’ve been much lower. I was also able to drive traffic to my website. Even though it was only 27 clicks, that translates to about two subscribers (the website is converting at 10% or so.

That’s two subscribers we didn’t have before so it’s a win.

If you’re going to throw your weight behind Medium, make sure you promote your articles and get accepted to publications. You can get much better results than me.

Exit Intent Popup