If you’re launching a new product, you’re probably short on resources and are looking for cost-efficient ways to grow. It’s only a matter of time before someone reminds you that Robinhood, Harry’s and Dropbox all drove significant hype through engineering virality before launch.
“Those things don’t work anymore”, you might think — do people really still share waitlist signup pages? It’d be nice to generate a bunch of signups for free. But surely people are tired of that tactic after being relentlessly spammed by friend invites for crappy products…
To find out, we ran an experiment with a new product we were launching, the Emerging Channels Report. I’m here to tell you it still works (in the right conditions). We drove 383 incremental leads from sharing (906 leads total), representing a 42.3% uplift from virality!
We’re making this post the definitive guide to pre-launch campaigns for your startup, by giving you the play-by-play of exactly what we did. We’re sharing everything, from the email and landing page templates, the tech stack we used, as well as the costs and performance broken out for each tactic.
If you follow along, and you have a compelling product idea, you should be able to replicate our success — if you go this route then best of luck, let us know how it goes! If you’d prefer to work with Ladder to help you execute, get in touch so we can talk more and see if there’s a good fit.
Product: Emerging Channels Report
We’re big believers in selling your sawdust — we’re already doing the hard work to help our clients grow, so we just add an extra step to document and productize what we learned. One example is the Ladder Playbook, a database of 853 growth tactics from 7,000+ experiments we’ve run. We’ve also launched SaaS tools internally and externally, and we share everything we learn on the blog you’re reading — responsible for 60% of our new client leads!
As one of the leading growth marketing agencies, we have a team of 40+ growth marketers testing every new channel, strategy and tactic we can think of. We also have access to a wider network of growth marketers we know from being active in the community. How can we leverage all of that R&D and inside access?
The answer was in publishing premium content, the Growth Intelligence Unit. For anyone who needs a deep dive into a specific platform, or wants to know what channels to test, or what trends are happening in the industry, they could pay for access even if they weren’t a client. We decided to test this out with an MVP launch. The first topic we found interesting was Emerging Channels — what are we finding success with, that we recommend people should test.
Results: Was This Campaign a Success?
Our goal for this campaign was to maximize the number of signups before launch. While revenue from this first launch would be a great value add, we know that primarily the long term benefit of this campaign was validating a new line of business, and generating some buzz around launching something innovative and new. Therefore we structured the production of this product so that revenue from purchases of the report cover the cost of production, at no risk to Ladder (other than the time committed to idea development, interviews and promotion).
Here’s how the numbers broke down from May 27th to June 10th 2020:
Referrals were our biggest channel by far! Over 42% of all our leads came in with a referral link, meaning they came from sharing with friends and coworkers. Qualitatively we also know this got a lot of buzz in various slack and whatsapp communities, and though it’s hard to measure, we imaging the 55 leads from ‘other’ also came from sharing activity.
The email newsletter was extremely important, because as you’ll see in the later breakdowns, this was the seed audience for much of the later referrals. This shows the importance of building an audience by adding value over the long term, so when you promote something new, people respond. That said, Facebook ads also killed it for us — with a budget of $500 that works out at a CPA of $2.76.
If you count all leads and add in the $600 we spent on design for all of the marketing assets (emails, ads, landing pages), then our total cost to launch (excluding production of the report) was $1,100, or $1.21 cost per lead. We’re typically comfortable with a cost per qualified lead of $200, so we only need 5 or 6 of the 906 leads, or 0.6%, to enquire about services and have the budget to work with Ladder to make this worthwhile.
Strategy: Engineering Virality Launch Checklist
Looking to launch a viral campaign yourself? Here’s a list of the various activities and assets we covered to get our campaign ready for launch. We kept it relatively simple, which is what we recommend — you can optimize but ultimately what drives virality is whether the product idea is interesting or not. No amount of investment will make up for a bad idea, so start with a small, simple campaign and build from there if you find you’re onto something.
- Facebook ad – static 1
- Facebook ad – static 2
- Facebook ad – animated
- Social sharing image
- Email header
- Landing page
- Sharing page
- Landing page
- Sharing page
- First 3 channels email (sent on sign up)
- First 5 channels email (sent after 3 referrals)
- Free report email (sent after 10 referrals)
- Influencer email (sent after each referral)
- Pre-launch email
- Last chance email
- Launch email
- Social sharing posts (Facebook, LinkedIn Twitter)
- Facebook ads campaign
- Unbounce pre-launch landing page
- Connect Unbounce to Kickoff Labs
- Connect Unbounce to Mailchimp
- Unbounce launch landing page
- Gumroad purchase button
- Google Tag Manager tracking setup
- Kickoff Labs sharing page
- Kickoff Labs reward-level emails
- Mailchimp newsletter emails
- Social sharing posts (Facebook, LinkedIn Twitter)
So you can see, even for a small campaign there is a significant amount of work here — there are plenty of items that without a checklist, you’ll forget about. What makes it difficult is that these are all very different skill sets. To run a successful campaign you need strong design, great copywriting, and technical know-how for setting everything up.
We used tools like Unbounce and Kickoff Labs to simplify setup, so we didn’t need to code up landing pages ourselves, and the marketer in charge of setup also wrote the copy, which saved time. Design was the one thing we outsourced because we wanted this to be an exercise in branding, but it’s possible with tools like Figma or Canva to pull together something halfway decent as a non-designer.
Channel: Facebook Ads
I’m covering this channel first, because I want to dispel the myth that your product is going to just ‘go viral’ without any investment. Every viral product needs an initial seed audience — how do people find out about you in the first place? If nobody’s heard of your product, they can’t spread the word.
Often this comes from the founder hustling to do 1 on 1 interviews, partnering with other companies or posting on forums, but that takes a lot of time. For us this launch was about quickly validating an idea for a new business unit — speed of learning was important to us. So we commissioned a small test budget of $500 to kick things off.
Not only did this buy us faster learnings and a bigger audience, but the performance of the Facebook ads themselves tell us a lot about the idea we’re validating — if our clickthrough rate or conversion rate from ads was below benchmarks, we’d know we weren’t solving a problem people cared about. This information is the most important thing to find out for an early stage startup, before you invest significantly more time and money into scaling.
Here’s our Facebook ads funnel:
|Cost per Mille (CPM)||$6.58|
|Cost per Click||$0.76|
|Cost per Lead||$2.76|
|Effective Cost per Lead||$2.21|
Interpreting the Results
This is way better performance than I expected. For a relatively limited budget we served 75,000 ads and got 6 hundred clicks and close to 200 leads we wouldn’t have had over a one week period. Don’t believe anyone who says they just ‘went viral’ — almost every successful pre-launch or kickstarter campaign I’ve seen bought some ads to add gasoline to the fire.
The clickthrough rate on this campaign was decent and the CPM was low, but what really made this campaign work was the massive 27% conversion rate. We’d typically expect a 5-10% conversion rate so this was far better than we hoped, and made the whole economics of the campaign compelling.
The referrals and K-Factor part might be new to you, so I’ll explain it. Because people who clicked and signed up from Facebook ads could refer users, it doesn’t make sense to just count those first order signups. Only 9 of these 181 people (5%) invited friends or colleagues, but when they did participate, they invited on average 5 each, bringing in another 45 or 24% more people than we would have had otherwise. The K-Factor is simply 45 divided by 181, or for every new user we drive, how many more can we expect from referrals? That brings down the effective CPL (cost divided by all leads, referred or otherwise) to $2.21.
Note that Facebook’s own reporting reported 264 Leads, and a $1.89 cost per lead — 45% off from the 181 we saw in Kickoff Labs, the platform that hosted our sharing page. I also checked Unbounce, which is the platform we used to build the landing page which also records the utm_source, and had 183. To complicate things further, Facebook also reported lots of downstream leads in other areas of our business, as well as 568 content views of our blog.
This landing page only received traffic direct from links in the Facebook ad, the email newsletter, or via sharing, So these were signups that did click on a Facebook ad, but ultimately converted via the newsletter or sharing links. It’s common that you have this overlap in attribution, so don’t blindly believe one channel — use multiple tracking methods and decide which one to trust.
Because this campaign was short-lived and successful, I didn’t need to dive too deeply into cross-channel attribution, and could essentially just ignore this and go off what Kickoff Labs reported. These extra conversions can be treated as a nice ‘halo effect’ — extra proof we did a good job, but not necessarily counted towards the channels’ ROI. If I needed to justify more Facebook budget or was continuing this campaign, I’d want to dive deeper.
Though I’ve personally spent upwards of $15m on Facebook ads for companies like Booking.com, Monzo Bank and even Facebook themselves (for their Workplace product), my preferred campaign setup is laughably simple. If there’s one thing I’ve learned, it’s to let the algorithm do it’s thing — gone are the days where creating thousands of micro-segments and manually optimizing them was effective.
It’s hard to beat a good clean lookalikes audience, with automatic placements and dynamic creative — it gives Facebook maximum control for optimization, which gets you the best results and doesn’t take days to implement. For this campaign I had about 30 minutes, so I went with just a 1% lookalike audience of website visitors for our top countries.
I set a lifetime budget of $500 to end the day before launch (runtime of about a week), and used automatic placements with no cost controls. This let’s Facebook decide the best way to spend the budget across time, location and placement, to maximize conversions. Note: I don’t recommend this setup if you aren’t optimizing to conversions — you’ll likely get a lot of low value clicks.
This left me with an audience of 6.6m people, from which Facebook estimated would give me a reach of 5 – 17k weekly. This is well within acceptable bounds — if it was significantly below 1 million I’d begin to worry about saturation, and if it was north of 10 million it’s likely too broad.
The creative used various best practices — we offered some value for free, introduced urgency, made it clear who the offer was for and how it benefited them. We also produced professional assets so it didn’t just look like another half-baked agency product. This included a video animation that really stood out for people browsing their timeline. You can see the assets here.
Then I added three primary texts, and five headlines, all of which worked with each other. Because this is a dynamic creative campaign, Facebook will choose which ones to pair up based on what performs, but it makes sense to avoid anything that could clash.
So how did each component perform? Overall Facebook did a good job of giving each asset a chance, yet doubling down on the clear winners. There were definite variations in Cost per Result for each asset, but by and large Facebook chose well. More importantly, they did this automatically — so I could basically set it and forget it.
For example on image assets, we found that the two static images performed more or less the same in terms of Cost per Lead, and therefore got similar budget allocations. Interestingly the video asset didn’t perform well at all — it was 35% higher in terms of CPL because of a lower conversion rate (30% vs 42% for the static images).
Naming the channels in the report seemed to generate more interest in terms of CTR (version 2) but overall worse performance than listing the benefits they’d get. Mentioning the number of channels was much more engaging than leading with the benefit then talking about the structure (version 3).
There was a lot of variability in headline performance — one variation got 76% of the leads despite only having the second best performance. However Facebook makes the best decisions across all the variables, so you sometimes see this when isolating just one. Both of the headlines that got the most budget mentioned the free download, so this is definitely a learning we’ll take into future campaigns.
Channel: Email Newsletter
I guarantee you anyone who’s made significant money off the internet will tell you that building an email list is one of the highest value things you can do. Because email is a largely open standard, you’re less susceptible to being cut off from the audience you’ve built, unlike for example social media followers. Plus unlike paid ads, it’s free — very few people unsubscribe, so you can reach an audience over and over again once you acquire them.
Email was absolutely crucial for our campaign, having built up our audience over 4 years to ~7,000 highly engaged people. We rarely send more than 1 email per week and invest thousands of dollars in new content every month to fill our newsletters with valuable information. This might not be an asset you have access to if you’re just starting out, but I recommend you find someone who digs your idea who has an audience they’re willing to promote you to.
Here’s our Email Newsletter funnel:
Interpreting the Results
Email was the single most important channel for this campaign — you can see just how much more engaged this audience was compared to Facebook (and we didn’t have to pay!). For example with 1/4th of the impressions we actually drove more clicks, because the CTR was far higher. The conversion rate for email was 30% higher than Facebook ads.
What really drove the economics of this channel however was the virality — a K-Factor of 0.70 was almost 3x what we saw from paid. That means for every 3 people we drove via email, we got 2 more through sharing for free. Referrals from email totalled 201 leads, 22% of the total. Only 39 of the 287 leads (13%) participated in sharing, but they brought in 5.15 new people each on average. Of course that’s highly skewed — the top ten people brought in more than 50% of the viral leads from email.
Additional Email Mention
These three emails were primarily focused on the Emerging Channels campaign, but we did include a mention of it in another newsletter email that went out. These emails go out weekly for the promotion of other blog posts, and we included the Emerging Channel Report in one of the promotional slots further down.
I didn’t include these in the final results because the clicks were minimal, for example the email sent to promote Ladder’s new Adaptive Growth Foundations system included it in slot #3. However this only drove 10 total clicks, so including it in the final results would skew the CTR downwards and make the results hard to interpret.
We started our email campaigns with a personal letter from the CEO, Jon Brody, announcing the initiative and linking to the pre-launch page. In this email we detailed the rewards for sharing (more on that in the Viral Sharing section), including getting the full report for free.
Following that email, we sent a ‘last chance’ email the day before launch, reminding them they could still get the report for free, and then a final email on the launch day to announce the pricing. You can see all the email templates we used here.
The pre-launch email is the first time we announced the new initiative, and it did most of the work to drive new subscribers. We wanted to show off the premium design we made so we put that front and center. This is important to showing the audience that this is a serious project — we didn’t just slap this together in five minutes, we’re putting resources behind it.
For the copy we went with a commonly used tactic, the ‘founder email’ — most people want to hear directly from the founder and CEO, so personal outreach can be an effective way to launch something new and make people care.
We stuck with that theme for email 2, the ‘last chance’ email. We wanted to communicate in this email how much response we had gotten, because people tend to follow the crowd, and participate where they perceive everyone else is too. We also reminded them of the benefit, that they could still register and get the report for free if they shared with friends.
For the final email we sent on launch, we re-used the same graphic from pre-launch, to breed familiarity right at the point where we start asking for money. We made sure to include a benefit (“actionable content”) and reminded clients that they didn’t need to buy. This helped provide a benefit to our client base, but also reminds the rest of the list that we offer growth services.
If we were doing this again, I think the combination of link plus button is confusing, and we’d probably remove the other posts we kept in the template at the bottom. This got the lowest CTR of any of the emails, and though it’s expected to see a drop, I’d expect we could have done more to improve the motivation of those reading it.
As expected with email, the more we mentioned the report, the less interest it got. We started with a 4.5% CTR on the first email, Pre-Launch, which decreased slightly to 4.2% by the second email, then finally we got only 2% on Launch. This is something you should take into account when planning or forecasting your activity.
One interesting thing to note is that the pre-launch email drove 168 signups from 321 clicks — a conversion rate north of 52%! The conversion rate decreased steadily from there, as you’d expect. In addition, this email also drove the most virality, with a viral factor of 1 to 1! This meant that the 168 signups we drove from the first email also invited 170 more people who signed up.
On a less promising note, we counted 135 unsubscribes from all emails, which represents about 2% of the list, and was definitely higher than usual. However with all the sharing activity our list ended up way ahead in size — and people who unsubscribe tend to be lower value and less engaged anyway, so we’re not sweating it too much.
Channel: Viral Sharing
Though Facebook and Email were essential to building an initial seed audience, virality ended up being our biggest marketing channel. We built viral dynamics into the campaign from day one. The mistake most people make is they expect people to just share, and they do in small numbers, but if you really want to crank up the K-Factor (how many new users you get from virality for each seed user), you need to prompt and reward that behavior.
For a digital product that part is relatively easy — we could offer a part of the product for signing up, and more sections to reward sharing behavior. After pre-launch signup there was nothing else we needed them to do other than wait for the full product. Most products completely waste the thank you page, but by prompting social sharing on that page we increased overall signups significantly. These are your biggest fans, who just gave you their email address — give them something more to do, and many will follow through.
Here’s our Viral Sharing funnel:
|Landing page visits||4,833|
|Referrals per Sharer||5.7|
Interpreting the Results
Like most things on the internet, this is a numbers game. The 18.7% conversion rate is something like 4x what we’d expect from a landing page — that radically improves the economics of every channel. With a K-Factor of 0.73, we did fantastically well overall in terms of virality — in B2B a viral coefficient of 0.20 is far more common.
I think the high conversion rate paired with well designed incentives to share led to this outcome. The trick is to add real value for sharing, and to apply most of that value in a kind of barbell — reward people who sign up handsomely, and reward your top sharers well too, don’t worry too much about the middle.
Layered on top of this, we only needed 7.4% of our audience to share, and they drove 42% of all signups! There was a power law in effect, where a small percent of users drove more than 10 referrals, and a longer tail of users drove 1-3. 70% of all referrals were driven by users who earned the free report by referring 10 friends.
Interestingly enough some people kept going after earning the full report — there was 1 person who sent 17 people to sign up. This might be because while they knew how much they shared, they couldn’t necessarily account for how many signups they’d get. Remember this only counts signups that came from these users sharing. There will be a lot of people with zero referrals, that actually did share on social, but didn’t get anyone signed up.
You might be wondering how did we get 0.73 K-Factor overall when Email and Facebook both had lower than that? It’s because virality is a loop. People who get referred, can also refer more people, who then can refer others. To figure this out, we have to look at users that didn’t have a referrer, these we know are the first branch. Then we need to see who they referred to (Kickoff Labs has a referrer email field), and then see who they referred to, and so on.
So from 523 leads without a referrer, we got 273 referrals, a K-Factor of 0.52. For that second branch, we saw a further 58 referrals, which is a K-Factor of 0.21. We could keep going, but because the K-Factor is significantly below one, it’ll diminish rapidly (rather than continue growing exponentially). We can also look at who influenced the influencers — for example there was one user that drove two people, who themselves brought in 10 referrals each.
Note that we did identify some incidents of attempted fraud — people entering in fake emails or duplicates (that redirect to the same place). This is going to be a factor in any campaign where you give away value, but thankfully it was relatively isolated and didn’t impact our campaign too much. Turning on double confirmation (making them log into the email and click confirm) would have hurt conversion, but that’s definitely a good option if you’re worried about fraud. In addition Kickoff Labs have various fraud prevention tools if you purchase their higher tier.
Also important to performance were the ‘reward-level’ emails set up in Kickoff Labs. To see the text of all of the email templates you can find them here. We’ve seen how important sharing was to the campaign, and these emails were designed to encourage that behavior. It only took a little bit of extra time to set these up, but it has the potential to make the world of difference to performance for any viral campaign.
On the morning of our pre-launch, we noticed a problem — 90% of our reward-level emails weren’t being delivered! You can see in the below screenshot we sent 255 emails and only 22 were delivered, 4 opened and 1 clicked. Knowing how important these emails are to encouraging sharing, we quickly contacted support.
Initially Kickoff Labs support thought it was something to do with fraud tagging (a feature we weren’t actually paying for or using), but in the end they figured it out — we had used a comma in the name of the sender, and that was silently breaking the email sends for the main service they used! To their credit they said this is the first time this bug has happened in 8 years, and they figured it out and fixed it in less than 6 hours.
We found a way to send out the emails manually, so everyone got their first 3 channels, though we had no way to track the performance of those initial sends, or how much it cost us in terms of virality (remember these were the most engaged audience, with a K-Factor of 1.0 despite this issue!). We had no further problems with the kickoff labs platform, and they credited us a free month to say sorry.
This was a relatively simple campaign, just two major inputs — email and facebook. The user journey was two-stage, from an unbounce landing page to a kickoff labs sharing page. After sign up we’d get a percentage of users sharing to get more sections of the report, and that would drive traffic back to the unbounce landing page, some of which would sign up, starting the loop over again.
Let’s take a look at how those shares drove people back to the site. First we need to filter for users who had a referral email, then look at the document referrer (what a website passes to your site when they send you a user). This will tell us how much sharing on social helped versus just sharing a link (which would show up as Blank).
In this case LinkedIn actually did really well — a lot of people say LinkedIn is dead, or full of spam, but this is just one of a number of times I’ve seen startups get good numbers from the platform. Don’t count it out yet! However the vast majority of signups from referral (93%) came from people sharing the unique link code.
You’ll notice we also saw about 6% of leads, or 55 leads come through as ‘Other Sources’. These are leads with no marketing campaign source and no referral email — meaning we don’t know where they came from. Every product has it’s ‘direct’ or unattributed traffic, but that’s pretty unusual for a campaign that could only be reached if you know the URL!
Digging into these leads, I noticed I recognized quite a few of them. Multiple ex-Ladder employees (great to see you guys still rooting for us!), current clients and other friends of Ladder from our wider network. Presumably there were also some leads in there where they found the page through other means or someone shared without their unique tracking link. This cohort punched above its weight in referrals — 68 referrals from 56 leads, or a K-Factor of 1.20!
The primary driver of performance for the campaign was the landing page conversion rate. We got such a high conversion rate because we followed a lot of great best practices and because we overinvested in the design of the page with a talented designer. Because the assets looked premium, the fonts were well chosen, it was on brand and the page didn’t look like a template, we drove significant results. To see screenshots of the landing page, visit here.
One important thing that many get wrong, is that you need to have an extremely prominent call to action — in this case our ‘Get a free preview’ button in green. No matter where you start on the page, this bright button draws you in. Because we’re offering real value for signing up, the first 3 channels for free, they may as well sign up and see what they get.
Further down the page we finished with some social proof — the impressive clients Ladder has worked with. This is important to building trust, as anyone who scrolls down this far doesn’t quite trust you yet. Then if this did the trick, they have another convenient CTA right there to get them to sign up for their free preview.
After signing up you get to the share page, built on Kickoff Labs. Just like Unbounce, it’s drag and drop, so it really simplified and sped up the process for launching this campaign. The page is pretty simple — include the report front cover to maintain consistency and breed familiarity.
The headline speaks to a strong call to action, that they could earn the whole report for free. We chose the social media buttons carefully based on what traffic we normally get from social, but in this case it was a very small amount of traffic, so wasn’t that important a choice.
The reward levels were intentionally chosen as a barbell strategy — you get a lot for just signing up, not much in the middle and a lot at the end for getting 10 referrals. The key insight being that getting your top referrers to share a handful more times, is way more important than incentivising the majority of people who don’t share much.
On that note, we did what we could to encourage sharing with our reward-level emails. In each email we made it personal (every email was from the CEO’s real email address), delivered some value, and reminded them of what they could get if they kept sharing. We kept it simple with plain-text, as that tends to work better anyway, and is less likely to be blocked or load slowly for users on mobile internet.
The reward-level emails didn’t take a lot of time to set up, but they did their part in incentivising sharing behavior. We set emails up at key reward levels — when users had made their first referral, had earned a new prize and if they were close to hitting another bonus. If you’d like to see the copy for all the reward level email templates, you can find them here.
In the first week of the campaign, we had close to a 50% conversion rate — this was the effect of primarily the email signup and early virality signups. Over time as Facebook ramped up and the email signups ramped down, we saw reversion to the mean, and conversion for the pre-launch page settled down to around 20%. Once the launch page went live, conversion was still defined as signing up for the first 3 channels preview, but this was lower down on the page — the main CTA was purchases (not counted here). Despite that, we still saw a 4% conversion.
Since the Facebook ads campaign has ended and we haven’t sent further emails, conversions have slowed to a trickle. If we had a K-Factor of more than one, we would have seen something different — conversions could continue to increase over time with no further investment. To learn more about how this works, check out this post on modeling viral growth.
The reward-level emails actually performed admirably once we got the deliverability issues sorted out. From the 606 emails delivered, we got 455 opened, an amazing 75% open rate! Even the click rate was relatively high at 15.5%.
Overall we were really happy with the performance of this campaign — close to 1,000 leads for a project that took about two weeks to pull together isn’t shabby! More than the leads though, I think this campaign was a great experiment that proved virality campaigns still work, that people still share products they like the sound of, when they are receiving good value. The actions that most contributed to success in my opinion, were overinvesting in design, keeping the Facebook campaign simple, and setting up a system that encouraged post-signup sharing.
If we wanted to make this campaign much bigger, I think the major opportunity isn’t in optimizing the landing pages or emails (though they could always be improved), it would be in baking virality deeper into the nature of the product. We could have interviewed popular marketing influencers who use these Emerging Channels, as well as the Product Marketers in charge of promoting these platforms, and worked out a cross-promotional deal. This would have given us a far larger seed audience without needing to invest more in Facebook ads.