Lessons learned scaling paid ads from 0 to $2+ million in revenue in 2017
Soon after Jumpcut was founded in late 2014, we started playing around with paid ads on Facebook. But there was one problem: nobody could dedicate their time to it. There were too many other fires that needed putting out.
That’s when we had a genius idea: let’s assign it to our intern. We put him in charge of paid ads, and then sent him an email with a profit-sharing goal:
Bonus is 10% of all paid ads profit during your time here. Very simple. Obviously, if we lose money on ads, we won’t be holding you to anything at all, but if you make any profit, you’ll see 10% of that.
He was smart, and we put a good incentive structure in place. How hard could it be?
After six weeks, we were $1,000 in the red, and after looking at how little sustainable long term progress had been made, decided to give up the test. Over the next two years, we tried similarly silly tests, in bits and spurts, to get paid ads to work. People would spend a week or two here and there. Nothing worked.
It was only towards Q4 of 2016, when when our burn rate was going up and our revenue was going down, that we made a decision as a company to do it right: put a full time employee full time on paid ads and just make it work. We put our best marketing person on the job, hired an external consultant who came with great referrals to advise us, and treated the initiative as a long term project, instead of a quick win.
It worked. Over the last year, we’ve grown revenue from paid ads to over $2 million dollars for our online courses after spinning our wheels for years. Paid ads have become the most important driver of revenue for us. Single campaigns on Facebook have made over 6 figures in revenue and are still profitable almost a year after creation.
Here’s just one example of an ad that has gone “viral” for us:
That’s 3.4 million views, 17 thousand likes, and over 7,000 shares on an ad that isn’t even our best content. That’s just one example of how we’ve gone viral through paid.
Doing this — mastering paid ads — is what has helped us transform from a cute, small online business to a real company. Because of paid ads, we’ve been able to connect with our core audience online and scale our course offerings to everyone who needs to see it. The revenue generated from paid ads have been the engine of growth for Jumpcut over the past year: our increased content production and headcount have all depended on our ability to sell through paid.
Of course, it wasn’t all roses getting to this point. We hit dead ends and burned a ton of money before we were able to hit profitability. We’re writing this post now to help you avoid wasting the same amount of time and money.
Here are the three biggest AHA! moments we witnessed over the past year with paid ads. We’re hopeful that you’ll be able to apply them to your journey as well.
1. Test everything, even when it doesn’t make logical sense.
In Q3 2017 alone, Facebook made over $10 billion dollars in revenue, mainly from showing ads to its users. Google did over $27 billion in the same way.
Just to put those numbers in perspective: we pay around $4 per lead we get from Facebook. If every company had the same cost per lead as us, that would mean Facebook collected over 2.5 billion email addresses in the last 3 months — or 321 email addresses per second from ads that they’re showing you (mainly) in the newsfeed.
These are huge platforms, and it’s all controlled by software. Their algorithm controls how your ads get seen, and despite having a really impressive amount of customization, the distribution mechanism is still sometimes a black box, where you don’t know how one subtle decision you make will affect what your end user sees. Because of this, there are holes and logical inconsistencies that don’t make sense. Innovative startups can use this to their advantage. That’s why you should test everything, even if it doesn’t logically make sense.
Here’s an example. When we were playing around with our ad sets one day, we decided that we would take an ad that was running at $200 a day and duplicate it. We created literally the exact same ad, and then ran both identical ads at $100 a day. The result? The 2 duplicate ads at $100 a day did much better, in terms of cost per lead, than running just one ad at $200 a day.
Why? Who knows. But it worked.
Here’s another example. One of the most annoying things about Facebook (thanks, Russia) is that when you submit an ad, you’ll have to wait a certain amount of time before your creative gets approved. Instead of waiting for the arbitrary yes/no decision from Facebook, we would just duplicate the ad we wanted approved multiple times, wait until one of them was approved, then delete the other ones. On average, this shaved nearly 2 days off waiting to run our ads. (For Adwords, this hasn’t worked — but what has worked is calling them, which has shaved our approval time down to 10 minutes.)
There are all sorts of things you can figure out when you play around with the software and really immerse yourself in it. Follow the step by step guides they give you, but go off the beaten path as soon as you feel comfortable. Question what you’re doing and think of ways to do it faster.
(Also, if you have a Facebook or Google rep, take their advice with a grain of salt. They are here to help you, but don’t trust them blindly. We’ve found that things they’ve said have crumbled under the weight of our own testing.)
By testing different ideas, we’ve found ways to improve how many people see our ads, lower our costs, and improve our profits. Here are just some examples of things that worked for us:
- Using the existing post feature on Facebook instead of making new ads each time. By doing this, we were able to show the same ad to different groups, and bring old comments along to those new audiences so that they were more likely to share and comment.
- Reusing old ads on new audiences. When an ad “deprecated,” instead of retiring it completely, we used it again on a new audience…and it was as good as new. No banner blindness.
- Using ads where the narrator is similar to the audience. We’ve found that using a female testimonial on a female audience works wonders.
…and here are some things that haven’t worked:
- Trying to target people who show search intent with Google Adwords (e.g. people who search, “how to make money on YouTube.”) After testing, we realized that just because they Google search the term doesn’t mean they actually want to pay for information — many of them are just curious about how people make money on YouTube, which (surprisingly) doesn’t actually equate to them wanting to pay for a course on how to make money on YouTube.
- Targeting people based on who Facebook thought were “high spenders.” Just because people spend money on other people’s products doesn’t make them more likely to spend money on yours — they are just more expensive to target.
- Using ads that worked on Facebook on YouTube and vice-versa. We had confidence in this one, given how much success the ads had in their respective platforms — but when we tried them on the other platform, it became the worst performing ad on that platform — low engagement and even lower click through.
Put together in the aggregate, the impact of testing has led to huge gains on both platforms. If we did everything by the book, instead of constantly testing, we’d be making 50% less than we are right now.
The takeaway? Test everything. And don’t get too caught up in losing money when you’re running these tests. View it as the price you need to pay to learn the platform, as long as you’re taking a scientific approach. Do things even if they’re not what’s recommended, or intuitive — you never know what will work.
2 — Use a marketing funnel.
Way too often, I see ads on Facebook that look great. They’re eye-catching, written persuasively, and direct me to a sleek looking landing page selling their product. But there’s only one problem: that landing page is usually the only contact I ever have with that company, ever. And that’s because once I exit out of their landing page, I’m done.
Especially for companies I’m not that familiar with, or those with a high-priced item they’re trying to sell, there’s no way you’re going to convince me to use my credit card information on the first visit, no matter how well the page is designed.
There is, however, something that I’m much more willing to do: give you my email address.
This is what 90% of marketers get wrong with ads. Instead of trying to capture an email, they try to make the sale instantaneously. The reason why we’re profitable on paid ads is because their sole purpose is to capture the user’s email address. Once we get it, we enter a weeks (often months) long relationship with them via email, as we establish our credibility and give them value. We’re not betting the farm on the one minute they take to read our landing page. Instead of losing them forever, we maintain a direct line of contact.
And after you get that email, what’s even more important is creating the right email funnel for the user. Our philosophy is pretty simple: the longer emails, the better.
Most people don’t know what Jumpcut is. We owe it to ourselves to give them the most valuable information. And we don’t hold back. Our emails are 7 pages long. Our videos are 20 minutes long. Because the content is interesting and valuable, it gets our customers to trust us more, and helps us develop a “true fan” relationship with them.
A mistake that we see startups making is sending the shortest emails possible after getting someone’s email address. We think that’s the wrong thing to do. Don’t be afraid to send out much longer emails — you won’t be bothering them if the content is great. In fact, we get customers who reply to us and say, “I knew you were going to sell me something at the end of this, but I don’t mind, because I learned so much from this.”
As you’re creating your email marketing funnel, make sure to focus as much — if not more — time on it versus your ads. In general, optimizing the value of the funnel is more important than optimizing your cost per lead. Why?
It’s all about sizing the opportunity. The ceiling on improving your funnel is effectively infinite. Every step in the funnel where you increase the click-through rate (e.g. decrease the drop-off rate), and every dollar you can add to their average purchase value has a multiplied effect on each other as your user buys from you.
But there is an absolute floor on how low you can get your ad costs — the CPL, CPC, CPM (if you don’t know what these mean…go look them up) — because it can’t go to 0, and it gets harder the lower you go.
We’ve found that it’s much easier to boost the value of a lead than to reduce the cost of a click. That’s what your funnel is there for. It will market your company, educate the consumer, and turn a stranger into a fan. If your funnel can turn someone who would have never bought something from you into someone who will spend $2,000 on you, it doesn’t matter as much that it cost you $17 to acquire that user, versus $5. You win when you educate your leads. Focus on the funnel.
3 — Targeting different niches through video works really well.
This is something that we’ve barely scratched the surface with, but targeting different niches on YouTube and Facebook through video is something we see barely anyone doing, despite how profitable it’s been.
Currently, when you’re on YouTube, most of the ads you’ll see playing are generic — right now, I’m getting a bunch of Grammarly and Wix. But they have this sleek feel that doesn’t talk to me, personally. It feels like an ad. That’s what you’ll get with most of the video ads you see online right now: a company will make one video, and then play that video to all niches.
In contrast, we’re running ads that are directed to a specific niche. For example, what’s worked for us on YouTube is creating an ad that speaks to every animator on YouTube. In the first 5 seconds, we will literally say, “If you’re an animator, you’ve got to listen to this.” And suddenly — you’re speaking to them! People pay more attention, and it’s partly because nobody else really does this yet. This type of personalization is undervalued.
The main reason why people aren’t making these custom videos for niche audiences is because it seems like too much effort. But there’s a way to scale it. In fact,we were able to make 7 different ads in the span of 2 days, all for different audiences. All you do is make a general video, and then change specific parts of it to fit every specific audience. Here are some of the successful ads we made, where the first 15 seconds are different from one another, but the rest of the video is exactly the same.
Here’s our ad for general audiences, which we use for targeting lookalike audiences and remarketing audiences:
Here’s that same ad for animators:
And here’s that same ad for musicians:
If you find yourself wanting to do this, don’t make the mistake of thinking that you have to make one new video for each niche. Find a formula you can tweak at the edges, and then apply it to every niche you want to hit.
Bonus — Create great ads.
The three above mentioned tips will work 100x better if you have great ads. Viral ads. Ads like these:
Keep in mind that people don’t think with logic. They think with emotions. The best videos will elicit an emotion from you: Happiness. Sadness. Anger. Nostalgia. Hope. Laughter. Inspiration. As you create your content, ask yourself this important question: “What do I want my audience to feel?”
Much of the best content online does this — and, what’s really important to note, is that it can be done without a huge production budget or slick visuals. Emotion can be captured on your iPhone camera; you don’t need a set of movie-production level cameras, like Blackmagics.
Make sense? You’ll need to evoke emotion if you want to capture the imagination of the viewer and get your ads shared organically.
Creating great content will have to become one of your core competencies if you want to succeed in paid ads. This isn’t one of those situations where you can set it and forget it. Ads fatigue over time. When people see too much of the same ad on YouTube or Facebook, they’ll skip over it. Banner blindness is a real thing, and we’ve found that creatives need to be refreshed once every 2–3 months to prevent costs from increasing dramatically. (Sometimes, it’s as simple as flipping a vertical subtitle bar to horizontal, or changing the title card from green to red.)
Creating great ads isn’t a passive endeavor. It’s an active, creative one, and while it’s a lot of work, it’s one of the best ways to make your company scale.
Closing thoughts: we’ve only scratched the surface
This isn’t a humblebrag, but we had over 50 strategies and tactics for Facebook and YouTube ads that we didn’t include here because this guide would’ve been WAY to bloated. Just know this: even if we shared all those tricks with you, we’re only scratching the surface. We believe we’re less than 5% deep in this journey.
There’s so much more to learn — Facebook messenger ads, coding our own bidding strategies, and creating ad funnels… just to name a few.
Whatever your skill level, I encourage you to join a group or community where you can share and aggregate your learnings on paid ads. There are groups for all levels out there: from newbies to those with some experience to those who have spent at least a million dollars. And if you’re truly just starting out, there are some courses that we’ve found helpful, and even agencies that we’ve worked with on a contractor basis to help us get over some early bumps in the road.
Above all, stay patient. The first two years of Jumpcut were marred by ineffectiveness with paid ads because we didn’t put in the time to experiment with the platforms. When you’re ready to dive in, dedicate at least one employee’s full concentration to it, and stick with it, especially at the beginning, when the learning curve is the steepest and most expensive.
Once you get over the hump, the rewards are more than worth it — once those base learnings are in place, you’ll be able to develop a culture of success in paid ads. Since we’ve started, we’ve been able to train and hire other employees who have never done Facebook or Google ads before, to the point where, within a month of starting, they were scaling our current initiatives 2x, cutting costs, and taking over the entire day to day. Create this culture of apprenticeship and knowledge sharing, and your organization will only continue to grow.
Good luck!
Cover Photo by Vivian Lau.
