The TikTok Strategy Behind 850K App Downloads w/ Sorce

Learn this proven TikTok strategy for getting mobile app users! 📈 Achieve profitable & scalable growth through paid ads here ➡️ https://perspectivefunnel.co/6983af76f4d6ec0083341606/698b94e6dfdc8963496731d6/

Ben Abbott built Source, the "Tinder for jobs" app, and scaled it to 850,000 users almost entirely through organic TikTok, with his third video ever hitting 1 million views. In this episode, he reveals why copying viral formats beats creating original content, the 5-second retention rule that predicts whether a TikTok will go viral, and how a team of engineers with zero content background now ships 750 UGC videos a month. If you're building a mobile app or trying to crack organic growth, this is the playbook you need.

⚡️Connect with Ben:
➡️ https://www.tiktok.com/@careerben

If you’re running Meta ads for a mobile app, this episode is a must-watch.

Subscribe for future episodes where we break down real growth systems, performance frameworks, and AI-powered marketing.

Transcript

SPEAKER_01 0:00

Starting to scroll TikTok for a couple of hours every day was a huge game, but it's really helped. I think you you just see what type of viral UGC is popping up on your feed.

SPEAKER_00 0:08

You have taken a mobile app, scaled it, and taken it through YC. What were you seeing the other YC consumer mobile apps do for growth?

SPEAKER_01 0:14

I think UGC is actually in a meta right now. Right now, we probably have around 25 TikTok videos going out per day. So I think organic can get the lowest cost per install. Something that we do that's different is looking up retrenching we have to done.

SPEAKER_00 0:26

If you're a new mobile app, what's the first thing you do? All right. Well, Ben, welcome to Valid Podcast. Uh you have worked with Valid for several months in the past and then also has taken a mobile app, scaled it, and taken it through YC. Um, welcome. Yeah, thanks for having me. It's great to be here. Um, maybe just to get started, how about walk us through what is source? How did you come up with the idea? What was the like what was the origin story? And yeah, what brings you here today?

SPEAKER_01 0:52

For sure. So the source of the Tinder for Dot's app and the origin story in the basically the app, me and my friends were applying to Fog Joe. So it would take hours every single day, filling out the same origin experience and same education over and over again. And kind of just started to feel like a waste of time. So we basically created this app that when you fly, you we sent up a browser in it to find the drug for you.

SPEAKER_00 1:13

That's super interesting. Like, why why why why source over using like a traditional job board or something like that?

SPEAKER_01 1:20

So traditional job boards just take a really long time to use. We were essentially filling out the same information, just built an AIA text to a for you. And as soon as we launched that, it went super viral. Can you expand on that?

SPEAKER_00 1:31

Like what it like, how did you how did you launch it? How did it go viral? Why did you choose a mobile app? How did you make it go viral? Um yeah, like walk us through, walk us through the virality there.

SPEAKER_01 1:39

For sure. So they just launched with actually on LinkedIn. We built a Chrome extension that would help you apply the jobs quicker. After that, we kind of decided that the easiest form factor was an app because if someone could swipe on an app like Tinder and we could handle all of the back end for them, we figured that that would be the most popular. After the initial LinkedIn launch, we just started making TikToks and we stumbled on format really luckily that worked. I think it was like our third Video ever that we posted, got a million views. So we kind of just discovered an organic format that worked well. We were the spam posting on TikTok, and then we just shift that to, I guess, probably at this point, like 850,000 users.

SPEAKER_00 2:13

I guess we need to get super tactical. Like how, like, what does spam posting on TikTok mean? And um, you went from a Chrome extension to a mobile app, and then things took off. How yeah, like sort of like why, like what prompted you guys to realize, hey, we should go from Chrome extension to mobile app, and why do you think that worked?

SPEAKER_01 2:29

And we realized that as we were making TikToks, uh TikToks with the mobile app would go way more just because it's so much easier to go from a TikTok downloading app versus going on your laptop downloading downloading the Chrome extension. There's just way more friction there. That's kind of how we decided on the form factor of the app. I guess what you would call UGC user-dever content. It's basically getting a bunch of creators to post on new new fresh accounts, just videos of them using your app in an organic layer. And you kind of want to scale through numbers instead of just using typical influencer mod, getting it super cheap per video. Right now, we probably have around 25 TikTok videos going out per day. And if one of them takes off, then it basically pays back the entire truck. So in a given month, we'll be having 750 videos going out.

SPEAKER_00 3:09

And then how are you like tracking videos and like how do you know what's doing well, what's not doing well?

SPEAKER_01 3:15

Good question. So there are a bunch of really cool tools like Sideship, viral. In terms of how we determine that a video goes viral, we found a few key metrics that contribute well to videos going viral in our format. So let's say a video has 50% of people still watching at five cents. If that's the case, then it has a decent chance to go viral. Because that's like not every video with really good metrics will go viral, but all of our videos in App One Vie have had these specific metrics.

SPEAKER_00 3:41

Interesting. So it's like 50% of people are watching through half the video. That means like, okay, this is a good format, it's worth recreating. It likely led to conversions and virality and all that good stuff.

SPEAKER_01 3:51

And literally all you do is look at the retention graphs, see if people are dropping off at specific points in the video, figure out how you could fix them, and then you just get as many people like to complete the video as possible.

SPEAKER_00 4:01

And then when you're launching, let's say like 20 videos a day, is it from the same set of creators? Is it the same video format over and over again? Like uh yeah, I'm so curious how how like how you put up that much content um and and monetize.

SPEAKER_01 4:14

At this point, you basically have a roster of 25 creators that are really good, have we really like so we we kind of trend through a lot of people because it's kind of a numbers game at the end of the day, like how at least 20% of people are gonna be good. We generally do 80% of the posts just in one format, and then 20% of the time we let the creators kind of decide where we'll come up with new hugs new formats to try out. But usually we like to just stick to our random money.

SPEAKER_00 4:35

And then do you do anything when it comes to I don't know, like competitor research or trying to figure out you know what is our next big format? Like is there like do you do that or is it like let's just do what works and do it over and over and over again?

SPEAKER_01 4:49

So the 20% for us is experimentation. A lot of my time is just spent scrolling TikTok. I used to not even have social media on my phone, so starting to scroll TikTok for a couple of hours every day was a huge pain, but it's really helped. I think you you just see what type of viral UVC is popping up on your feed.

SPEAKER_00 5:04

Yeah, I almost wonder if you guys have this within your company. We have it at Valid, we have this Slack channel, we call it Awesome Creatives, and it's just ads that are sent over and over and over again. Um, it's also like really interesting because you would think that um like the same type of content will resonate with everyone, but it's uh it's really interesting to see um what type of ads it like engage what type of person, also what type of ads are served to what type of person, because some of these ads will see them and like this is an amazing ad, but that would never show up in my feed. Um, do you do you guys do stuff like that within source to almost like send content to each other and uh have a have like an ever-growing squake file or something like that?

SPEAKER_01 5:38

Yeah, we're not as official as to having our own Slack channel just for this, but I everybody on my team will just text me a bunch of videos that they see that go viral. It's really interesting the factors that can cause videos to go viral. For example, you might have a video that goes viral when your creator is like an older, like looking like a mod type for if you're like talking about remote jobs.

SPEAKER_00 5:57

I did want to ask you about uh copying. This is our first strategy, our favorite strategy at valid. We say like valid think the same. Um, so what I mean by that is when we're trying to come up with new ad concepts for our clients, what we try and do is find ad concepts that are working for their competitors and then we take them. Do you do that in organic? Does that work effectively? Or um yeah, I'm curious if you uh like how how do you how do you copy, if at all?

SPEAKER_01 6:21

Yeah, so as I was saying earlier, I don't actually come from a background of scrolling TikTok and being brain audited. So generally, like I'm really good at operational stuff, but I'm not super good at coming up with content from scratch. So we definitely do copy.

SPEAKER_00 6:33

I love that. I think one of the craziest things that uh I see people do is come up with original content. Um, I just never. So let's let's copy. That's easy. That that works. Um I also want to I wanted to chat a little bit about paid ads as well. So I think organic is the is the lifeblood of source. Um You obviously ventured into paid ads with valid. Um what sort of took you down that path? What are the how how do you compare organic versus paid? Um what what kind of uh users do you think should, or like what was not users, but what sort of apps should be thinking about having a more paid uh heavy approach versus a more organic heavy approach? If you're in the mobile app space and are struggling with paid ads, we'd love to chat. We'll even go to your ad account and tell you what you should work on. Just head over to valid.co, click book demo, and we'll get in touch. Now let's hop back in with Alex.

SPEAKER_01 7:20

So I think the best way to think about this is that the best way to create big content is through organic, and then you want to scale the content with that. So I think organic can get the lowest cost per installs. I think if you get to the point where your LTD is greater than your cost to acquire the user and the cost to maintain the user, then you should just start like using a bunch of paid ads.

SPEAKER_00 7:37

Yeah, that makes sense. So sort of like start with organic, get the get the first set of users, start to understand your your unit economics, and then use paid ads to kind of like take that knowledge and scale it beyond. Um yeah, that makes sense. We've we've seen that across Twitter, across Cali. Another one I wanted to ask you about is uh, and this is is like how um when when we worked together, Source didn't have uh like a monetization event. So on on paid ads, like what was the the objective uh in in running in like running paid ads at the moment? And like uh yeah, how did you how did you measure success?

SPEAKER_01 8:09

At that point, our objective was kind of just to understand paid ads. I think we're only able to do this because we're like a VC backed startup. So Andrew Boostrubb, it may not be as good of an idea.

SPEAKER_00 8:17

Yeah, no, that's fair. Um, that's that's one of the benefits of doing uh of like being a VC back startup, right? You get to experiment faster and kind of like play like your four steps ahead of where you actually are. Um we benefit from that as well. We're also a VC back startup. Um actually, along the lines of a VC, uh, you went through YC as uh mobile app. Yeah, can you like walk me through what that process was like? What was what was your interview? What did they ask you? Um, yeah, and then how like you know, what what were those kind of two, three months like?

SPEAKER_01 8:46

Yeah, so our interview was actually pretty unofficial, but I put the deadline. The whole experience of YC was really great too. They generally segment the batches by industry type. So we were in a consumer patch. Can you walk me through the interview?

SPEAKER_00 8:56

Like, what's the what what did they ask you? What did you say? Um, what do you wish you said? And uh yeah, some super curious.

SPEAKER_01 9:03

I don't think there's anything that we wish we said differently. A tip, I guess, would be to find someone that you know that did the interview and ask them for as many tips as possible. But generally, all the tips, generally all they're gonna ask you about is they just want to figure out about your company and about the founders if they're qualified or not. And your answers are generally supposed to be less than a minute to a minute and a half. So super concise answers, just telling more about your company.

SPEAKER_00 9:27

Um, and then you you mentioned that in YC you're put in these these batches where you're like working alongside a bunch of other companies. I I assume you're with other consumer mobile apps. What were you seeing the other YC consumer mobile apps do for growth?

SPEAKER_01 9:40

I think UGC is actually in a meta right now. Everybody was either using an agency to run their UVC or creating their own organic content that's scaling it through PayPal.

SPEAKER_00 9:49

Then actually one other question, Brady. So you mentioned some are doing doing it in-house, some are using an agency. Uh yeah, like how how did you see that split? Do you think it was uh like yeah, like what like what what types of companies were working with agencies versus doing it in-house? Um, maybe what's the what's the trade-off?

SPEAKER_01 10:04

I think usually smaller teams were doing it with agencies just because they didn't have the manpower to start a growth program from Smash. I think agency will definitely give you better results than this car. I think the benefit there in doing it in-house is just that you gain domain knowledge that an agency will never, I guess, give you for free.

SPEAKER_00 10:20

As a as a as an engineer ensuring like TikTok brain rot, what what shocked you? And then like what maybe what's what's really weird about your approach that maybe like a traditional agency or someone who's from the like you know, for from that world would uh like would not do.

SPEAKER_01 10:36

I honestly found it interesting how addictive these products are. If they myself off social media really helped my productivity, but it's really the only way to understand the algorithms, find viral concepts. Something that we do that's different is looking at retrenchment graphs a ton, all of us are engineers at source, none of us come from a content creation background. So it's really interesting to break down the videos by who's watching Apple. We don't really do any other, like I guess, qualitative analysis. We kind of just see, oh, 20% of people dropped off between seconds and four, have the move fixa. We'll hypothesize, test with the creator, and then if it doesn't work, we'll test it again and just until do you um like you see with with your creators?

SPEAKER_00 11:14

You mentioned like 80% is from like is you want them to run what what what's tried and true, what works, 20%. You want some some new shots on goal. Um with that 80%, how are you conveying that information to them? Is it is it briefs? Is it very is it like hopping on a call? It's um you know just kind of like getting very tactical. And like how do you how do you scale a UGC program?

SPEAKER_01 11:34

We usually just create briefs. Uh that is the most scalable way of that beef valve. We can text out a brief to everybody just through messaging API, which has been super helpful. In terms of feedback, we're usually really hands-on.

SPEAKER_00 11:44

Have you messed around with like AI UGC at all? Have you seen that work or not work?

SPEAKER_01 11:48

I'm actually experimenting with it right now. We've been able to get a lot of use with it.

SPEAKER_00 11:53

What what what's your what's your stack for AI UGC?

SPEAKER_01 11:55

So far, I've just been using Cling and add Open App.

SPEAKER_00 11:57

Nice, yeah. How is it? Like, is it extraordinarily painful or is it pretty quick?

SPEAKER_01 12:01

Or it's actually a pretty painful process right now, just because I don't understand the tooling. But all of these providers have APIs, which is really cool. So I think eventually we could set up a really good content form, which is kind of why I'm looking into it in the first place.

SPEAKER_00 12:12

Yeah, no, that makes sense. Yeah, that's actually like one of the big uh like focus areas for us as well, as we now do um like we we do video creative, but we actually uh recently like moved off of all of our creators and now we only do AI UGC. Um but it's extraordinarily painful to make the make the content and to especially to make it look good. I think those are actually like particularly interesting for uh like organic and paid ads because um with paid ads, if you have an influencer, you can like basically run everything as a Spark ad or Dark Post. Um and like Spark ad is where like you know your creator posts it and then you just take that post and you scale it. Uh dark post is you have an ad, it might not even be from that creator, but you run it from that creator's page, then that'll actually reduce CPMs by 10% or so. Um so having these like AI creators let's you have a bunch of accounts that uh you know, assuming they're calibrated to your audience. Like if you have, I don't know, like an account that's about nail polish, probably wouldn't help source much. But if you had an account that was about um, I don't know, like maybe like something related to interview prep, uh, and then you ran ads for source from that account, I would assume it would do pretty well. So there are actually agencies out there that all they do is dark post. So they just uh they they produce um they basically produce uh like influencer accounts, and then they have no in they have no interest in like traditional influencer marketing where it's like the influencers that you know it's like you might have an influencer be like, hey, like buy water, it's it's good for you. Um they they have no interest in that. All they want to do is uh sell the rights to that influencer account to other people who are buying ads to make in Dark Post because it reduces CPM, they charge you a percentage of spend. Um so yeah, it's uh it's it's pretty cool. I think also with AI, if you're if you set up a good content play wheel, you can scale that infinitely and go into a bunch of different niches. Cool. Uh definitely we've chatted for a bit on uh lots of different things here. Uh one final question for you. If if you're if you're a new mobile app, um actually two questions for you. If you're a new mobile app, um, what's the first thing you do? And then if you're a mobile app at 100k trying to get trying to get to that next tier, uh what do you do?

SPEAKER_01 14:14

I think if I was a new mobile app, I would just make Pygmin designs and try to vote a bbc. See if the video went viral, and then if it goes viral, just start creating ship a beta version within a week and hit up everybody in that reactor click interest in the video and just have them keep beta testers. If you want to scale an app for the 100k amor, I think you start you have to start using K ads just because organic content has a limit. And I think scaling paid ads once you get 100k mr is the way.

SPEAKER_00 14:38

Awesome. That makes sense. I appreciate it. Uh, any any final notes of wisdom for folks uh who are either starting mobile apps or are interested in getting into YC and scaling a business from there?

SPEAKER_01 14:48

In terms of getting into YC, I think all you need is a bit of traction and they'll be very interested. You I fly with a different startup that had very little traction, and they were super interested as well. Uh so really just if you want to get into YC, start building something, get some users, and then apply.

SPEAKER_00 15:03

Well, what do you what would you call like a little bit of traction? Again, you have to kind of like how much traction do you need to peak YC's interest?

SPEAKER_01 15:09

I think at that point on my other start, there's more B to B thing. I was commenting customer service service for some people, but I think I had 20 customers. Okay.

SPEAKER_00 15:20

Awesome. Well, thanks so much for coming on, Ben. It's uh yeah, honestly. Love, love working with you and like oh I was like super interesting to hear about how you're how you're scaling source. Uh any anything that people should know if they're interested looking for a job.

SPEAKER_01 15:31

Um well, so I think source is the best way to buy it a job, obviously. We helped automate a bunch of applications. But other than that, nothing really. Uh thanks for having though. I really appreciate it, and I hope we can work together. Yeah, awesome.

SPEAKER_00 15:42

Thanks, Ben.