![](https://cdn.prod.website-files.com/63da75e1ec536b89a802545a/64f6794bb493652894fdd9d2_Karan%20Jassar.jpg)
Karan Jassar is the Founder and CEO of Socioh, a digital advertising platform for e-commerce brands to create and run profitable ads on Facebook and Instagram. Karan is the technical architect of the company’s first-party pixel and branded catalog and consults eight and nine-figure brands on improving meta-ad performance. With 17 years of experience in the technology industry, he has in-depth knowledge of advertising and ad optimization with a specialization in AI. Before founding Socioh, Karan was a Senior Manager at Yahoo, where he helped create its new search optimization engine.
Here’s a glimpse of what you’ll learn:
- Karan Jassar shares how Socioh resolves marketing issues for e-commerce brands
- The effect of pixels on ad campaigns across social platforms
- How Facebook’s new model is influencing the marketing industry
- The difference between cash and accrual-based accounting — and the advantages of each
- How long should you run a test before determining a campaign's success?
- Advice for digital marketers on how to create successful ads
- Karan divulges how personal struggles have shaped his leadership abilities
- Strategies for maintaining mental health during stressful periods
In this episode…
Evolving regulations on social platforms directly affect how e-commerce brands manage marketing campaigns and conduct business. While major platforms like Facebook are leading the branding revolution, improvements must be made to prioritize building customer relationships. What are the downfalls of these upcoming developments, and how can you discern the most efficient ad campaign strategies?
By 2024, it’s estimated brands advertising on Facebook will be required to set up a shop through the platform. A one-stop shop is strategic from a business perspective, but until the user experience improves, e-commerce brands are hindered from building customer relationships. Karan Jassar, an e-commerce branding expert, explains Facebook’s strategy is ideal for attracting new customers, but lacks in providing brands with essential customer data. He recommends reviewing campaign performance to determine which channels are yielding desirable results.
On this episode of the Up Arrow Podcast, William Harris welcomes Karan Jassar, Founder and CEO of Socioh, to discuss how ad campaigns contribute to brand success. Karan shares how pixels affect ad campaigns, Facebook’s influence on the marketing industry, and the differences between cash and accrual-based accounting. He also advises how long a campaign should run and strategies to produce a successful ad.
Resources mentioned in this episode:
- William Harris on LinkedIn
- Elumynt
- Karan Jassar on LinkedIn
- Socioh
- The Facebook ROAS Death Spiral: Fix it Now — or Lose Your Shopify Store by William Harris
- Feeling Good: The New Mood Therapy by David D. Burns
Sponsor for this episode...
This episode is brought to you by Elumynt. Elumynt is a performance-driven e-commerce marketing agency focused on finding the best opportunities for you to grow and scale your business.
Our paid search, social, and programmatic services have proven to increase traffic and ROAS, allowing you to make more money efficiently.
To learn more, visit www.elumynt.com.
Episode Transcript
Intro 0:03
Welcome to the Up Arrow Podcast with William Harris, featuring top business leaders sharing strategies and resources to get to the next level. Now, let's get started with the show.
William Harris 0:15
Hey everybody, William Harris here. I'm the founder and CEO of Elumynt. And the host of this podcast where I feature experts in the D2C industry sharing strategies on how to scale your business and achieve your goals. Today's guests I have with me is Karan Jassar. Karan is the Founder of Socioh and technical architect of Socioh's first-party pixel and branded catalog. He currently consults for multiple eight nine figure brands to help them improve their meta ads performance. Karan has a graduate degree in computer science with a specialization in AI from the University of Southern California, and has been working on digital ads since 2006. Before founding Socioh, he worked at Yahoo Inc, for six plus years where he was a part of their display ads team and later helped build their new search marketing system, think AdWords, but for Yahoo, Bing search. And Karan I believe we met through Twitter, I don't know if there was any introduction, we just started engaging on each other's content. We liked a lot of what we had to say. There's some big uproar, I think about something I said about Simpsons paradox and maybe taking some things into your own hand, instead of just trusting Facebook could do everything. It was a wild, wild west, a little bit on Twitter, and we've just kind of gone back and forth from that. Does that sound right to you?
Karan Jassar 1:21
Yes, I think you're I seem to have a very shared point of view on how attribution and Facebook Ads work. And we tend to agree a lot, sometimes disagree, but mostly we are in agreement. And that's how I think we connected.
William Harris 1:36
Yep. Yeah, we have we have to disagree a little bit. That's the whole point of it. Right? We have to have disagreements and disagreements, what are allow us to both collectively learn and grow from each other?
Karan Jassar 1:46
Perhaps so I think, I think as long as you have strong opinions heading loosely, where you're open to other people's feedback and opinions, I think you grow and learn
William Harris 1:57
with that. Absolutely. Yep. And I think that that's something that, let's just say like that's a an allegory into social wellness as well. Right. And everything else that we see that's going on in the world is being able to have yes, you're gonna have opinions, you're gonna have ideas, you could have thoughts, and you're going to have to at least hold some of those loosely, so you can learn from each other and grow as well. Yes, absolutely. So let's get into the quick intro here. This sponsored by Elumynt. Elumynt is an award winning agency where we optimize campaigns around profit. In fact, we've helped 13 of our customers get acquired the largest one sold for 800 million. And we were ranked by Adweek as the 12th fastest growing advertising agency in the world, just last year. So we're pretty excited about that. If you want to learn more check out element E-L-U-M-Y-N-T.com. But that's enough of that boring stuff onto the good stuff. Karan, tell me a little bit about the backstory of Socioh what made you say, this is what the world needs? I'm going to dedicate my time and attention into building this.
Karan Jassar 3:04
Yes. So as you mentioned, in your introduction, I used to work for Yahoo. And at that time, my wife had visa issues, she couldn't really pick up a full time job, even though she had a master's degree in communication. So she started first an eBay store, get an Etsy store, and she was a small seller. And I was working at Yahoo, doing digital ads, and she was struggling to sell stuff online. And it's like, how is it possible that you don't have the right set of tools? I mean, there is so much happening in digital marketing and advertising. So I decided her out. And I figured that a lot of the tooling a lot of the the software, and I'm talking about like 2008 Nine, then that was really sure Shopify was barely coming. See that that point? Right. And there used to be a company Groseclose or something like that, that they acquired Profasee Right. I think they paid us to acquire that company. So it's like way back then. Then we'd open the pros tools website. But that's how you got into e-commerce with my ad Yahoo My job was Sir react launcher. So why worked on display ads, being the ad media, the exchanges and then gone to AdWords equal and, you know, the search, search search advertising word, then trying to help my wife, I realized there was like in Google an opportunity to kind of help the long team of sellers. So and at some point, Yahoo sold their search product to Microsoft. And at that point, we had the option of either going in joining Microsoft or just you know, taking the severance package and moving out. I choose to do the latter because a I got it The so we I decided to take the severance package when they wanted me to stay six extra months. And they said, if you stay for six months, we'll give you two years salary. And that is hard to use. So absolutely, I took that option. I had all the data and light systems to Microsoft. In fact, I was the guy who did the data migration across the two companies. So we are licensed on Yahoo. And we can have moved into a live system on Bing without dropping. But that was my last corporate job. And then, you know, I decided I took some time off because and we'll get into this later in the podcast, because my wife's health wasn't well, we were going through like a very difficult personal time, it was really stressful doing that project, which was really high intensity, you know, just taking a, I think it's at that point $2 billion of revenue, and moving from one system to the other. Nobody stressed. At home, we were also dealing with a lot of personal stuff. So I just decided to take some time off and see what I wanted to do next. And part of that journey, I realized that I'd like to do something for small businesses. So I tried some things, nothing really stuck in I kind of decided to go back, you know, what we were doing with my wife earlier, which was e-commerce. So we started focusing, you know, on building tooling for Etsy, which was smaller sellers, like my wife was one of them. We knew that market, we knew the seller. I think like within a week of watching, we were the number two app or the now dead Etsy App Store. In fact, we still are we shut down the Etsy part of that business. But our which is still running on Etsy is still inside embedded in the dashboard as like one of the five or six recommended apps marketing. We don't focus on that business at all. Somebody just keeps trickling in, we keep helping those merchants. But for the most part, that's that business is dead. And then from there, we pivoted to Shopify, but we were doing organic marketing that thing, because Zalk decided, you got to pay for stuff. Sure. And that was beta play. Yes. And that product started making revenue, we raised money to the founder of WhatsApp for that. But that was around the time, Facebook dial down the organic reach. And I stole that and it was like, you know, we we actually became profitable very quickly. But they, I could see that one year down the road, this is just dead. So we started to look at other things. And in the process, we ended up starting an agency because we knew paid was going to be big. It was already very big by 2070 80. And we started an agency to understand what exactly, you know, does the customer want. And we ran that agency for years till we figured out what our paid off is going to be. And we came up with this catalog, which was basically a lot of the technology for the catalog product was built for organic product. But we figured that it's much better utilized. Enter. And that problem is stuck. Like it resonated with merchants. And till date, that's our bread and butter. That's where we make all our money. We have bands like blue plastic, our garden designs, specialized bikes, a bunch of really big e to 90s brands, now we are getting product to do almost all the advertise the feed advertising.
William Harris 8:43
Yeah, so this was born out of, you know, intimate knowledge of a need, basically, that you were experiencing by running people's paid ads and said, There's got to be a better way. A lot of those catalogs and the emphases that were coming straight out of Shopify, or any of the other ads platforms or sorry, the other shopping platforms were absolutely, they're garbage, they're you know, nothing more than really just kind of like sending through some some basic information. But there was very little customization that you can do to that. And we witnessed that firsthand, as well as an agency trying to, you know, optimize. And that happens to be one of the bigger places that a lot of e-commerce people are going to put budget is towards their catalog ads. And so if there's an opportunity to gain an edge on what you're doing from an advertising perspective there that just makes sense.
Karan Jassar 9:32
Yes, and this became like so when I was at Apple, nobody knew what was going to happen. It was like a eight month Wild Wild West era where companies like dribble we're not being we're born with the analytics that need massive and for a moment there we sent like, Man this is and you know the performance tank overall. If Facebook got their shit together and things started looking better again, sure, but you We ended up being a big beneficiary of that trade. And that that's something that we understand now. And that will be scary, right for the minute. And then COVID happened. I was good didn't happen. It was like a crazy two year period for e-comm. So I think everybody industry knows that. But now it's not.
William Harris 10:20
It's not. We're dealing now with recession and everything else. So yeah, we're, we're just year after year after year of some very interesting stuff in our space. Yeah,
Karan Jassar 10:29
look, now I think I basically changed my perspective. I think that's just the nature of business. Absolutely. It's, it's just the way things change markets change every day. Yep. But what happened is part of the iOS thing was Facebook, Austin godersi. stuff. You know, before I was put in Facebook, nearly exactly, you and I were, each and every website, we've been to each and every architect the visited, he changed the variation, checked out the color size in copy. Now they can work with data card loss due to Apple's quote, unquote, privacy protection, which is very self serving for them. But honestly destroyed a lot of small businesses. I think that was a terrible thing that Apple did. But yeah, but here we are, I think Facebook's done a brilliant job of recapturing that signal. And one of the ways the recapture that signal is by going back to honor. So there is nothing Apple can do to prevent Facebook from capturing anything that happens inside of the Facebook. Right. So if you're on Instagram, and you're interacting with products that are tagged in the sea, that's a very positive signal on Facebook. Right, if you're looking on those products, and going to the Bradshaw, and going to browsing those products, on on Instagram, why they are sharp, that's very powerful signal, Facebook. Likewise, if you're running catalog ads, you know, those carousel ads that people get, the level of engagement, the products that you interact with the products that you click through to go to the website, all of those guys are really powerful Facebook. And my, my understanding, which has been unofficially confirmed by many people at Facebook, some friends, some people on Twitter, is that that's where Facebook, the shop smarter, the feed based products, the tagging of Instagram feed, which is also driven by product, see, they're all way Facebook is capturing all of that signal. And we can see some of those results in the latest quarter that Facebook just had.
William Harris 12:38
Yeah, and so you know, while Facebook lost access to a lot of data, and a lot of ways that are tracking some some things, they also gained access to a lot of data. And I think a lot of people forget this part of this. And I've seen some people hint at this. But prior to this, a lot of people hadn't set up copy right conversions API, then all of a sudden, there was a significant need for them to do that. And so to a point now, Facebook does have knowledge, it has a record of every single purchase that you have made on every single website, like basically right, like let's just say within reason, almost now, it has that now it's coming through in the API, they can't necessarily tie it to which ad was originated that but they know now every single purchase with even more, let's say accuracy, they know the purchases that you're making on other websites, more so than they knew even before they had the pixel which there was some loss on the pixel before?
Karan Jassar 13:34
Yes, yes, it took them some time to figure out exactly how to go about filling those gaps. But you've done a really good job with the basic integration that Shopify has worked happy. I think that's rock solid. But also there are third party apps, though, that are doing a really good job as well. I think hellobar is one of them. And there are others that are not selling that will for Facebook. And I think you'll like absolutely, I think they've taken some signal from here and some signal from here. And they've patched together enough data to kind of give deliver on target ads all over again.
William Harris 14:11
Yeah, and you know, it kind of reminds me of, I've seen, let's just say like some compression algorithms applied to pictures, whether intentionally almost make something look eight bit or something like that. And what's interesting is, it takes like, you know, at least was simplified ones, it'll take like maybe nine Pixels 123123123. And it'll take the average of all of those, and we'll make one pixel from that of the average of all of those colors, and it makes this image that's a blurred version of it, but you can still tell what it is. And I think that's more or less what happened is Facebook's got now enough data from all of these different sources to where they're saying, Look, we can understand within a pretty reasonable you know, amount of certainty that this is what that color of that pixel is supposed to be or this is what originated that that purchase and things like that. And there's so much data there that it gives us a pretty good understanding of what's actually taking place.
Karan Jassar 14:59
Yeah, to your Scotland. And so don't do it, do it like things like advantage plus shop. Right?
William Harris 15:07
This is where I wanted to go here. Yep. Yeah.
Karan Jassar 15:09
So it's a natural segue into that. more easily the founder of native recently asked on Twitter, like how is the advantage of shopping different from let's just say, broad targeting, or how Facebook's ad in its previous iteration. And we had a really good conversation with somebody from Facebook, pretty much before, I was saying, but essentially, what's happening is Facebook's lost some data points, but it's gained some new data points here, that this is returned a different algorithm, right. And part of doing that is they took away some controls from the media buyer. So normally your interest based targeting you layer on demographic data, then they taken away some of those controls. Partially because, you know, if you take one of those eight bits away from that nine bit, you know, matrix that you were talking about in that analogy, you'd have a blurry picture, right earlier, it is like they're 20 bits. And all they needed was to figure out that eight irrelevant base. Now they actually are precisely Ebates, they don't want you to take one of those eight weeks away, because then we impact the algorithm, right? So they figured out that basically, if we have the pixels, we have caffeine grace, and we have the catalog, the shops product, we've got tagging going on all of these things, we can still paint a very reasonable picture of who's there to buy your product.
William Harris 16:33
Yep. And so when I was talking to you about this earlier, we were talking about how like ASE in Facebook early weighing on some of these catalogs and retargeting and daba. But Facebook kind of more or less forcing people to go through their own shop, right product, taking some of that stuff. Walk us through what you're seeing there. Because from from what I understand, from what I remember reading here, Facebook is more or less moving towards a model where you absolutely have to go through essentially Facebook shops here by I think is like mid 2024. You have to
Karan Jassar 17:10
Yes, it yes for this shops product. So you would still have the option of running an app and driving traffic to your site chart. I can easily drive some traffic there shops are it gives them the opportunity to collect more data. And essentially, in the end, all of this is our data, right? What what happened data is impacting what ASE is doing what you're doing with catalogs, what one is doing with TRP, this is all just filling the holes in that data warehouse in our phones, they need to fail to be able to predict accurately who's going to buy your product. Yeah, forcing people in their shops to go to the checkout on Facebook gives them a lot more. Right. And I think, sure, I think they've done enough to kind of get the performance back where people will continue to spend more of their budget on Facebook. But I think it's probably opened more options in their mind that hey, what if he did this one more thing? I think we'll probably be able to collect this much more data, then probably further optimize the ad delivery. Like the reason Facebook walked like a machine before I was good. It was like they had so much data privacy concerns. Like if even if you didn't set out their hands, they probably still know exactly if it was money. Sure. Right now they have to, but heart. And my feeling is that eventually, it's about like, way back in 2000 Facebook, or Mark Zuckerberg was very keen to buy the phone. And they decided that's not a good idea because he was worried about the exact same thing, either Apple or Google having a whole chokehold around Facebook's neck. Because, you know, these are the two dominant mobile platforms. Yep. And that's exactly how it played out with iOS 18 thing. And now obviously, they don't do much about it in writing. A lot of Facebook's VR obsession comes from the fact that we want to own some platform, right? We don't want to be dependent on either Apple or Google. But I think a part of that same thinking is, can we win the check, right? We want to have as a DA, and as much activity in platform so we can get more data can still optimize ads, because that's really where all the revenue comes from the needle, given all the other two undertakes that Facebook does money still comes from ads. Right that that is the golden goose. Yep, like so. And I
William Harris 19:40
go, I'm sorry, go ahead.
Karan Jassar 19:42
No, no, I was just like, all they're trying to do is move everything either to remove Apple interference, and Google's interference into you know what's coming down the road.
William Harris 19:52
Yep. And I have seen them invest heavily into learning in iterating. In their shops to be able to control the checkout experience to your point where, you know, they're giving brands $100,000 or more a month to basically send ads to their shops experience in basically for the sole purpose of almost learning as much as they possibly can as quickly as they can. So they can iterate that product and make it better. And they have, they've made it considerably better over the last couple of years, when it first came out. You know, I was very disappointed, I thought it was horrible, in that the experience was terrible, people didn't like using it, I'd say that has come a long way. But in here's one of the issues that I that I want to get into, because you've talked about cash versus accrual based attribution. And here's where I want to tie this into this. One of the things that I've seen with shops still that I'm hesitant on going all in on other than the fact that you have no choice on it two point this is coming 2024, you pretty much have to be all in on it is that I don't think that it dries as good of a customer or as good of a customer relationship yet. And so when we're looking at this from a cohort perspective, we do not see the same repeat purchase rate, the same lifetime value, even the same ale vs oftentimes from people that are coming through Facebook shops, as opposed to getting them to your website. And so when we're talking about, you know, cash versus accrual based, almost accounting, or just let's just say understanding where the most profitable purchases, a purchase from somebody on shops, I don't want to pay the same CAC for somebody that I acquired on shops as I do for somebody that acquired on my actual website. What do you think about this, that you're
Karan Jassar 21:43
absolutely spot on? Part of the problem is Facebook again, back to data, like they are fronting like, you know, 20% discounts, you know, they are actually taking that cost themselves? Sure. So you're selling something on Facebook, they just discount it by 20%, or basically give the buyer a 20% coupon. And that's not necessarily the buyer, you want the right I mean, you want people to buy your product for what it is, it's a high quality product, it meets their needs, you know, aesthetically pleasing, but when you start attracting purely discount shoppers, you will get lower you will be poor retention will reap rebirth, like poor lifetime value and repurchase rate, all of those things will happen. If you just acquired customers using discounts, they'll have the same thing will happen on your website as well. Right? Sure. And that is why like a decent brand is generally either you're a discount brand. Or if you are not a discount brand, you can't constantly be giving out coupons, right? Because that's just attract the roll kind of bucket. Currently, Facebook is doing it to kind of meet their needs. And sure, and we'll get some sales, but that, but there are two things that you use as the brand new one is, of course, the quality of customer is not the same, you're 100% Spot on that. The second thing also is you also lose that customer journey. Like you'll get the sale. So we provide a free first party pixel for Shopify merchants, right. And one of the most requested features we're working on is giving more analytics about shop purchase. Yeah, you'll get the data like okay, William bought something from my store for $100. But what you don't get is like a How many times do you click on my ad? Which ad? Did you click on? Like, what was the UTM? What was the campaign because the creative, all that information is lost. So whereas Facebook's picking that information up for their targeting, we as like people working with the brands or either on the agency side, or even in house marketers, we use a lot in terms of like the journey that we're used to see. The second thing, of course, is Facebook pushing these discount codes, which is good. If they're learning something about how to serve my ads better, I'm okay. But at the same time it is attracting that. Not that ideal customer that she'd want. Right. So it's it's, I mean, it's a double edged sword. You want Facebook to have more data to serve your ads better. But it'll come at a cost of, you know, either you losing data or getting not the ideal customer profile, or both, in many cases,
William Harris 24:21
yeah, it becomes unnecessary. And I would say, Yeah,
Karan Jassar 24:25
I agree. I think we'll have to deal with it. There is no choice. Yeah, yeah.
William Harris 24:28
Yeah, I don't think you're gonna have a choice. The same way that I think, you know, Google is pushing towards a very similar model and a lot of the things that they're doing as well, where they're limiting your choices of what you're able to do. And so that's something that you know, you're either gonna play in their sandbox or you're not, it's no different than what Amazon does as well, to the point where it's like, Look, if you want to play an Amazon sandbox, this is what you have to do. And so I think that we're seeing right now a fragmentation of commerce, which is interesting. And I'm trying to look at this from the future perspective of, you know, we've got omni channel and it's looking, let's just say very fragmented, and being able to understand all of the different things that are going on across all the different channels, it's becoming less of an omni channel. And it's just like, you know, every man for himself channel, you're just having to play in all the different sandboxes and play the best within each one of those sandboxes that you can.
Karan Jassar 25:19
Yes. And that ties back again to attribution, right? You as a brand owner or as a performance marketer, you want to understand which dollar is most effective. Right. And that then will tie it back to cash versus accrual at the minute, because I think that's what we started. Yes. But what anything, you've talked about this before? I, at least, I think I think you have, but correct me if I'm wrong. Like when you run a Facebook campaign, you will see an update on brand search. I think that is something you talked about it.
William Harris 25:50
And article is the specific article is The ROAS Death Spiral. And I show charts and graphs of that outline that very well, if they want to go check that yes,
Karan Jassar 26:02
that's an article, it's a very well written article, I think I would recommend everybody, maybe we should be naked in the show notes. That's a great article makes that very simple point that ROAS is hard to measure, right? Whether you have a first party, so while zero party pixels, or a second, or a third party pixel or any platform doesn't matter. It is just a murky picture. It always will be. And it always was. It's just that in the past, because Facebook was such an effective platform, people just cared less about us, because surely Facebook people they for I'm sure it's at least he went off. And that was good enough. Sure. Right, as long as it wasn't eating a certain level of profitability. People were okay with it. And now that's become a little bit harder to track. So there's this group of tools that's trying to sell you a will help you understand your attribution. And yes, it does help having like a first party pixel, it starts giving you more and more data. But it's never going to be the complete picture. Like you'll never be able to clearly tie your Facebook ad to your update in brand search, you know, by an order lead. Right? Right, you will probably do some statistical modeling around it. And you can probably reasonably estimate, a I got like 13%, right? That's all very, that's all very, you know, if you're not an IT, you're really savvy when it comes to data adding you're one of the more informed people. But for a lot of performance marketers, that's actually not their strength. And even if brand marketers in house, they struggle with that stuff, right. And there are tools to help you. But really, who knows. And that's where the whole cash versus accrual thing comes into picture. One of the advantages of accrual based accounting for your advertising is that your, your base like let's say you want to do one day click base requests for Facebook. And I was actually looking at an account right before a call there one day click in for people who are not familiar with cash versus accrual, I just spend a minute kind of explaining what we're talking about. Cash based accounting is, let's say you shown the ad from the first to the fifth of June, and the day we're recording. And let's say you showed an ad on first of June. And that guy came back today directly to your website and purchase on the fifth of June. They clicked on a Facebook ad on the first of June, the impression was on the first the click was on the first, but they came back on the sixth of June and purchased that revenue will be attributed to fifth of June for the same campaign. That's cash based accounting. You're attributing revenue on the day you collected the cash. accrual based accounting will say that, well, this ad impression and click was on the first of June. So I'm going to attribute this revenue back to first of June. So you've got no revenue from this fifth of June when the actual sale happened. But that revenue goes back on the first of June. And that sounds confusing. Like why would you do that? Right? Right. And the whole idea is that you could use that information to predict, you know, what is the delayed impact of your advertising. So let's say on the first of June, you had one sale, and then on the system view, you had another sale from the same set of impressions, or the same spend you had on the first five days. So you can say it with enough data if you're collecting data at scale. Okay. On my one day click basis, my attribution window is just one day on an accrual basis on the first of June by reverse was 1x. So on the very same day, I had the impression I got a click and that reverse is 1x. That's easy to attribute right because you got that click the same day you can confidently say is that like I have the dogs I can show you these other youtubers there's the cake there's the say, but that reverse is what makes you and then you can also statistically model what's happening in the next seven days, 14 days, 28 days, and you can with reasonably with reasonable accuracy predict like that 1x us is gonna go up by 15% or 20% or 30% Over the next X days. Mm. And why is that valuable? Like it, it actually becomes very valuable when you have multiple channels. So if you're also doing Amazon, if you're also doing Google aggressively, you're also doing Facebook, you're also doing tick tock. The one day is very accurate. And if you have that data that is accurate, the prediction is a lot more accurate. Yep. Whereas in on a cash basis, you're getting revenue today from an impression that was like three months ago, six months ago, three weeks ago, is really hard. So is your current campaign doing really well is? Or is it something that you shut down? Because he didn't fully understand the delayed attribution? That guy doing well, right. So so that's where like, if my guideline, you know, because we do this for free for our users, my guideline is typically if Facebook is your dominant channel of acquisition, anybody can reasonably small standard Google or even if it's just Facebook and Google, you're probably only doing cash, like butter. If you have multiple channels, you should focus more on accrual. Or if your conversion cycle is really not like if you're selling, you know, furniture, it couldn't last 30 to 45 days before you actually see the sale. And when you see the sale, you think like oh man, my my campaign that I ran today got click on that customer convert it. But really, the sale was done by the ad group, like 45 days, right? That's the ad that got them on your website, introduced a brand, something in that messaging resonated with them, so they thought I should click. So that's been a crew helps a lot.
William Harris 31:58
And so I used to talk about this, like bowling is the is the illustration that I used to give. In bowling, you get a strike, it's like, well, for that frame, you get 10 points for that, you know, well, it's not all you get for that, because you got to see and you don't, they can't even put the score that you got for that frame in there yet. Because you got to see what you get for the next two balls. And if if the next time you roll, you get another strike, they still can't give you the score that you got for that first frame. Because you got to look at the next one, you got to go all the way to get the third frame now then you build another strike. Now it's going to take those 20 points, it's adding it to that you got 30 points. Now for that first one, there is a delete attribution that takes place. And I think that a lot of times maybe we forgotten about the late attribution, simply because the way the attribution is being done within platform has changed significantly. And so there's already an inherent delay that takes place within attribution anyways, to a point because Facebook can hold these, or Apple can hold a lot of these for up to three days before they even send over these transactions. And so there's already an inherent delay in there. And so but to your point, you know, to get something that's a considered purchase, the delay is very much likely longer than whatever you're seeing there within platform and a lot of it then ends up just not even getting the opportunity to be attributed in platform at all. And so what we like to do is we like to use holdout tests as a way to just let's level set this data if you spend $1,000. What does the actual impact of that $1,000 have on your bank account, right? And so, let's just say that you're starting a new channel, you're doing Tik Tok, and you haven't done ticked up before. And you advertise in these five states. And you don't advertise in these 45 states. If you spend money on these five states, you spend $10,000, how much more money do you get in those five states? The you know, versus the ones that didn't receive the as, what's the difference there? What's the what's the percentage change growth that happens? And now you can level set that and you say, Okay, real hard money, we spend $1, we get $5. Maybe I look in the platform, I look in Tik Tok, I look in Facebook, and it says a two to one. Okay, well, I know now that a two to one, essentially in platform correlates to a five to one in actual real money. Now like is to operate around this two to one model, like you said, but I can know in my head that if I was using like an accrual based accounting, that two to one is really probably closer to five to one that's just not going to all show up in platform because of the delay and attribution. And it's going to be so delayed that it just doesn't even show up at all.
Karan Jassar 34:28
Yes, yes, absolutely. Facebook currently is using a seven day window actually they started reporting now again on 28 day window, but you know, there are many factors that complicate this entire situation. One of them is you know, Apple holding out events are much stronger. Second is each brand has like a different conversion Psych. Third is that it is just not possible to track everything. Not at least from like, like even as a first party pixel. There are limitations like people claiming that they are modeling views. view based conversions. But the reality is first started pixels don't have any view date, especially from Facebook. So you really have to take into account what the platform is, say at least a part of that. And also look at some hard data that you have. Also do some stuff on your end, like what you said about all our tests. That's like, that's a very high level of data accuracy. It is very, very hard for most agencies to do that. Right? It's very few agencies who do and, and to be fair to the smaller agencies, it's also not viable for smaller.
William Harris 35:37
And it's not necessarily for everyone. But if you've started to get that level accuracy, it's it's important to understand the difference between that two to one that you maybe think you can't scale on is actually five to one that can open up an entire world of scaling possibilities.
Karan Jassar 35:51
Absolutely, yes, absolutely. And I'm so glad you brought up, you know, this topic, because it's just one of those things that nobody will ever agree to handle on Twitter, or otherwise. Right. But what you're saying is absolutely spot on. Like you have to spend the energy to understand what your dollar is bringing, right and sometimes in platform can go either ways, like sometimes they go overboard sometimes under report, my opinion is is exactly what you said in your example, for the most part, the platform was not able to capture the impact it is having on revenue, right? So you have to both hold out tests, you have to go to certain advanced tools to figure out exactly where that money's going.
William Harris 36:33
Yep. And I think it depends on what type of campaign it is, if you're going after a prospecting for net new people, that one's usually going to be under attributed in platform because that that conversion might happen a few more days out. Whereas if you're doing a retargeting campaign, that one might have a higher than necessary reported attribution, because it took credit for something that was already going to take place anyways. And so it depends on the campaign that you're looking at, and how to evaluate each one of those campaigns and who you're going after, to make the most sense of this. Which one of the things we were talking about before too is how long should you run a test? So let's say that you are running tests like this, and you're looking at this? How long should you run a test to know whether or not you've gotten good enough results to say that you feel confident in what you're seeing between this ad working better? Or this ad working better?
Karan Jassar 37:22
Yeah, that's a really difficult question to answer. Yes. You know, simply because, well, okay. I mean, there comes a point where you have enough data to make that decision. So one of the things that we are working on, and I think you and I were talking about this on Twitter, maybe like six months ago, is how long before you shuttered? Anything Taylor Holliday was of the opinion never. Right. And you and I were single, at some point, you've got to cut the ad, is it's not performing. But But that's a very subjective timeframe, like how long is it going to be? So I generally think in terms of multiples of a movie, right. So if you is 100 bucks, I would typically at least spend like seven to 10x A movie before cutting that. Now, it really depends on the brand also like, like, we serve a lot of smaller brands as well, because we're an app. So sometimes, they just don't have the capacity to go that far. And sometimes they'll shut shut and good at you that could have done better, which is Facebook needed more time. But in most cases, if an ad has not given you, like on an accrual basis, or a one day click basis, if an ad is underperforming, as compared to your other ads, I think you should shut it anytime. I mean, as early as three to 5x up but no longer than 10x. To me. At least that's how I do it. But I'd also love to hear your thoughts on the topic.
William Harris 38:53
I well I like I like your your methodology there of looking at like, you know, multiple of ARV because to a point, you almost need enough data to know whether or not there's something there. And if you're only acquiring, you know, somebody for you know, it's $100 ARV and you're only getting a couple here or there that's it's not enough to know whether or not what you're doing has made a difference. One of the other things that I like to look at with this is looking at the LTV of who's coming in as well to understand and so I like looking at this from like a cohort perspective where you'd say, well, maybe this ad is acquiring people a little bit better. Let's say that ad a is acquiring people for $20 for ad B is acquiring people for $25. So your CAC is better on ad a? Well what happens if you look at that three months down the road, six months down the road, which one of those is actually acquiring the better customer and you might find that the one that acquired people for less money was not acquiring the better customer. And so the $25 CAC was actually better for you in the long run over the course of six months or 12 months. So when I talk about running a test, a lot of what I talk about is you can run it for your initial leading metrics. And let's just say that you can run it as short as just looking to see is it doing a better job of getting clicks to your website and things like that you can look at those leading metrics to say, is it worth at least continuing to spend money on for the next foreseeable future? Then how's it doing from an actual acquisition performance standpoint? If it's acquiring people, at least at a relatively good amount, then you say, Okay, well, let's let this potentially run a little bit longer and see what happens on it from an LTV perspective. Where I would say shutting an ad off, though comes from and when you say like, this is a clear case, or at least a case worth trying to shut this ad off is looking at some of the Google Analytics data. And so if you can look at some of your analytics data, and you would say that you can see that maybe add a says that it's getting a whole lot more than purchases. But ad B, when you look at it in Google Analytics, all of those are actually coming through on ad B and not ad a, then you have to wonder, did facebook get its attribution incorrect, because remember, it's not a one to one attribution anymore. It's a model of attribution. And so you have to say, maybe they didn't quite get it. Whereas on Google Analytics, it was based on the actual click through data, it at least gives you pause to say, what happens if let me go ahead and shut off one of these ads here, the one that maybe isn't getting the actual direct click through and seeing if I actually get more revenue overall now as a result of this, or less, or how that impacts things?
Karan Jassar 41:32
Absolutely. So a few few, I have a few thoughts on that. Number one, I think like so, as I was talking about our free first party pixel, this is one of the features that I will really want to build down the road. Whereas we can go back and look at cohorts of customers acquired through specific acts and specific campaigns. I think that is like a killer feature. That is absolutely the best way to look at, you know, which campaign which ad did well, the slight issue, though, is that you really can't make real time decisions, you might have to wait a few months. But it is such a powerful piece of analysis that every branch should do. Simply not not just for the app, you're also like, what was the messaging on the app that did? Well, I think the I mean, there are two ways to look at ROAS as customer acquisition cost and lifetime value, what is like my day to day management? Hey, should I shut this ad, you know, and in Google Analytics are different data is getting more click through rate. But Facebook's reporting like this ads, driving revenue, you know, those, all those tactical things? So I'm going to talk about how I do that. Sure, but the URL is also what you said, Hey, can I go back six months later or three months later, and understand what ad worked, and why but that I really need what messaging work, it's not really the model or, you know, the hope that no, maybe the hook too, but essentially, hey, when we were talking about, you know, durability, that added like quarters, the better class of customer, then maybe put a disco or when this influence upon does or we use this, you know, influencer driven ad, their audience actually had a much better LTV than let's say, another act that a different message. So stuff like that. And that's the kind of analysis that you know, you have to have a long term view to, for your, if you have a long term view of your ad account, you go back and look at these things, you will improve significant, right? They don't be, you know, giant leaps once you understand the messaging, not like today, like what's my iOS today, but versus what's my LTV from that core? Yeah, in the short run, you know, I personally find Google analytics information and little bit limiting in your GFR is a whole different beast altogether. But in and that's why we did the first party pixel and in the years of free simply because we want more people to have reliable data. But we have this thing called a big advice. And what it does is the number one thing we'll do is we'll break out new versus repeat customers. So to your point, a you know, Facebook will over report on the middle or bottom of the funnel, you know, because that customer may have any late bought a but a new customer, you know, that's more valuable. So that's one way of looking at an ad performance like okay, we're getting more new buyers. So maybe even if the deck is higher, it's worth it. Right. Yeah. Yeah, that's that's like one of the really important things and then click through rate matters. You know, the model traffic, the CPM, CPC, those things are I mean, if they're terrible out of the gate, that's a bad sign, but for the most part I find like The decision comes down to either ROAS or the new versus repeat customers that, like some ads will do fine. But as you look at the numerous repeats, that is like 70%, repeat buyers, right, by so that's another data point, at least in the short term decision making that I find extremely bad. And that's harder to do with Google Analytics. And we get that for free for all Shopify stores. So hopefully, somebody will listen to it and try it out.
William Harris 45:28
Yeah, I love that. So and I want to transition in a little bit into some of the personal stuff of of who is Karan Jassar. But before I do, what would you say, are your top tips for doing ads? Well, then so given everything that we've just said, if you are going to say like, Hey, these are the top two or three things that you should be thinking about as an advertiser in e-commerce space, in today's environment, what would those be?
Karan Jassar 45:56
So we're talking about specifically brands, like north of 10 million already, so they figured out a lot of their basics, and they're talking about, right? Yeah, yep. I think the stuff that you mentioned is really, I think, the understanding your cohorts and your lifetime value. And then even going back and seeing, okay, why did that ad on that campaign, or that channel had higher LTV, you know, higher up, or it will be for that matter, I think that's just really important. Then overall understanding of your, where your dollar is coming, and not really just, you know, looking at in that form and saying, oh, Facebook's wrong, or I need this to then that two tones are fine, we all need tools to kind of make our job more efficient. Otherwise, we just spend more time in spreadsheets. And that's the value of it too. But like, for the most part, you need to have a mental model of C and D, like, right? Ear, you shouldn't lose your shirt advertising, you need to know like when to put the pull the brakes. But I find like the other problem is more acute than then that I'm going to make a loss. If you're doing like 10 million plus a year, most likely you're not spending in or scaling in, in the right places versus cutting back. Right. And then you know, of course, contribution margin is really important. I think that's your True North Star. As long as you have any contribution margin, the problem comes down to which channel which, which which created which campaign, what messaging, what channels are more effective than others? I think that is the part that brands in US marketers, performance, marketers working in agencies need to spend more time, you know, just getting better, was easier. It's a lot harder now.
William Harris 47:45
Well, I mean, your point, I think that, you know, once you get to the 10 million mark, you've done a whole lot of things to get there. Yeah. And when you say, Well, what specifically is working, that's when it becomes harder to say, I don't know specifically, what's working. I know that everything that we did has worked to get us here. But if you're saying well, what do you need to lean into more next, that's where that attribution becomes harder to discern in and harder to make that next move into your point they're under investing in partly because they're not quite sure where to invest next.
Karan Jassar 48:15
Yeah, a lot of bends, they find that they have this fear that's holding them back that if I just go all in or too aggressively, like it's going to be terrible outcome. And I think a simple fix for that is looking at your contribution margin, that should not be in the red. But as long as your contribution margin is green, just go for it, you know, yeah. And see what happens. Take it out where you need to, you know, put the gas
William Harris 48:43
in, sometimes the best way to test it out is to just go ahead and do it, and see if it has the effect that you want, right to a point. Okay, who is Karan Jassar? One of the questions I wanted to ask you is, why are you successful? What is it about you? Is there something in your childhood or something that pushed you to be who you are to say, I'm going to be able to start this company and run this company, amid all of the craziness that's going on in the environment, I'm going to be able to persevere through this.
Karan Jassar 49:14
I think there are a couple of things in life that shaped us. For me, I think there were two things that shaped me. One was like my, my early education was somewhere in the interior of India, where my English, my exposure to English was negligible. I was very good with math and science. And then around sixth grade my family goals simply because they felt my education is suffering because even the schooling wasn't very good in rural India. And we had to be there because of my dad's job. But my mom and I returns to the city, like a bigger city and I got into a very nice school. But the problem was, the medium of education was English, and I didn't understand it. Not At least to the level where other students were. So I had a tennis self learn from an early age, you know, just the teacher was teaching in a different language I had, and my parents were, I mean, they did everything they could to support me, but they didn't fully recognize the fact that, you know, that's what's happening in school needed at the school. So that made me a little bit more independent, like, Okay, you figure it out, you shouldn't go to the Hey. And then then the second thing that happened was, you know, I met someone in grad school. So you know, so I went from, like, a school where I didn't understand the medium of education, right, and to kind of self learn to get into my masters where I was completely on scholarship, I didn't have to pay a penny, I got there. And I came to the US didn't have to pay anything at USC, I was working in AI, much before, like in 2001, and two, so 2003 to 2004. But and then someone there and we got married. And then that. I mean, she had met my wife, my current wife, were still married. She had a long, two decade journey of a very complex illness that included depression, anxiety, fibromyalgia, chronic fatigue syndrome. So that journey shaped me more than anything else in the in the last two decades. I think just being there for her. I mean, I had a very happy childhood. I had no idea like, like, for a lot of people who did not dealt with depression, were like, Hey, man, come on JIRA. But that's just not what it is. Right? It's not that simple. But, you know, I remember this one incident where this but I think 2010 We were at, we were under the care of Stanford Hospital, they put us on opioids without fully educating us what this thing is. So my wife, very bright academically, we've never really done anything wrong in the sense we'd never done drugs or illegal drugs before. When no idea what what opioids were, right? They, I mean, as, as most people aren't, you go to your doctor, you expect the doctor to guide you, you follow their guidance, because you know, they are the subject matter expert. They put us on opioids without fully explaining to us what it meant because my wife had very bad body pain from fibromyalgia. Sure. So that complicated her illness in our stomach lining gave up your gallbladder lever, everything was suffering. She and her depression became a lot worse because she couldn't keep food down. It's just, it was like, we were spiraling downwards. And I remember this one instance, where the doctor, we went to the emergency room at Stanford, and it was like, you know, this is my wife. We have a great life. We love each other, but she wants to take her life. And the doctor said, go to the emergency. Wow. It was weird, because everything in life on one paper was perfect. Like I was making so much money, doing very well at my job. Everybody was recognizing, you know how well I was doing at my job. We loved each other. And yet, it was so difficult, like she had depression, anxiety, fibromyalgia, she was an opioids. Almost everything in in her hand was complicated and made worse by Stanford. They and then we add to it took us like 10 years to dig ourselves out of that whole lot of therapy, a lot of alternative medicine. But that made me very resilient. I mean, I know firsthand, understood human emotions, which now makes me a better pounder, a better leader. We've been able to help our team through really difficult times, like somebody's having difficulty conceiving, miscarriages, interpersonal relationships, problem with parents, but even like as a dad, I think both patient better dad now because of that journey. But that's one thing that I feel like we don't talk about enough still, as a society, right? I mean, there is still stigma attached depression. There's still a stigma attached to feeling anxious. Yeah, I think that's the one thing that if I have an opportunity, I always like to talk about.
William Harris 54:33
I can't thank you enough for being that open and vulnerable and sharing that. I think that one of the things that I appreciate is congratulations to you and your wife for pushing through that like, congrats to your wife like for pushing through that that is a horrible feeling and for you for sticking around. It's very hard to be around somebody who's depressed and nothing you do, can can cheer them up and that's very difficult and I just I I applaud you in both of you for being able to stand up to that. And I think what you said about depression, that is really, you ordered it so well, which is on paper, it didn't make sense. I think that's the thing that I think people should understand about depression. I'm not a psychologist, but from my understanding, I do have a background in in nursing, as well. And from my understanding, from what I've seen in people's lives and friends who have taken their Lives, unfortunately, is that that is classic of depression is that it doesn't make sense that almost you know that this, this is depression, because if you're sad, because somebody just died in your family, that makes sense. It's not necessarily depression, that's just a natural response to exactly what's taking place this external stimuli. Depression for oftentimes, though, is when you are not even necessarily sad, but just unmotivated, or whatever it might be. Despite everything that's going on, that might be in a good spot. And so good for you guys for recognizing that and pushing through that.
Karan Jassar 56:06
Yeah, I think it was hard, we'll do get wet. I think that got his overtime. And even like, you know, shout out to anybody else who's dealing with a family member going through depression. I think it's hard on the caretaker as well. But yes, it's important to remember that that other person, the person who's going through depression, they probably their demons failed to understand why do you care about like, for my wife, it was like, why are you still around? Like, I mean, I am no good I ever was that at that point, there's, they're just the reality. So distorted. I don't even see why anybody should love them. Which was the hardest part for me, because I love my wife theory, and it was sad to see your, you know, being so down on herself. But I'm happy, we're past that. A lot of that is just her will to get better. It's just one of those things that I think as a society, we need to do a better job.
William Harris 57:16
I agree, and being able to have conversations just like this, to just call it out and say that, hey, you might be struggling with this. If you are, you know, don't be afraid to get help. Whether this is you or someone you love. Help them through this, stick with them, even when they maybe are rejecting your help, because that's probably when they need the most help at that moment as well. And, boy, it's hard to transition from that. But I'd say similarly, we talked about what it's like, well, how do you specifically deal with the stresses of being an entrepreneur? You know it because because that that can also be something that sends people into this? Are you do you have a strong routine or something? Or what is it that allows you to stay mentally strong, during very stressful situations.
Karan Jassar 58:06
So one of the silver linings of this episode, this journey that I went through with my wife was I sat through a lot of therapy with her. And we get this specific form of therapy that works for us. It's called Cognitive Behavioral Therapy, or CBT is, which is a board director. But specifically this, there's this book called Feeling Good. And they have this way of expressing your own emotions, and talking back to your negative thoughts. So it's called the daily mood. The book is by Dr. David Burns, and we were lucky enough to work with one of his star students, Dr. Matthew Bay. So we I, I still do that. I mean, I do that for my own mental health now, not enough and not I mean, that's one, one thing and my wife tells me now the exact year we've been through this journey, you got to take better mental care of your own mental health. So do those roadblocks. It's a tool that is invaluable to me that you basically what you end up doing is you start by writing down everything that's negative, every negative feeling and emotion that you have. And then you talk back to that, that thought in a very methodical way. It is really hard to dig into your negative thoughts and put them on paper because that's the whole point of those negative thoughts is you don't want to face you don't want to see like a man I'm a bad entrepreneur. I'm a bad leader. I'm a bad sure you know, he's not smart enough to build a good company. There are various negative thoughts that all of us have, you lose a client you feed out, you know, something a deal doesn't happen. You see. There is this tendency that you know we pick up somewhere around sometimes while we are growing up to blame us, belittle us stupid or talk down to us In a way, we will never talk down to another person. Yeah, and just, you know, get those, the we're so harsh on ourselves, it's, it's amazing. But if you can find the will to write those thoughts down, like people talk about writing gratitude, I think that's a beautiful way of feeling better. But if you can find the will and courage to write down how you feel about yourself on paper, I think that's like 90% of the journey, look at feeling better about yourself, then the last 10% is really just defeating those negative thoughts, just logically, the rest of it is purely knowledge. Right, I think I think I learned a few tunes during this journey with my wife that I use the way that and this is one.
William Harris 1:00:48
I think those are great tools. It reminds me very much of let's just even say, retraining the algorithm, even within an ads campaign, right? Like, like to bring it to that. But it's, it's that idea where it's like, hey, something's not it's got into this weird feedback loop. And you almost have to start feeding it the right data again, you're saying, Wait, that's the wrong data? What's the right data? Let's put that in there. And it reminds me of, you know, regardless of what somebody might might believe, and I'm not imposing my beliefs on you, or anybody here, but there's a Bible verse that I like, Proverbs 1624, it says, Pleasant words are like a honeycomb, sweet and delightful to the soul and healing to the body. And I think that maybe we don't understand just how healing Pleasant words are to our body, like actually physically healing to our body. And to your point where you said, where it's like, we oftentimes would never say the things to somebody else that we say to ourselves, we have to recognize it. Like we have to maybe treat ourselves very well as well as speak pleasantly towards ourselves as well, because it is health truly for our body.
Karan Jassar 1:01:55
Absolutely, you're absolutely spot within and the thing that you said about the algorithm, it is exactly that, like I could beat myself up by saying, Hey, man, you you're not going to be a great hospital error. But then you have to talk back and feed some good data, a yellow the proof points done, this does this. And slowly and steadily, your mind will get back on that positive track that you need to have to go to chi sub something value.
William Harris 1:02:22
Yeah, Karan this has been absolutely amazing. I want to wrap it up with two quick things. One, is just a quick, fun, silly game that I like to play at the end, which is just called What's that mean? And I give you an absolutely random English word that you shouldn't know the definition of, but you make up a definition that you think sounds believable for all of us. Okay, so the word is hallux H-a-l-l-u-x? What is the definition of Hallux? That we should believe?
Karan Jassar 1:02:53
Alex All right. Well, it's it's a support structure under a building that all set to a hillside. So, you know, if your house is on a steep hillside, what you need to do is put in some hallux is to kind of make sure your foundation is still
William Harris 1:03:10
That's brilliant. I love I mean, it sounds really believable I but I just I love whenever I play this game with people when they come up with things that I never would have thought of like that that first go. If you want to know what the definition of Alex is for real in unless you believe that your Karan is right. If you want to look it up, look it up. H A L L u x. Also Karan, if people wanted to work with you or follow you connect with you, what's the best way for them to get in touch with you?
Karan Jassar 1:03:38
Well, my email is karan@sociohcom. I'm available on Twitter. Generally, it's the easiest way to get a hold of me. It's Karan Jassar, K-a-r-a-n J-a-s-s-a-r. Twitter's the easiest. I enjoy meeting people like yourself. I've met a lot of amazing people on Twitter and would love to connect with more.
William Harris 1:04:01
Cool. Well, Karan, thank you very much for coming out and sharing your knowledge and your time with us.
Karan Jassar 1:04:06
Thanks, William for having I enjoyed it.
William Harris 1:04:10
Yep, thank you, everyone else, have a great day.
Outro 1:04:14
Thanks for listening to the Up Arrow Podcast with William Harris. We'll see you again next time and be sure to click Subscribe to get future episodes.