Unlock your next $100Million+ in e-commerce revenue: The creative testing formula for growing paid social in 2024

Episode transcript:

Gabrielle Stafford:
Hey, guys. Welcome to our SuperSummit session on how to unlock ecomm growth through creative testing. I'm Gabby from Supermetrics and with me I have the wonderful Shabazz from VentureBeyond. We recorded this session a couple of weeks ago on a very rainy Thursday in Birmingham and that enables Shabazz and I actually to be available online through chat with you right now. So as we go through this, if anything occurs to you or you have any questions to ask, reach out in chat and Shabazz and I'll be able to have a chat with you guys, which will be awesome. So Shabazz, you are co-founder and CEO of VentureBeyond. Venturebeyond are one of our Supermetrics customers. And you guys came to our notice because of the amount of queries you are running on a monthly basis. Tell me about yourself. Tell me about VentureBeyond. And where are all these queries coming from?

Shahbaz Khokhar:
Yeah, I think you messaged us saying we're pulling 750 million rows of data, which is-

Gabrielle Stafford:
Yes, it's a lot.

Shahbaz Khokhar:
... insanely big number. Yeah, so I'm CEO of VentureBeyond. We do growth marketing. So basically we grow e-commerce brands direct to consume e-commerce brands. We use data to basically inform our creative testing. We produce creative, we do the media buying, we do Google Ads, Facebook ads, TikTok email, et cetera, et cetera. So we're kind of like full service based around whatever our clients need. The main thing is we use data to actually drive every single decision that we make

Gabrielle Stafford:
Using Supermetrics obviously.

Shahbaz Khokhar:
Of course.

Gabrielle Stafford:
So you guys specialize in DTC and ecomm. You're only four years old. You've grown hugely in the last four years with DTC. I think e-marketer says DTC is going to be 160 billion this year. Tell us about direct-to-consumer and why it's so different and why that makes what you guys do unique right now.

Shahbaz Khokhar:
Yeah, I think the key difference between, I'd say direct-to-consumer and brand marketers is brand marketers are not looking for an immediate conversion. It's more like a long-term play. It's more like what do people feel and think about the brand a year, two years down the line from now. With direct response, we get paid when we make sales. So basically we have to drive revenue immediately, like now today. It's we've got to hit targets every single month, every single week, every single day. So there's a whole different dynamic into how we operate. Overall, I'd say the speed at which we operate has to be super quick. So what I've noticed from looking at people who do brand and people who do performance, generally speaking, people who do performance are much more speed-orientated.

Gabrielle Stafford:
How do you get that speed? Because I know from being a marketer for a very long time myself, creative usually takes time, there are approval processes, but DTC is direct response. So it's what's happening now and reacting in real time.

Shahbaz Khokhar:
It's kind of getting everyone aligned to a certain mindset, including the brand if it's a client or internally, if it's just a team. You've got to get aligned to this whole kind of idea of speed is really important. You cannot wait three months to get a shoot done. You cannot wait for X rounds of approvals, second round of approvals, re-shooting, et cetera, et cetera, et cetera. By that time, guess what? The competition has taken over and you've lost your potential edge. And also I tend to find the kind of content that people shoot for brand campaigns doesn't always resonate to what you need for performance. You want things that are a bit more product-focused, things that really call out features, benefits, pain points of the customer, and a lot of the branded content that people produce isn't really going to get you the cuts that you need to produce those ads that actually get you those direct and immediate sales.

Gabrielle Stafford:
I think you told me when we were having a chat earlier that for one of your clients you're actually producing a thousand unique pieces of creative a month. That's a huge amount. How are you making that work and why is that scale required to drive the growth?

Shahbaz Khokhar:
Yeah, I mean generally, I mean I'll cut to a graph and we can have a look at it. So as you can see, this is a particular brand we did a study on. This is one that was spending 4 million a month at its peak. And you can see the number of ads launched tracks very closely to the amount that they made in revenue. The whole principle here is, and you'll see in the second chart, there's a quite strong correlation between number of ads launched and revenue that you make.
And the whole thing is really is you're trying to produce as much creative as you can, as quick as you can to get the insights back as to what is working to then go and make more stuff. If you are waiting three months for a shoot to come back and then you're launching a campaign and hoping for the best and then waiting another three months, the speed is just not anywhere near fast enough. And that's why we end up producing so much creative. And on average the success rate in our industry and just our success rate is typically between 5, maybe 8% of creators actually spend any money and do any revenue. So to get ones that win and overall have a strong growth, you've got to have a lot of testing.

Gabrielle Stafford:
And what are the mistakes people are making right now?

Shahbaz Khokhar:
Taking too long an approval, being too precious about using a model of a content creator, not using the data to actually inform what happens next, trusting their gut too much. I've had my ego hurt so many times by the fact that I thought that creative was going to win. I came into this whole industry thinking, "Oh, that type of creative is definitely going to win." And it looks beautiful, it looks amazing from a branded point of view. And then you come to launch it and you're like, "Oh wow, that didn't do anything." It had 0.0 impact and nobody saw it. So then it's a lot of effort for no reward.
So you've got to really fine tune your efforts to being towards what is actually going to drive sales. So it's people are not using the data enough to inform their decisions and they're not being fast enough to enact what they find from the data. And one of the key problems is like your data teams and your media buyers and your analytics people are very opposite in the way that they think to your creative teams, your video editors, your producers, your directors, they just have a different way of thinking and often like the heads butt. So you've got to have a way where, and we have this creative strategist role that sits in between those two different teams that really pulls those pieces together so they can help talk each other's language.

Gabrielle Stafford:
And I think very often our customers use data as that language that unifies those groups. You've also mentioned creative strategists, so the data and the creative strategist, how do they bring this group of people together and make sure everyone is going in the same direction?

Shahbaz Khokhar:
So I think the key skill set of a creative strategist for me is understanding the data, but then also understanding the consumer psychology, understanding what is making them tick. So while creative strategists, they're sat there going through Reddit reading comments about their customer avatar, they're going through reviews, checking out what everybody's saying, they're going on ad comments to see how people are commenting and responding to ads both of ours and our competitors. So the whole idea is the creative strategist is a marketer. They really understand the psyche, the demographic of the customer, and then they're connecting that to the data and what the data is telling them. So they're working in sync with our media buyers to say, "Okay, this is what the data is telling me, this is what I'm going to produce." And then overall they project manage the execution of the creative as well. So they're the ones who kind of really dictate the priority and the execution and what gets done.

Gabrielle Stafford:
So they're using the data to join together the art and science of direct response essentially?

Shahbaz Khokhar:
Basically, yeah.

Gabrielle Stafford:
So Shabazz, let's talk about creative fatigue. How important is a solution like yours to overcome that creative fatigue that obviously impacts the 5 to 8% success rate?

Shahbaz Khokhar:
Yeah, I mean that's another thing around producing volume creative. One is because your success rate is low, so you have to produce a lot to get a winner. But the other thing is ads don't last long. I've got a chart that I'll actually show you. So this brand is spending around about 2 million a month. The typical pattern is in month one, the spendage or spend share is actually kind of lower, and then it peaks out around month two, and then it dips down month three. And then by month four, the average spend is actually is mostly gone. There is a legacy effect where you have... Always think about it's the area under the curve rather than just necessarily that data point. I'm always trying to think ahead of, "Okay, how much creative do we need to produce?" And it's often a question I get answered and it isn't always a thousand a month. It depends on the situation of the brand-

Gabrielle Stafford:
And I was just going to say, is it different for different brands?

Shahbaz Khokhar:
The way I like to answer this question is because I always say, "Do as much as you can." But the truth is we can only do so much. So I try to reverse the answer and I say, "Okay, how much resource do you have access to right now? What does your creative team look like? How many editors, copywriters, designers, post-production managers? How many people have you got access to right now? Let's break down. How much can they produce in a day? How much can they produce in a week? Now let's work out how much creative we can produce off the back of that if you also amplify that with AI tools, good best practice, et cetera, et cetera."
So the way I answer this question is like's not a number. It's like let's work out how much you can produce right now, let's build efficiencies into your current resources. Once we've built efficiencies into your current resources, then we'll add additional people on top. There's no point in adding people to an inefficient process. So bring your cost of production down, bring your success rate up, then start to add people in. As the team gets bigger, your nodes of communication increase exponentially. So you really got to think about how to structure your teams in the right layout. And I'm happy to share one org chart that we've got that shows how we overall structure the different teams, which I can share to you guys in the chat.

Gabrielle Stafford:
So let's get right down to the detail here. Again, having a chat earlier, you told me your process is seven days to get an ad into market, which seems for consumer ads very, very fast, especially at this scale. Talk us through how you guys do that.

Shahbaz Khokhar:
Yeah, I mean I feel like seven days is too slow. I'm trying to compress it down every single quarter where one of our goals is to get it down and down and down. I'd love to get it to three days or two days, but the thing that ends up taking the longest is actually the approval process. But I can talk you through the steps if you like the stages and unpack each stage in a bit more depth.

Gabrielle Stafford:
Let's do that. But before we do that, let's talk about the approval process because it can be a bottleneck for a lot of getting creative into market. How have brands approached, the companies you work with approached your approach in terms of approving fast?

Shahbaz Khokhar:
It's painful at the beginning, I won't lie. Right at the start, we've got to convince people that they've got to be fast. A lot of clients will want to approve things twice a month or once a week. I'm like, "No, you need to approve this every day." If you want ads to be launched and them to be based off an insight that was working yesterday, we had an ad that spent and made revenue yesterday, you've got to approve today if we're making an iteration. So it's just getting them into that mindset of like, "We've got to approve fast." And usually on the client side, we'll always have one person dedicated to approvals and it's up to them how they get it driven through, but they absolutely have to get things driven through on their side. So that's absolutely key. I think one of the things that's helped us is we built our own process and software that connects it together. So the software that we built bespoke for ourselves actually enables them to do the approvals all within the work OS. It's a work OS and approval software all built together.

Gabrielle Stafford:
So let's have a look at the process because I think people will be really interested in that. And when we were chatting earlier, it's very agile and uses a lot of the same kind of phrases and procedures from the development Agile process.

Shahbaz Khokhar:
I mean it's been inspired completely by Agile really. So we typically have a two-week sprint period in which everything gets delivered. And we have the first stage, and we can bring this up on screen as we're talking about it, is the ideation stage. So here, we are capturing ideas and then a creative strategist basically is in charge of capturing all the ideas, logging them, maybe putting an inspiration image or file next to each particular idea, and they'll write a hypothesis against this idea. So there'll be a reason why they will think this idea will work. So they'll always have some kind of an insight either from a competitor, from what a consumer has said or what has worked in the past.
And we split these into new concepts and iterations. So it's either an iteration of an existing ad that we've made where we're trying to improve it. So for example, we've got a call ad, it's doing some performance, but the hook rate at the beginning is too small. We think we can improve that. So there we will change the hook on an existing ad or maybe it's the second or third frame and we can see there's a sharp drop-off in the view rate on an ad, and so we'll go in and swap that segment out. So that might be like an iterative concept that we capture. The other type is we've just got a brand new idea that we've got and we want to just test it and we try and split things half off between iterations and new concepts.

Gabrielle Stafford:
Going to be my next question, what is the split?

Shahbaz Khokhar:
There's never a perfect balance, but I'd say generally speaking, iterations always perform better than new concepts because you launch one ad that works, you'll only get so much yield out of it. So the whole idea is there was something there that was working, so just keep improving it, keep adapting it, and usually one ad turns into hundreds and hundreds of ads, which actually overall in total make the millions in revenue. You can't depend on one ad to do it. We do have breakout ads that end up doing hundreds of thousands in sales, but honestly, you can't predict your revenue around that.

Gabrielle Stafford:
Which is why you have to use the data.

Shahbaz Khokhar:
You have to use the data.

Gabrielle Stafford:
So let's keep going on the process.

Shahbaz Khokhar:
So we capture the idea, we create hypothesis, we then have a sit down, we have an ad planning meeting and we prioritize. So we might have 20 ideas or 10 ideas or 15 ideas. We kind of narrow that down to the best 5 or 10 that we want to work on in a particular sprint. Each idea actually gets built into several variations. I could give an example. When we launch one idea, it's basically one ad set. Within that one idea, we have four to six variations. So each variation might be we're swapping a bit of copy or we're swapping an image. You'll have an ad plan that's basically got maybe 10 ideas and within each idea, you've got a set of variations.
So the team will sit together, agree on that, we'll then present usually to a client to get their buy-in. And then once that's done, the next step in the process is we actually go to briefing. So this is where we brief the.., and we built this all into our own custom processing system because we were using monday.com for ages and I've got all the love in the world for Monday.com, but we broke it pretty quick because we did so much automation and scripts and things to make it work in the background and we couldn't do it anymore. So we ended up building our own custom software. So we do our briefing in there, we write a hypothesis, we write a full instruction for the video editor, usually in a storyboard format.
Well, we do open and close briefs, which I can talk about in a second. And we do image briefs with layouts. And then once a brief has been written, the copy goes to compliance to check that we're saying things that we can say, claims we can make about the brand. And then we have an auto ad naming convention, which actually is a key part of what we use to get creative data back out of Supermetrics. You get the on platform data that comes back, so you get all of our CPCs, CPMs, breakthrough rates, et cetera. But what we actually do is we name our ads in a very particular way, we have parameter, underscore parameter, underscore parameter.
And what this allows us to do is to break down all of those into a table when we pull the data and actually extract what was happening in that creative. So as an example, we'll have... And I'm happy to share our naming convention with anybody afterwards or in the chat, just ask and I'll give it to you guys. We have the product, underscore, what offer we are running, underscore, which content creator, underscore. There's a whole set of parameters that we actually analyze.
So if we're using that, we can actually cohort or subset our data. We subset our data and look at it and we're like, "Okay, is this the factor that was actually making these ads win? What was the element that actually contributed to the success of this particular ad?" So we actually have to do that naming ourselves. We've automated it in our own software. So we have a way where, as the brief gets filled in, it populates the ad name and then that goes into the ad account, then it comes back through some metrics into a sheet, and then from the sheet goes into our dashboards and things like that. So that's the briefing segment.

Gabrielle Stafford:
And let's definitely share that naming structure because I think people will be really interested in the data.

Shahbaz Khokhar:
No problem. I'm happy to share any of this process, whatever you want.

Gabrielle Stafford:
Great. Data governance is a big topic for many of our customers. So let's share that after this.

Shahbaz Khokhar:
So next stage is you go to editing. So this is where post-production happens. So work gets allocated to a specific editor who's free, who's capable to do that type of edit. If they need motion graphics experience, they'll go to somebody who's got that. If it's somebody who's more of a graphic designer, we lean that way. Our team is typically compromised 70, 80% video editors and 20, 30% graphic designers only because graphics are quicker and faster to produce, you need less people to produce the same number of creatives. What we end up actually launching is around about half-half between videos and images. So the videos actually cost us more to make, and generally the images actually perform better. So there's a bit of a balance there to actually get the resource allocated effectively. It goes through post-production next stage. After that, it goes through approvals, which is my biggest bugbear of trying to get things fast.
We have an internal approval process. You've got a creative strategist who basically looks at the ad after it's been produced, looks at it and says, "Okay, did this hit my brief?" Sends it back to the editor. "No, I don't like it. Yes, I did like it." I target the team to only have one revision in total across internal revisions and external revisions. It's quite hard to get that on average, but that's what we do target. Some things fly straight through, other things go through five, six rounds and you're like, "Why did I even start this?" But yes, it goes through an internal approval process first with our team, then it goes to our compliance team. They check the copy to make sure it's on point with what we can say, and then finally it goes to the client. And usually that's the slowest part in the process. And like I said, I'm always trying to convince them to go faster and faster, but they then sign off the ad, comes back to us, we then launch into the ad account using that automatic naming convention.

Gabrielle Stafford:
So seven days to get to that point. And at that point the ad is out in the world and consumers are reacting to it. What happens then?

Shahbaz Khokhar:
Yeah, and I'd even caveat by the way, within that seven days, most of that time is stuck in approvals. So the actual time to produce things is a lot less than-

Gabrielle Stafford:
It could be faster.

Shahbaz Khokhar:
It could definitely be faster than that. So I'm constantly trying to optimize this process and put clients on SLAs to actually improve their internal review time. But sometimes I win the battle, sometimes I don't. And then the final step is the conclusion. So here you are basically looking at the data of what happened, looking at the ad and looking at what actually happened.
So you're looking at all the different creative metrics. You're looking at... Like I mentioned, about subsetting the data to understand which aspect of this ad actually made it work and perform. And from that, you're drawing a conclusion against that original hypothesis that you wrote during the briefing stage. And then from that you document your learnings. And I'm always asking the team to actually share their learnings internally with everybody else. That's within the client account itself and also across clients. Because we find that sometimes we get a winning concept over here, we just copy and paste it over here and it does work. The overall concept works with a different product and kind of brand image.

Gabrielle Stafford:
So where Supermetrics comes in is really in the conclusion phase where you're looking at that data with all the different cuts to figure out what's worked.

Shahbaz Khokhar:
There's three places overall that we use the data we get from Supermetrics. One is for internal reporting. So this is where we're discussing things internally, making charts, making graphs, try to understand what the data is telling us to make our next decision. The second place is we actually have dashboards that we make. So we make dashboards that we then present to clients. So they'll go into a deck or we'll screenshot or screen share and show a client. And then the third place is really in our media buying process. So this is actually what you guys call the activation. It's like putting the data back into the system and letting it optimize the media buying process.
So in the past we did use Google Analytics data to automate part of our ad buying decisions. So if a particular ad hit a certain threshold in what GA and Facebook said, we would've scale the ad, we'd change the budget, change the bids in an automated way, 24/7. We've got logic in the background that actually does this for us in a software. More recently, what we do is we use Shopify data from Supermetrics to actually create fail safe. So for example, if the amount of sales in a particular hour is significantly less than we're expecting, there's probably a bug in Facebook. And that's happened in the past where we've spent 50K in an hour or two.

Gabrielle Stafford:
I think it's happened to all of us.

Shahbaz Khokhar:
It's definitely happened. We did get a refund on some of the money, but it's always a bit of a panic when that happens. So Shopify basically gives us that data and we can cut ads immediately if we see there's something outside of the expected thresholds of what we expect to happen.

Gabrielle Stafford:
This is a process that you've obviously honed over the last four years and is now really working for you guys and more importantly, driving growth for your clients. If we have somebody listening who wants to start on this journey, what should their first step be?

Shahbaz Khokhar:
I always kind of like to zoom out first when I first work with a client. I like to think about the traditional four Ps of marketing, which we laugh about in the office because it's kind of a bit old school, but it does work. You've got your product, so you've got to think about which product do I want to focus on first in my range? Don't try and promote everything. Just ignore the stuff that doesn't work, that's not profitable. Just focus on the stuff that you really want to push because it's shown profit and shown sales. So focus on that first. Get that honed in. Second thing is your offer and your price. So those things are kind of combined, makes your prices competitive against what your competition is doing. There's no point in launching ads similar to your competitors, but their price is half or 20% or 30% less. You're always going to struggle.
And have a good offer that's going to compel people to purchase. So you might do a bundle or a kit that's only available online through your online store, or you might have something based on urgency or scarcity. Having a specific time of the year where you have a sale and you have a cut off. And the final thing is choose the people, choose the avatar that you want to target. Once you've got those pieces together, you've got your product, you've got your offer, you know who you're going to be targeting first. You then start to build your creative testing plan. So I'd say sit down with Reddit, sit down with your ad comments, sit down with your reviews, read these for a day, two days, three days. Just go head deep into this stuff. I like to use Miro. I just map all this stuff out on there.
And you want to review your customer's ad libraries to see what they're up to, what's working for them, what's been running the longest, and you want to look at what they're doing, what you've learned about your consumer, choose your product you offer and your avatar. Put all those things together to make your first ad plan. Come up with 5 to 10 concepts and within each concept, yeah, have a few variations of things you're going to test. Start with that, produce that first, and then from there, then go through this process of... And it's a never-ending process. It's not something... I've drawn it linear, like left to right, but it's not. It's a circle. This thing just keeps going round and round. And the faster you can make it go round, the more money you're going to make.

Gabrielle Stafford:
And I think that circle, the 360 is super important because it's put the stuff out into the world, look at the data, use the data to drive your next action, put the stuff out into the world, look at the data, use the data to drive the next action, and it goes around and around. So I don't think we can talk about creative in 2024 without talking about technology and AI. Are the new tools helping you do this as fast as-

Shahbaz Khokhar:
Yeah, I think the other of the day we actually did analysis. We have about 60 piece of software running our entire marketing effort, and I think probably about seven or eight, maybe even nine of them are AI tools. We've been using AI to produce. We've done in excess of 10 to 20 million at least, I'd say probably in that range of revenue with ads made using AI. I can give you a quick summary of the stuff that we use. So we've been using Midjourney from the beginning when it was launched to produce all kinds of statics. We use Photoshop Generative Fill to then take parts of the static and do things to it. We use FluxLora, which actually gives much more realistic faces than what Midjourney does. So if we're trying to produce faces, that's what we use. We use ElevenLabs to do voiceovers. Sometimes we train a voice based on what content creator has... Content they've given us, and we use that to train a voice that then gets put into a script that's used as a voiceover.
We use Arcads to actually make content creators move their mouth and say things with their sign off that we want them to say. And that's quite a very quick way of producing. There's a whole raft of things that we use and it's always changing. I'd say have somebody in your team who's dedicated towards researching AI tools. So we have somebody who literally sits there, looks at tools, tests them out, produces ads. If it works, they then get the rest of the team to do it. They train the team on how to do it. And I think that's literally the most important thing right now to get on because the cost of production is coming down and down and down, and it enables to produce volume at speed. And if you're not adapting to using this AI technology, you need to be. And people who are traditionally, let's say brand focused, brand marketers get really concerned about this kind of stuff. But I think there's a way of doing it, where you can achieve the brand aesthetic without compromise and still get that speed.

Gabrielle Stafford:
And I think even in my own network, I've come across brand marketers who are still a little, "It's not going to be good enough." But I think testing again is the thing. Don't take that on face value. Test it and find out. There's so much potential.

Shahbaz Khokhar:
One of our mantras is literally in our values, is test it and see. We'll have a battle internally. No, I think I'm right. No, I think I'm right. No, you're not. Just test it.

Gabrielle Stafford:
And I think when the industry standard for creative is between 5 and 8% works, you have to test.

Shahbaz Khokhar:
You do. Yeah.

Gabrielle Stafford:
So Shabazz, thanks so much for coming today and talking us through how you guys are actually using both creative and data to do this at scale and to drive growth. And I think if I was going to take anything away from today, it would be creative data activation, creative data activation, creative... I'll just keep doing it.

Shahbaz Khokhar:
Yep, yep. That's pretty accurate.

Gabrielle Stafford:
Guys, Shabazz is actually going to be in a lounge right after this session. He will be available to answer questions. I know he's likely to give some stuff away, like some of his templates or show you some graphs. So I would definitely catch him when he is there and go and have a chat. And Shabazz, thank you so much for sharing your experience with us today. It was great to hear about your success and the success you're driving for your clients.

Shahbaz Khokhar:
It's been a pleasure. Thank you.