Ep 172: Which Data You Need To Scale A Business with Nima Gardideh

Jaryd Krause: (00:00)

How would you like to know which data you need to scale your business with ads? Hi, I'm. Jaryd Krause host of the Buying Online Businesses podcast, and today I'm speaking with Nima Gardideh, who is the co-founder of Pearmill a tech power gross studio that combines the power of artful creativity with precision targeting of major digital platforms.

Now, Nima is an expert in growth experimentation, organizational design, operating systems, agile software engineering, funnel optimization, and user design. Plus, he hosts his podcast. So he is a great speaker. Now in this podcast episode, Nima and I specifically talk about how he was able to scale a hotel company to 80% capacity through COVID with paid ads.

We talk about how they order a business to see if the business is actually scalable before they work with somebody. Then we talk about how they scale ads based on data sets and which data sets you actually need to scale the business.

Then we talk about ad creatives like videos and images and how they create ads, images, and videos based on data alone, and what that process actually looks like, which I think is important and fascinating. I also share some of my philosophies in this podcast episode about investing in advertising to collect data and why we all need to do it because the data is so valuable.

We also talk about what actually nets better on cost per acquisition basis, search advertising, or paid social, and why there's so much more in this podcast episode that we unpack but if you own an eCommerce business or a SAS business or want to run ads on your business, this is a great episode for you. You're going to love it.

Do you want to build or grow your content website? Niche website builders have helped hundreds of people to take their content websites from a few hundred dollars per month to over tens of thousands of dollars per month with crafted content creation, buying age domains, and link building strategies.

These strategies have helped people increase their traffic authority, monthly earnings, and their website valuation too. Head to nichewebsite.builders/bob/ to get 10% off, any link building or 10% more from their content creation services.

Hello, Nima. Welcome to the podcast.

Nima Gardideh: (02:34)

Thanks for having me. Jaryd Krause: (02:35)

When I first got introduced to you and what you do some pretty cool stuff, and I wanted to unpack what Pearmill does not just for myself to understand alone a bit more but for everybody listening. So what is your company Pearmill? What do you do? What's the goal when somebody comes and works with you?

Nima Gardideh: (02:57)

I'm happy to walk you through that. So we are a paid growth studio. What that really means is we bring on essentially three separate disciplines together to help companies grow. We bring in the sort of traditional growth, marketing out operations, media, and buying discipline. We bring on creative production, creative agency, discipline, engineering, and data science disciplines together to build a social platform to help companies.

Scale platform is really a combination of processes. Plus some technology for reporting and understanding the ad spend across the brands to help companies scale and sort of maybe 50% of our ad spend is consumer a little bit over 50% is consumer either e-commerce or consumer SASs and the rest is marketplaces like all over the place in terms of clients that we work with but the process is essentially the same.

Jaryd Krause: (03:46)

So I want to unpack more of that in how you sort of user data with creators and your framework for that you mentioned in your content on your site but before we hit the record buttons you mentioned something pretty cool that you help brands buy other brands by working out what sort of brands would be a good fit for them.

So a lot of people here listening are wanting to buy businesses and have thought about the ideology if I need to grow a business, I can buy clients through paid advertising and marketing and stuff like that but you can also buy an audience via buying a business.

So what are some of the things that you have done to help other brands acquire the right businesses to put underneath them, and scale that's very broad but I guess we'll unpack that?

Nima Gardideh: (04:39)

We'll get through it. So the way that we can help with these types of things. We just understand what I call sort of auction dynamics of these paid channels. So Facebook and Instagram and Google and Bing and all these major institutions that you can advertise on all have their flavor of the same thought process on how to price an auction and price an impression.

So what we can help with, and this is some of our clients are these types of companies that buy other e-commerce brands and scale them is try to get an understanding of, is there some clear path for growth in a certain market for that specific brand? And really what we can do is if you let's say, have access to a Facebook ads manager, you can get AdSense if they're going after audience X, the size of the audience is a million people.

Nima Gardideh: (05:30)

And we think the current conversion rate will hold how much more revenue can regenerate from the specific audience, and can we expand that on the audience in a way that will maintain the current customer acquisition costs? So then we can scale the brand top line over an X number of months, and it takes, I'm overly simplifying something that does take a while but that's kind of the gist of it, is there space in the auctions at their current sort of rate or is there a way to increase the size of the auction?

This is a thing within Facebook and Instagram specifically if you're going after larger auctions, you're going to end up getting cheaper clicks. So is there a way to increase the size of the auction where it's still relevant?

It's still people that we could put by the product but you're going at it in a way that has a larger potential audience on the platform, and the reason you end up getting cheaper clicks is to do with the machine learning models underneath it. We don't have to get into the details of that but you end up getting cheaper clicks to larger the audience targeting audiences on Facebook right now.

So there is an easy path for us to sort of estimate this space and then on paid search, it's much easier. They give you this metric called the percentage of impressions shared that you have on a certain keyword, and if the brand is been let's say operating for a couple of years, and they have a few keywords already out there that they're spending on you just have to look about around, how competitive are they?

Is there space to grow? How much of the share of the impressions are we currently sort of buying to get? And if we buy a hundred percent of those, what does that growth look like? Is it going to be still sustainable in terms of price? Because the bid increases as I try to increase my share because other brands are willing to also compete with me on it. So some calculus has to go behind that but putting these two things together, you've been spending a couple of years already on Facebook and Google we can get a sense of how much further can you do it without fundamentally changing the approach you're doing to advertising.

Maybe your brand has to get a lot more attention to it. And you need to produce more products and more skews or whatever. This is like a baseline of your current strategy. Can it scale a little bit further? How much further can it go? And it's a very good snapshot to take of brands if you're trying to buy them.

And if you, promise your investors or yourself that you need to get a 10% return within the 12-month period then this is an easy way to get a sense of, actually I'm going to be on the specific brand and then create a model around it.

Jaryd Krause: (08:12)

So it's essentially seeing if the business is scalable and the only way that you can see the business is scalable is through the data and those main data points are what is your cost per acquisition, and then your audience size and can that be expanded. That's really cool, and I guess that doesn't just go for buying, seeing if you're going to buy the business and see if it's scalable, when we look at different types of business models, like we go for e-commerce SAS in terms of that sort of paid advertising but also how much traffic is available to be acquired through content if you are running a content site.

So I think it's pretty cool. Not just if you're going to buy a business or acquire a brand under another brand but also if you've just, I guess you go through that process for anybody that comes to you that's like, Hey, can you help me grow my business? That's probably the first point of call, would that be it?

Nima Gardideh: (09:08)

This is one of the reasons we do free audits we want to know if there is enough space for you to grow. So then we can be part of that future. A lot of agencies would charge you for just taking a look at the account. We'll take a look at the account and give you a ton of feedback without even charging you for it.

If you're part of our sales process because we do want to see a certain growth rate, we don't have a lot of clients like our whole model is to just have sub 20 clients at any given point but they should be soon should be spending millions of dollars, and for us to see if you have the potential to get there we need to take a look at these metrics ourselves.

We're pretty honest about that upfront we'll tell you, Hey, we'll just say no to you but we'll regardless give you a reasonable amount of feedback on the accounts and what we would do if we were to take over but there's a chance we'll look at the account and say, we just don't think we can help or there's not enough space for us to grow your brand within the next 12 months.

And it doesn't make sense for us, it's kind of like a hedge Fundy way to run a growth shop. A growth marketing shop but it's the way that works best for us. I think because one is we get to work with some really great brands we're scaling rapidly and we get to be part of the journey. And the second thing is it's just more stable.

We make very long commitments to the brands we've been working with of them for years. And I don't know if you've been in this part of the world but technology and specifically advertising, you end up having a lot of folks that get jaded by having to work on new clients all the time, and they don't get to really feel like they have ownership.

And so we completely eliminate that to some extent. And so that's kind of the model for us, it is a little bit harder in that we have to, we compete with folks that have a lot of money on the other side as well. And that becomes a little bit harder to scale your efforts but it's worked for us over the past few years.

Jaryd Krause: (11:02)

You're using a long-term game plan rather than short term. Let's get anybody in the door and make them a bit of money. And then the life cycle of the growth of the business may not last well into the future, which means you have to go back to the drawing board and get more clients, and just can be messy.

Nima Gardideh: (11:20)

Totally, and I think if you think about it, all these networks like Facebook and Google if you're spending on them, they are caring about longevity. Now a lot more, it's literally at least on Facebook's end one of the probabilities in how they price your auctions is how likely this user that's seeing this ad is to come back to not only the platform Facebook but also to your site in the future.

So there is also an advantage of thinking long term for each brand that's going out there on how do you build relationships with folks that you're selling your products to such that they keep coming back and they use your products and services and your community and whatever else is you're putting out there. These networks are trying to incentivize that because that's how we end up building more stewardship-based companies that are trying to build very good software or very good products out there for people to use.

Jaryd Krause: (12:11)

I think everybody listening to the podcast probably had it drummed into themselves enough by me saying, it's not just about trying to make a sale. What comes before the sale is building a relationship because that's where the trust comes from and how strong the relationship is built.

So I think that's a really good point. I wanted to talk about the framework that you guys sort of used to navigate through advertising with Facebook and Google, like to sort of unpack it a little bit and maybe how you used it for say a brand that you helped scale. So you've got five different things. You've got ideation prioritization, implementation analysis, and then systemization. So if we were to use a brand that you have helped scale that has got, you're pretty proud of working with them and you've got them a great result and they're happy.

Would you be able to walk us through, like, how do you first start with ideation when they come to you? And is this after you've sort of done your audit to see if that brand is scalable?

Nima Gardideh: (13:15)

I can walk you through an example. Ideation essentially starts after the audit. So let's say you come to us and say, Hey, I want to scale my brand. We'll take a look and say, we think we can help, and we go through this process every month and I'll walk you through it but cool. There is a general version of it that is applied to microwaves.

So that general version you just talked about, there was ideation experimentation, analysis, and systemization is essentially what is known as the growth experimentation process. There are growth leaders like Sean Ellis, Brian Belfor, Andrew Chan, and these folks that have built.

So these Silicon valley companies have built this general process. It's quite simple to think about. You're just coming up with ideas, you experiment with them, see what the result is, and use the results to build a system around that idea.

But the problem with this general approach is it's hard to actually implement. So let me walk you into some examples of how we do it, and how we fall into our business. So every month, we have a week we call the sprint planning week, and we've stolen some of these concepts from engineering because they also are just experimenting at the scale where we try to come up with all the different ideas we have on what we could do to improve the account.

So there might be what we call targeting or structural ideas, and that is how to campaigns or assets and ad groups are structured in the Facebook and Google accounts. It could be targeting ideas. Do you know what I just told you, hey, maybe we need to increase the audience size by 20% to see what happens.

There's a bunch of ideas in that area, there's a second lever which is creative. So we're going to come up with a bunch of creative ideas on how do we try to pitch this brand in a slightly different way? What are the past learnings that we can use to iterate on and all this sort of stuff? So that's another lever.

And the last one is data which is what are the learnings we have around what works and what doesn't in terms of maybe the conversion rates of different ad groups, audiences, or first and last touch attribution that we've gone across the board. How can we use that to then run experiments, and then there's a newer area. We still fall within data for us. Is that the signal you pass back to these ad networks?

So if you tell Google that I want this type of user essentially to take Facebook, I want this type of user by telling them that they convert it at the right part of your product they will sub-target your audience for you in such a way that they are hoping to get people that are more likely to perform that action.

A simple example is we have a client named Saunder and I'll walk you through a bunch of examples across these four levers but let's start with data. One Saunder is a distributed hotel, we've been working with the right since right before the pandemic. So they had a tough year and the first year we worked with them because the pandemic was happening and they were hotels.

So they were able to actually maintain their margins quite miraculously by changing their hotels to be sort of shelters for doctors and nurses that needed to be away from their families. And they did that effectively and did it through paid social advertising but there is really no great targeting for, Hey, I want to target nurses and doctors in a certain city that maybe have families and are on the front lines of COVID.

It just doesn't exist and is looking for hotels right now. It just doesn't exist on Facebook. So what do you have to do? You have to essentially target everyone in that city. So the targeting was anyone in the cities in which Sandra operated in, and then you have to tell Facebook, I want this action which for them at the time they wanted longer stay term stays. So we were optimizing towards a long-stay purchase event.

Instead of any purchase or any booking, we were optimizing towards a specific type of booking. So what this taught Facebook over time was that we're looking for clearly this persona, it was doctors and nurses, etc that were looking for this thing, which long stays over maybe a couple of months period after we spent quite a lot of money to feed the algorithm.

We were able to bring costs down dramatically and start ballooning and became quite a great account but it was a very hard period, not just for their business but also for the trust of their team to know that we know what we're talking about. It's hard to see it because you're feeding the algorithm in that portion of time until you hit the scale that you need and it works. So that was a very clear experience.

Jaryd Krause: (17:50)

Can I just pause you there? I just want people to understand this when they are doing paid ads, what you're doing is you are putting money in to collect data at the start. That's all that you are doing, and when people if they're trying to run it themselves or they're going through an agency and they realize like, I've just spent so much money, I haven't really got the results.

That's pretty difficult at the start because you are just giving money to Facebook and Google to get data, to work out if it's going to work and how will it work and what are the right ads you should be optimizing. So I just wanted to jump in their name, I am sorry for cutting you off.

Nima Gardideh: (18:28)

It's a very important point, and especially over the past year since the iOS 14 changes, this has become even more of a reality because unfortunately Facebook's algorithm still needs the same amount of data but it's not getting less of it. So that means you have to just pay more to get to the scale in which the algorithmic sort of optimizations or reductions come into play, and that's sort of the gist of the overall process.

And walking through an example of creativity and how that works, for a data piece, that's simple, hey, we want to see if we can use a different signal to see if we can reduce. So we're going to create a different event optimized for that event. Wait a certain number of weeks, and see if the change dynamics of the account change but for the creative, it's actually much harder. I would say, it's not as clear.

Jaryd Krause: (19:19)

This is for ad creatives for images and videos.

Nima Gardideh: (19:24)

A creative that could be images or videos that you're producing on Facebook or Instagram.

Jaryd Krause: (19:29)

Yeah.

Nima Gardideh: (19:29)

So we do use the general similar process. We come up with a bunch of ideas. We run them, we try to see if they worked or not, and then use the learnings to go on, to do the next step but the part that's much harder, and this is what we've learned over the past few years is you can't isolate the learnings as much. After all, it's really hard to produce a video in 30 different ways with all the small nuances and the differences that you would want and then run that same idea 30 different times because it's just going to cost you millions of dollars.

You'll have to isolate every ad and run enough money behind it that it gets conversions. So you see what the conversion rates are, what per ad is and all this sort of stuff. So it's just going to take you a couple of million dollars for every great idea you have because let's say it's like a 15-second or 3o seconds video and there are hundreds of different variations of it that you can come up with.

So what we've tried to do now is still attempt to do that and isolate as much as we can but we do it essentially based on a structural way where we try to test out the simplest versions of things first. We'll start with an image ad that just is, let's say for this specific brand Saunder it's just the apartment and nothing else. There are so many ways to think about how do I shoot the apartment before the user sees it.

And we went through quite a lot of iterations of that maybe hundreds and hundreds of different ways of shooting apartments until we got to a point where we understood three things need to be in the shot for it to work.

And there were things around this like the windows have to show there should be kitchen countertop in it, high ceilings. There's a bunch of like visual aspects that kind of were the thematic things that were in all the high performers. So that's level one, and then you grab those learnings to build, what if I put the discount code or the discount amount or the price of the, of the hotel or their logo or maybe right before you see the photo of the apartment someone tells you about the value props or maybe the value props are over it, there are so many different ways.

Now I can build on top of the core of what works, and you build these learnings and compound them over time, and if you go on look at Facebook ads. On the Facebook ads library for Saunder, you'll see wildly different ideas there. But the core of it is that we know what, and how to take the shots of the apartments. And then we built so many different learnings on top of that idea, it's much harder to nail, I'd say it took us many years to master this and it's like constant learning for us.

And it's also very hard to get people that are doing a production of creative, who are interested in this way of thinking, which has been the hardest part for us. I could imagine it's not just knowing that this is a thing you have to also find the videographer who thinks that is interesting and they want to play around with it.

Jaryd Krause: (22:33)

Yeah, 400 times.

Nima Gardideh: (22:36)

And so that the combination of those things is where the magic and creativity happen. It's just very hard to find the right people and the right process to make it work but when you do nail it works extremely well because you essentially go after the people that want your product, they want it the way you are presenting it, and they end up wanting, buying and discovering what you're trying to sell. And it works sort of perfectly for everybody involved.

Jaryd Krause: (23:03)

I think it's a great process, if I come back to what I was saying before, around when you first start running ads, paid ads, you're putting out paying money, you're investing money just to collect data. And that's what you're doing throughout this process and then honing it in to perfect.

Only put in the things that you need within the short timeframe that you can have the ad in front of that person and give them exactly what they need to see to be able to have a connection with that video or image to go further and start that relationship.

Nima Gardideh: (23:38)

And you only have so much time and money to spend to learn these things of course but you do have to spend it to get that information. So you understand what works and what doesn't right. There are a lot of nuances around creativity in itself because now the way we think about creativity is that it's just a lever for targeting and that the right person can look for a way of looking at the same brand and buy it versus you might have a different taste and sens aesthetic sensibilities compared to me, and you might click on different ads, even though we're both in the same target market for, let's say brand X.

So not only do you have to figure out what is the composition of these videos and images that work also what are the different versions of those that work for the different parts of your clientele and your potential audience? And so there is an infinite amount of things to test. And you do have to spend money to get the data back to see if it works or not but hopefully, you're sort of doing it in a way that you aren't wasting too much money.

Even with Sandra, as an example, they were profitable within a month, one on the ad spend or not profitable but basically breaking even, and then now they're doing quite well many times over but in the beginning, it was a big commitment for us and them to uncover this to make it work, given the situation was dire, their other channels had completely stopped giving them business at the time, which makes sense. I think to sum it up, I think you're correct. It isn't to say that you're just going to go out there and waste money though. It's an investment that will give you the information you need in order to figure out how to make it profitable.

Jaryd Krause: (25:24)

I always like to use that as a principle for all things, even when somebody's like I'm not into starting businesses, which is why I tell people to buy businesses because 90% of startups fail. But I even, if somebody was a startup business, what you're doing is you're investing your time to work out. What's going to work and you're just investing that time to collect that data but what I think it covers, it goes across all sorts of fields.

I think it's more of a philosophy in investing to be able to collect the data, which is critical to have the confidence, to be able to scale, right. If you're taking on an investor to grow your business, you want to be very confident about what return you can get. And that can only be based on the data that you have of the scalability. So tell us a bit about the results.

It could be figures, it could be bookings, it could be anything that Saunder was able to get through that period of working with you, especially through COVID, which I think you did exceeding well to help with company grow in the hotel space over COVID.

Nima Gardideh: (26:20)

I don't think I can share exact numbers but I can tell you that they had close to 80% occupancy within the month. One of working with us, which was right around the pandemic and they maintain that. And these numbers are public, so you can look them up because they've just iPod actually in a couple of months ago. So they've gone through a massive growth spur actually over the last few years.

So it has been a good couple of years for them even given the situation and the market has changed. They are quite a nimble team and we're able to sort of remarketing their company for that period as a way to help people survive through the pandemic better. And now they're going back to their sort of bread and butter which is helping business travel and travelers to find space while they're staying in different places but their average stay has increased and they're scaling quite rapidly now.

Jaryd Krause: (27:12)

Excellent, well done, especially for that period. I'm sure that the first month would've been quite tedious. I wanted to ask you about the differences between Google and Facebook and Instagram. And why would you lean into one or the other depending on the different brands?

I have my ideologies around that based on tents and where the audience for that is particular or how the audience likes to consume content as well but how do you go about choosing, let's check out these platforms or which platforms to test for which brand.

Nima Gardideh: (27:44)

I think this is a very good question and you have nailed some of them immediately, if we think that there is existing intent for what you're selling search is an obvious path, and the good news about searches, is you can just look up if there is intent for what you're trying to sell or there are keywords, platers, and plenty of companies that do the analytics there.

The problem with search, even if you have intent is that there is a limited amount of it. You're finding people who are at the bottom of the funnel, almost like they're literally looking to buy now. And sometimes there may be researching but the intent level, in general, is much higher than anyone you would find in what I call sort of discovery based channels. So these could be Facebook, Instagram, or YouTube.

These are where you passively consume content and you could stumble upon something. An app that tells you about X product that you may be in the market for a couple of months down the line or even tomorrow because you didn't know that existed until today. So search problem is scale because there are only so many people that know what to look for and even forget about looking for exactly what you're selling, and so I do encourage people to look at search.

I've seen only a handful of businesses be able to scale search the problem, and then the reason for that is, you're maybe setting a couple of products and there's only a handful of keywords you can buy traffic from, and they have a maximum, the businesses that can scale search are the ones that are either constantly generating new skews or new products or they're fundamentally shaped in a way that gives them an advantage on that channel.

So a good example here is probably something like booking.com. Every combination of Cities, destinations, and dates is right there. Those are so many different sets of keywords they can go after. They are the number one advertiser on Google ads and they have the biggest team in turn. They have like 70 people or something like that on Google's side that just helped their team because they spent billions or something like that.

So that business is just primed to be excelling at search, but other businesses might not be because they don't have these like obvious generative content that they can use to run search outs towards, and so for, and then on the face discovery channel side, or paid social plus maybe YouTube and those channels where you're passively consuming content.

I do think it's exactly, as you said, are they in the right place. If you're selling to younger people, maybe you should try TikTok and snap first, before you try Facebook and Instagram. TikTok especially has, I would say stolen a lot of our budgets from Facebook to TikTok.

We do TikTok as well, but we've shifted a lot of our budgets to TikTok because it just works right for a specific subset of brands. So you have to think about audience a lot and then if you have no intent, it's obvious you have to go out the page social channels, and then you have to be aware of the competition there. So you have to find a way to stand out.

Jaryd Krause: (30:49)

I wanted to ask around the acquisition cost by acquisition between the different ones because search is like somebody's typing in where to buy a watch or something like that, will be because somebody's pretty close to purchasing it with the high intent from search. Would you typically see a lower cost per acquisition via search than say with these social platforms or is it just depends on the business, the services and products and everything like that?

Nima Gardideh: (31:20)

If you put a gun to my head, I'd say, yes. I would say there's a lot of nuances there. That's it could be in a very competitive market such that the keywords are so expensive to pay for it. The clicks are so expensive that actually paid social is a better path. A good example in that the booking.com example is that hopper is competing quite well with booking because they're just doing social well and booking and the biggest travel sites, we're just not spending on social.

So there was a lot of space in that area and they were able to compete and capture a lot of audiences. So it does depend on your market but yes it should be cheaper on search assuming that it's not super competitive, and you're able to capture some of the clicks.

Jaryd Krause: (32:09)

It depends on how your remarketing is as well.

Nima Gardideh: (32:15)

And there are different levels of intent on Google. If I'm searching top coffee brands of 2022 that's some intent I'm looking for coffee but if I'm searching how to deliver coffee to my door tomorrow, that's a different level of intent. I want the product now, it also does depend on the level of intent. So those obviously higher levels of intent search keywords or phrases are going to be more expensive.

The ones that have lower conversion rates, which means you have to do retargeting. Maybe they are less expensive but the overall costs still could be too expensive because they were retargeting strategy but the lower funnel with very high intent sort of sets of keywords could be more expensive but higher conversion rate which ultimately could have a lower rate. So it just does depend on what space you're in and what the cost dynamics are.

Jaryd Krause: (33:08)

The platform Nima, thank you so much for coming on and having a chat, hugely beneficial so much value in this conversation, where can we send people to find out more about Pearmill and what you're doing?

Nima Gardideh: (33:21)

For sure, they can find us on www.pearmill.com and send us a note there. We also have a Twitter account @pearmill and we're happy to work with brands that are scaling, and working on paid search and paid social. So feel free to reach out. And I'm always happy to answer questions. So hit me up on Twitter.

Jaryd Krause: (33:40)

If you have an Amazon business and you want to get off Amazon, a lot of people listening in that boat. Thank you for listening. If you do have an e-commerce brand or somebody who does or is looking to acquire one, please share this podcast episode with them. It will be hugely valuable. It's great for us to grow the show and it's great for you to get that info. So thank you so much there'll be links to everything in the show notes below, and I'll speak to you guys on the next one.