Why This Startup Spent $50k Building an Email List (Spoiler: They Had a 5x Return)

Transcript

Felix:今天我加入了Gaurav Agarwal,副总裁of Growth from Molekule. Molekule is a technology company who has reinvented the air purifier, and was started in 2015 and based out of San Francisco. Welcome.

Gaurav:Thank you, Felix. Thank you for having me.

Felix:Yeah, excited to have you on. So, what did you guys do to reinvent the air purifier? Tell us a little bit more about the product itself.

Gaurav:So, the fundamental difference in one line is we do not capture pollutants. The current devices that are out there, they capture pollutants and they trap pollutants on filter surfaces, which is primarily hyper technology.

What we do differently is we do not capture, but we completely destroy. So bacteria, viruses, VOCs, mold spores, pollutants that are in our air and that are causing us the most harm at the microscopic level. We completely oxidize them and convert them into carbon dioxide water vapor.

So, instead of collecting pollutants, we completely destroy pollutants.

Felix:明白了。That definitely sounds like a better way to purify the air, so appreciate you guys inventing this. We were just talking before we got on about how you’re the VP of Growth, but you’ve been there from the beginning. You know the founders previously, as well. Tell us the origin story. How did this … where did the idea behind a product of this come from?

Gaurav:So, let’s go 20 years back. So, 20 years ago, the chief inventor, Dr. Yogi Goswami. He’s a renowned [inaudible] in the field of solar energy, and he’s currently a professor at University of South Florida.

20年前,他工作在一个职业:博士ject for water purification at Tyndall Air Force Base, and Dr. Goswami happens to have a son who has [inaudible] allergies and nothing out there is actually helping his son from airborne aspiratic conditions. Apparently because of allergens in the air. And that’s the time that he thought, “If I can adapt my solar research to purify ground water, can I do it to purify air?” And that’s how he got started.

So, the process that he came up with is called photocatalysis. And over the next 20 years he would then optimize the process and bring it to a place now where it’s ready to be commercialized.

So, it wasn’t like someone came up with this idea and this company came into being. It took us 20 years to be ready with a piece of technology that is super effective, completely destroys pollutants, does what it say we do, and then, extremely extremely effective. And hence, we are here today with Molekule, which is his life’s work.

Felix:明白了。So, were you familiar with this technology from the beginning, or when did you first get wind of this technology, this product?

Gaurav:So, I’ve known Molekule since we were in our hardware exterior program called [inaudible] One, which is in San Francisco. So, I have known … at that point of time we had the technology, we did not have the product. So, I have known the company from early 2015 when we had this technology and we were thinking, “What do we do with this? Do we go ahead and do we talk to the industrial houses? Do we talk to commercial real estate developers? Do we go ahead license this technology to automobile and the airplane industry?” So, I have known the company from the time when we had the technology, but we did not have a product.

And then we decided that the first thing we really want to do right now is build up a strong consumer plan and then go after B to B verticals.

Felix:明白了。And what’s your background with marketing and growth? What were you doing prior to Molekule?

Gaurav:So, prior to Molekule I was a product manager. I was the lead product manager at a marketing media lab company that was trying to do smart cameras. And before that I was in investment banking with Barclays Capital, and a lot of my work involved credit default swaps and trading bonds for Barclays Capital across. And I spent time across [inaudible] in Japan and Singapore. So, that’s where I really perfected my craft on being [inaudible] and understanding less deeply, understanding growth opportunities, understanding business.

And then my time at LensBricks, which was the MIT media lab company where I was a product manager, helped me understand business from a very product perspective.

Felix:明白了。So, you’ve never had the title of marketing and growth before, but you’ve picked them up from previous roles and combined it together. What attracted you to taking on this role at Molekule?

Gaurav:Well actually growth is my calling. In the sense growth is the perfect balance between building a business, but doing it through building a product and community. So, when I look at my role, I can play across anything that I want it to be. I’m very quantitative, but it’s just important that I choose the right levels that are going to enhance our business.

So, oftentimes, people confuse growth as marketing, which is fine, but that is because if your business is very marketing driven, then the biggest opportunity to grow the business is within marketing. But if your business is very product driven, for example, Dropbox, Facebook, and you have network effects and marketplaces, then you should really think, “What is going to drive the core growth premise in your company?,” and then begin there.

So, even for Molekule, we believe that we can keep spending a lot of marketing, but ultimately, if you build a great product that’s when people’ll talk about it, and at some scale, the word of mouth will eclipse our paid marketing.

So, even though my role is growth and I focus a lot of time on marketing, I also focus a lot of time on retention and how do we grow organic word of mouth?

Felix:明白了。So, how do you know if you are a product based business, or a business that’s more powered through paid marketing?

Gaurav:So, that’s a really good question. I think in the long run, product based businesses survive, paid marketing businesses do not survive. So what I mean by that is … and, by the way, it also depends on what scale you are talking about. But let’s talk about … let’s take a couple of examples.

If you are a brand … let’s say, if you are a skincare brand that’s trying to sell online, you might be able to get your first customer in through paid marketing, but eventually what would matter to you is, is that person buying again and is that person buying other products that you have to sell? Which that means, you are a product based business. Marketing was just a way for you to get your customer in the funnel and for that customer to buy from you, but after that, what is rally going to build your business is, is that customer happy with the purchase and is that customer going to come and buy again?

So, that’s a perfect example of a [non-tech] product, but if very well you are selling a product or you are selling a service so you can pay marketing dollars to get people to use the product or service, but eventually the longterm growth and the viability of the business would depend on how good of a job you do.

Similarly, say if you are an eCommerce company and you are a marketplace provider for other people to come and sell products, eventually people would see how easy or difficult it is to buy on your site. And if it is easy, they will keep coming again because they are getting what they need from your site in a very easy fashion. That’s why we post up on Amazon and not on Jet.com because it’s fairly easy to shop at Amazon.

So, ultimately I do think in the long run. It’s the product and the experience that’ll decide if your business is viable or not.

Felix:明白了。So, it sounds like there are two core benefits that I heard you talk about when it comes to focusing on the product. One is to make sure the customer is happy so that they will spread your product through word of mouth, and the other is so that they will become repeat purchasers and come back and buy again.

So, Molekule, there’s only one flagship product, right? So, how does it work when you’re only essentially selling one product?

Gaurav:So, we sell subscriptions. We sell filters, replacement filters. So, there’s definitely people buying more product. And then there’s word of mouth, where people would get their friends to buy this product because, “Hey, I have this cool product and it’s interesting.”

Another thing that happens when you have really happy customers is they would write a lot of good reviews for you, which would help, which would make the funnel more efficient, serve it, and which would make your marketing dollars more efficient. Because the next time you spend dollars to increase awareness, those people were to search online, they would find good things about your brand and that should help and only making your oral marketing more efficient.

Felix:明白了。So, what kind of steps do you take to make sure that the customers are happy? How much involvement are you taking on when it comes to the product? Are you talking about design a product a certain way? Are you talking about communicating in a certain way? What are you actually doing to make sure that customers are happy with the product that you guys are selling?

Gaurav:So, design the product in a certain … even before that, to figure out what a good insight is on which you can build the business on which you want to build the product. Then you design the product in a certain way, then you build it in a certain way, then you talk about it in a certain way, where you’re not over promising, but you’re also not underselling when people use the product. And then, for how do businesses offer physical good business, it’s important that your packaging is right, you’re shipping at the right time, and small small details, such as unboxing experience. When people use your product are they getting the core value that was promised to them or not?

And if they are getting the core value that was promised to them then the next question would be, “Great, you have someone who is a happy customer. So, how can you convert a happy customer into an ambassador? Are you giving them the tools and resources necessary to go talk to other people about it? Do you have a referral program in place? Do you have something that allows them to quickly share a good feedback on their social media?”

So, you have to look at it across the entire funnel, and then how do you work on the feedback that they have given you to go produce another product or improve your product? So you have to look at it from this and then the cycle gets completed.

Felix:正确的。所以,像你说的,有很多steps along this funnel. When you guys are starting out or when you talk to other entrepreneurs that are starting out, do they have to create this entire funnel from the start, or is there a good starting place if they are limited in their time or resources?

Gaurav:So, I think the first place to begin is … and I think that’s the most crucial one, is do you have a product that’s actually solving a problem? That’s the biggest one. So, do you have a product that’s actually solving someone’s problem or creating potential for someone that they haven’t explored before? So, do you have a product that’s actually delivering on its promise? That’s the first thing because if you don’t have that, then you should rebrand it because you will be spending marketing dollars on a product that people won’t like. So, I think that’s the first step.

Felix:Obviously, there is a product that is solving a problem, but how did you guys know that you were solving a problem that the market wanted?

Gaurav:So, we are an air purifier and a lot of air purifiers are being sold already, but we knew that people wanted this because there are 60 to 80 million sufferers in the US alone who suffer form allergy, asthma, bronchitis, COPD, so many other respiratory conditions. And they are constantly looking for ways to improve their life.

But now how can a new entrepreneur do this? Even before you have a product you have an idea of a product. And then that’s classically your startup principles. If you have read that book, I would recommend everyone to read it. It’s called Lean Startup by Eric Ries. But the goal is even when you have a product concept, what can you do in a quick, hackish manner to see if people are willing to pay for your idea or for that concept.

So, for example, if I’m making a system where someone can book a home decoration service via my portal, even before I go live with that portal, I can see if people are interested with it. And if people are interested in it, are they willing to pay for it? And if they are willing to pay for it, then great, let’s go and build this product. So, it’s testing from the concept stage to building the product to final resolution. The more tests you can do, the better it is because then you know that is this product going to work or not?

Felix:正确的。What test did you guys run to determine that … to follow the startup methodology to run these little experiments?

Gaurav:So, we made ugly looking versions of our device, which were just metal boxes. And we made 20 of them in our own lab. And then we sent it out to people on our mailing list who said that they wanted something to help them breathe better. And we sent them these 20 units. Each person got a unit for four weeks and they tested it out. And along with the unit, they also got a survey where we were asking them to report their allergy symptoms over the four week period.

And what we found after doing that experiment with 20 people was people’s allergy symptoms were coming down drastically. Significantly at the same level as a Flonase. And these people would then come to us and say, “Hey, you did something and now I’m feeling much better. How can I buy this?”

And that was a proof point for us, that people on our mailing list would convert. And people on our mailing list are people who actually suffer from allergy and asthma. So we had a market opportunity there, so let’s build a company out of it. Let’s package this ugly box into a beautiful looking device and sell it to our customers.

Felix:明白了。So, the mailing list that you built, which you sent the prototypes for and got that early feedback, how were you able to generate a mailing list without a product yet?

Gaurav:So, that’s a very good question. That was a very small mailing list that we had built to test the concept. So, you may have a very good question. Once even when we had the product … so we have the product [inaudible] … or we had the idea and the design of the product by November 2015, but we actually launched the product in May of 2016.

And between that time period we never revealed our product, but we still had a mailing list of 25, 30 thousand people. And what we did was we used a tool called Unbounce. On Unbounce we created a quick landing page where we were not unveiling the products, but what we had was a product that was covered with a piece of cloth. And what we said in that landing page was, “The future of air is here. We have groundbreaking technology that completely destroys pollutant and gives you relief. If you are interested in knowing about this technology when it launches, sign up here.”

So, we never unveiled the exact product. We never showed what the product looked like. We never talked about features and exact functionalities. But we created a landing page on Unbounce based on the promise of the product and then collected leads on that landing page. And we spent some money on Facebook to drive traffic to that landing page.

Felix:所以,25000年到30000年,这是你米数entioned. That’s not a small mailing list, especially for a prelaunch. How much did you guys have to spend to drive the traffic to that landing page to get that kind of size email list?

Gaurav:About $50,000 to get there. So, the way we did that, we just did a reverse math and this is a skill that I learned in investment banking. If you have to sell a thousand units, for example, I need to figure out what’s my worst case scenario. How can I sell as much as possible? So, we just did a math that out of those 25,000 people on the mailing list, if even one percent of that list converts, then that list is going to end up paying for itself. So, we just did a few scenarios where we said, “What if the list converts at half a percent? What if the list converts at one percent? What if the list converts at two percent? And then, let’s look at the net CPA because of a certain conversion rate and then, decide based on it.”

Felix:明白了。,是it safe to say that it was worth it at the end?

Gaurav:Absolutely. That’s been our strategy with probably five times after that.

Felix:Wow. So, your strategy is still now to drive traffic from Facebook, paying for essentially email leads and then,marketing to them through email?

Gaurav:We did that when we were not selling. So when we’re not selling when the product is not available for purchase, you can either stop advertising or you can drive a lead campaign. So, every time … in July we were sold out, we didn’t have any inventory to sell. So I moved all our marketing budget towards lead capture and that allowed us to build a strong list we can apply when the stock is available. So when the stock was available again in August, we saw a huge uptick in demand than during the launch phase because what happens when you build a lead is you can also nurture the lead, right? So, they’ve been to our website, they already know that the product can do a few things that can be extremely beneficial to them and then as soon we launched, we triggered that list and they ended up buying.

Felix:明白了。So, nurturing the lead, what are you emailing them? How frequently are you staying in contact with them during this phase where you don’t have a product to sell yet because of it’s out of stock or you’re in your prelaunch phase. How much should you be communicating with them?

Gaurav:That’s a very good question. There’s no hard-and-fast way of doing this, there’s no right way of doing this, but you need to ensure that you also don’t want to be very aggressive. You have to watch your frequency, which is how frequently are you reaching out to your customer. So, I would say once a week or once every two weeks is the right amount and instead of trying to sell them product because you don’t have product, you should try and educate them about your product and educate them about the unique aspects of your technology.

Felix:明白了。Makes sense. So, when you first got this mailing list during your prelaunch phase and you were sending the prototype out, the twenty or so, were you also trying to collect other kinds of information, doing a survey, how do you make the most use out of an email list when you’re trying to understand more about the market?

Gaurav:One of the best ways of using an email list actually is to create an audience on Facebook and then … there are lots of tools that you can use that give you a demographic breakdown of your audience … and then, that’s very, very powerful. So then you understand that, “Okay, I have these 30,000 people who signed up, but do I know who these are people are? Do I know what are their other interests outside of Molecule? What are the other pages that they’re following? What are their market affinity to other markets? Are they more likely to travel? Are they more likely to buy a car? Are they more likely to buy a fridge?” So that gives you a lot of information on your customers that is super helpful, like what’s their age breakdown? So, the one use of that email list is to figure out demographic and psychographic data on your customers.

Felix:That’s very interesting, I never heard anyone talk about this, but you’re taking the emails that you’ve generated that you collected from Facebook, and then reuploading them back into Facebook to get a profile of the demographics of your list?

Gaurav:是的, you can use the Audience Insights tools. So, the reason we are uploading that, you can also do it based on the Facebook Pixel, but the reason you would upload it is sometimes there’s a [inaudible] difference when you upload a new audience versus when you base it out of a Facebook cookie. But, Facebook has this tool called Audience Insights and you can use Audience Insights to figure out who are these people and what are they doing and so on. The same thing you can also upload this list into a specific audience in Google or you can just use the Pixel, if you have Google’s Pixel deployed on your website.

You can also go to Google Analytics and get audience information. Google Analytics has a report called Audience Breakdown and Audience Demographic and you’ll find so much more information and you learn so much about your email ID’s that you would not have known before, like what are their other interests? What do they do? Are they married? Are they not married? Are they homeowners are they not homeowners, and so on.

Felix:明白了。And, you’re able to upload an email list to some kind of Google tool as well?

Gaurav:On the Google tool, I would recommend using a Pixel.

Felix:明白了。

Gaurav:So, if you have the Google-

Felix:Like Analytics or something?

Gaurav:Google Analytics. Yes. If you just have the broad Google Analytics, you should be able to do it.

Felix:Okay. Now, what did you learn from the demographic data out of your Facebook or Google and were able to actually take action on that learning?

Gaurav:这是一个很好的问题。What we learned was … Initially when we were building the list, we thought because it’s a tech product, it’s a cool product from San Francisco, people would be interested in buying this are people from San Francisco Bay Area and New York because that’s where a lot of tech products are, that’s where a lot of early adopters are. But, when we looked at the list and when we looked at the demographic breakdown of that list, we found that our list subscribers were from all over the country.

We did not see a huge bias in the coastal region or San Francisco and New York, in fact we saw a lot of people signing up from Florida or Texas, Chicago are, Seattle area, Charlottesville. We got a lot of people signing up from all kinds of different places and then that gave us a unique insight that we are actually not a tech product, we are more of a health product that is needed by population across the country. So, it’s not like cool gadget that only people in San Francisco would buy. The other thing that we learned was our mailing list was … there were as many people on the mailing list that were 45 to 50 plus, than people that are 50 and below, in terms of age.

So another thing that we learned very quickly that people interested in this product, in this concept, are also people who are not Millennials. So it’s very important that when we design the product or when we design the landing page, when we design the website, we just don’t optimize everything for a younger audience, we need to be very mindful of the older demographics as well and create something that can talk to everyone.

Felix:明白了。So, without this knowledge, without doing this exercise of looking at analytics or uploading your email list, you guys would have had the wrong kind of demographic in mind.

Gaurav:是的, and we would have started optimizing to the demographic and we would have missed out on the bigger picture.

Felix:Wow. I can imagine that there a lot of people in that situation now where they have different picture of who their audience is than the people that would actually be most served by them. So, when you are optimizing your landing page with this new demographic that you guys discovered through your analytics, is there an internal team that’s putting together the copy and the images or do you guys have that outsourced, how do you guys make sure that the message that you’ve learned from analytics is permeated through the rest of the company so everyone is up-to-date on the demographics and the brand, essentially?

Gaurav:So, the company’s growing now. Right now we are about 75 to 80 people. So, and a year ago we were 20 people. So when we were 20 people, I would just flag everyone and they would see it. Now we are many, many people and sometimes it’s important for people who are not in marketing to also know about things that are directly related to their work. So, what we do now is every quarter, we would do a presentation for the rest of the company to get a pulse on who is buying, why are they buying, what are they doing with the product, and so on.

Felix:明白了。So, you also mentioned that one important thing when you do have a happy customer is to make them an ambassador or make sure they have the tools and resources to become an ambassador. Can you talk a little bit more about this? What is the difference between a happy customer versus and ambassador?

Gaurav:I think the difference is an ambassador is a happy customer, but then, an ambassador also has experienced parts of the product and parts of the marketing journey that makes it easier for them to share about their happiness. So, an ambassador is a vocal happy customer and they are vocal because they have been given the right tools and the right incentives to keep sharing about the product.

Felix:Can you give some examples of tools that you should have prepared for your happy customer so that you can create more ambassadors?

Gaurav:So, for example, we use a referral tool called Talkable. So every time a customer is happy or when we see that they’re engaging with the product, we would send them an email or there would be a push notification asking them to share about the product with a potential buyer and if the potential buyer buys the device, then they get a $75 Amazon coupon, which is their incentive to refer. Now, even without Talkable, people would still talk about the product, but at least with Talkable there’s more incentive to continue to talk and then, share the word out there because people can make some money back.

Felix:正确的。How do you know if they are a happy customer so that you can engage them to enable them to become an ambassador?

Gaurav:That’s a very, very interesting question and that’s also a very difficult one to predict. A lot would just depend on how sophisticated your analytics engine is. You can look at things like … there are some easy metrics for you to look at, for example, you can ask people to fill in an NPS survey, or a Net Promoter Score survey, after maybe a month of using it and from the Net Promoter Score level you would know if they’re happy or they’re not. That’s a direct way. The indirect way would be, say you are tracking what are they doing in the app and are they replacing the filters and are they using the device.

然后,基于你能做出预测model to say that any person who tries to connect the device with the internet, because it’s an IoT connected device, anyone who’s trying to do that and is successful at doing it within the first 24 hours of unboxing the device, there’s a good chance that they are going to be a happy customer. So, you’ll have to build a somewhat sophisticated predictive model to predict if they’re going to be happy or not.

Felix:Is it harmful to assume that everyone is a happy customer and then just offer the opportunity for them to be an ambassador and hopefully the ones that are actually happy will take advantage of it?

Gaurav:Actually, that’s a really good idea, for an ambassador program you can do that, but if you are also thinking of a review program, then that can be dangerous. That’s one. Second, you can of course expose everyone to the ambassador program and the ones who are not happy would not interact with it, but maybe a better opportunity is only expose people to become ambassadors when they’re really happy because at the core of all of this, it’s important for people to be happy first, and then become ambassadors. So, the way in which we do things is if they’re happy, let’s make them ambassador. If they’re not happy, let’s make them happy first, and then make them ambassador.

Felix:明白了。So, there’s a stair-step approach that makes sense. There’s an order of operations that makes sense. Now, you mentioned that one of your focus’s is around performance marketing. What kind of performance marketing are you guys doing these days?

Gaurav:我们有一个健康的网络,这是社会、display, search and then, we also do a lot of offline, actually, that spans across broadcasts, radio, a little bit of TV, mailers, print, and so on.

Felix:And when it comes to offline, this is definitely a channel that I think a lot of listeners either don’t get to in their business or haven’t gotten to yet, what made you guys decide to also include offline into your marketing mix?

Gaurav:这是一个很好的问题。我认为的方式look at that is run experiments across different channels, and then, see if you’re getting incremental sales because of the channel. And, if you’re getting incremental sales because of the channel, then you should include it in your marketing mix. So that’s how we tested it. We tested offline a few times across different modalities and we saw that offline was driving sales, and because offline was driving sales, it made perfect sense for us to consider scaling offline as well.

There’s no hard-and-fast answer. Technically you could say that offline is good for brand building, but that’s not true for all offline. Yes, TV and billboards and out-of-home, they have more of a brand purpose than a conversion purpose, but broadcasts, radio, print, they can also drive conversions if you’re targeting the right audience and targeting the right audience with the right message.

Felix:正确的。And, it sounds like with offline, is it harder to track than? I guess, how do you make sure that you are able to track this performance?

Gaurav:Offline is certainly very hard to track. The way we do this, is we would give out coupon codes on different … so, if we’re testing a print campaign, we would have a specific coupon code in that print campaign that we can use to, later on, measure what the success has been. But, that’s one of the biggest problems with offline ’cause you cannot measure the true impact, but by actually giving out discount codes such that people would end up using the discount code rather than buying it without any discount, you will at least be able to measure what is the channel doing for you.

Felix:正确的。And, when I think of offline, you’re talking about things like out-of-home and radio and even podcast, it sounds a lot more expensive and more of a commitment to test. Are there ways that you’ve been able to test at a smaller scale, maybe not as small as online marketing, but in a way where it’s more budget-friendly for smaller companies?

Gaurav:So, I wouldn’t say it’s necessarily more expensive because the CPM’s on some of these offlines can be lower, but it is definitely a higher commitment. And the way to do that, perhaps is to use a lot of … to buy remnant. So, any sort of traditional media: TV, radio, print, they also have remnant media purchases. Remnant, meaning, that it’s not primetime or it’s not something that’s been already paid for. So we ran a lot of remnant media on print to first see if it’s going to work for us or not, and it worked. And, because it worked, we were able to then scale it because remnant media’s available at a much discounted price.

Now, you don’t get to control which page of the newspaper your ad is going to come in, but you are still able to get a directional readout if this channel is going to do anything for you or not. So, one approach is to try remnant media, the other one is some of the other offlines, right? For example, direct mailers, shared mailers, those can be done in low-budget ways because your experiment’s gonna be small. The only problem is you may not get enough data to get an accurate readout, which I think is a challenge. So the way you look at that is: was the channel performing really, really well? If it did really, really well, then you would see that in the data. So-

Felix:It would be clear. It would be clear. It would be the clear winner. Right.

Gaurav:Absolutely. If the channel also didn’t perform at all, which is really, really bad, and that would also be a clear data point on a clear understanding. So then the only situation where you are stuck is you don’t vary, where the performance was comparably okay, it wasn’t best, it wasn’t forced, and then you can decide if you want to keep testing it, or you can decided to drop it.

But my suggestion would be run small tests and find channels where you are seeing really, really good performance, and then start scaling them first. And then come back to other channels that were lukewarm, and that they did not do a lot, and then test them out again and see if they are going to drive any change.

Felix:正确的。

Gaurav:最后一个,你需要做的就是与other small companies that you may know, and then together you guys can do an experiment, for example you can do a billboard experiment, or sorry, not a billboard, you can do a direct mill experiment where you can bring in three companies, and you can pull in your resources together to test it out. Although that’s very challenging to find three companies at similar stages who are willing to go after the same audience as you, so oftentimes I found remnant and small experiments to be the way to scale these channels.

Felix:明白了。And based on your experience, can you give an idea of what budget we’re talking about, if you’re just looking to test this out?

Gaurav:For example, you can test out something like Wall Street Journal with remnant add, with something as low as maybe $3,000.

Felix:This is Wall Street Journal, what is this, the website, or … ?

Gaurav:No, this is the print.

Felix:They sell remnant ads as well, just their print?

Gaurav:是的, they sell remnant ads through print, and then instead of testing out nationwide audience, you can say, I want to test out Wall Street Journal east coast, or Wall Street Journal west coast, and you can run a test like that for $2,000 to $3,000.

Felix:Wow.

Gaurav:But then you know that if that works, then, okay, the CPA here is good, the costing is not bad, we can do something here, let’s see what we can do to sale it.

Felix:That’s definitely reachable, a lot more achievable than I would imagine getting ad into Wall Street Journal would actually cost.

Gaurav:That’s also remnant, if you decided without remnant, you’ll end up paying 15k to 20k.

Felix:明白了。Now what about the direct mailers? What are some starting budgets there?

Gaurav:I would say you would need $20,000 to test that out.

Felix:Okay, so a little bit more expensive to test that. Okay, so you also mentioned that there’s a lot of personalized marketing that happens during the customer’s journey with you, can you talk a little bit about this, say how personalized are you, what you make your marketing for a given customer?

Gaurav:So what we do is we use … We track the UTM parameters really carefully, and that allows us to know when someone is a … When if you’re a buyer, and is your interest because you’re interested in allergies, or if you’re interested because you’re asthma, or if you’re interested because you’re living in a mega city by a freeway.

所以the first thing that we do is understanding the different demographics of why people are coming in, and we do that based on the ads that we run, we have a UTM parameter on those ads, telling us who was the direct audience for that ad. So based on that we know why are those people interested in Molekule. So when they come in they see different versions of landing pages, and they see different versions of follow up ads, which are re-targeting ads.

The next one is personalization based on where you are in the funnel, so if you’re very close to buying, we are trying to deploy a system where it would show you ads that are very relevant to the purchase mindset, which is, “Hey, we are giving you an X dollar discount, or we are Y dollar shipping discount, or our warranty period is two years and our returns are free, and you have 60 day return policy.” Whereas say, if you are in the consideration phase of the funnel, then we are going to show you a testimonial and we are going to show you why this technology can change your life, and so on.

Felix:What will you use to try to track, this sounds like a lot of data coming in, a lot of different variables that could easily overwhelm someone if they obviously don’t have the right tools, what do you recommend using?

Gaurav:It is, you’re absolutely right. If you were to deploy this in full scale, these are very expensive tools, like Telium, or Litix, or Segment, to be able to do that, but however, you can do a version of this based on being very diligent about how you are placing cookies and defining events on your website. For example, you can define Facebook events on your website that measure [inaudible] abandonment, that measure lead capture, that measure the different pages that person has been to a website, and based on that, you know that, okay, this is a person who has been to five pages on our website, and he has filled in the lead capture, so I should re-target that person differently.

Similarly, this person was come to our website and they just bounced within five seconds, I don’t even want to recontact that person, because that person is not market. So you can do a lot of interesting things just based on the pixels that you have deployed on your website, and start postulizing based on the place in the funnel. If you’re not dong that, that should be a low hanging fruit to do very immediately.

Felix:You would recommend personalizing that over personalizing the ads and landing page customization?

Gaurav:The first one is definitely the place in the funnel, because that will also define how much a frequency should be. For example, if someone has abandoned cart, you need to hit them fairly hard so then you can get them to buy, right? So I think, similarly, if someone is not attracted with your website at all, you don’t need to spend any returning money on them.

So you could say that, I would suggest that the first thing to personalize is for the place in the funnel, and then what you should do is, based on the ad, based on the targeting group, personalize the ad and the landing page.

Felix:Got it, and what about that, do you guys just have … Do you create the landing page just by hand, or are there tools to scale up the personalization between the ad and the landing page?

Gaurav:所以there are definitely tools that allow you to do this. We use this tool called Optimizely, which just is an AB testing tool, but which also allows us to create different URL’s wherever people can land. In the future we’ll be evaluating a bunch of different tools that can allow us to do it, like Unbounce was a good tool that would have allowed us to create multiple landing pages, and then if you are a more mature organization, of course you have things like Hop Spot and Sales Force that allow you to build more and more landing pages, to a centralized tool, but I feel like that might not be necessary ina consumer company.

So build as much as you can by your hand, and then upload the pages on Optimizely or Unbounce to service different people differently.

Felix:明白了。So you’ve obviously seen the growth of the company from the very beginning to now, I think you said 75 employees. What is the first place that you find that most companies that start off and are ready to start scaling, how do they make the decision how they should be investing in, who they should be hiring, where they should be investing into the business?

Gaurav:The way I have done this is first [inaudible] your bread and butter, so for example, find one channel where you can at least get some sales. People want to conquer everything very quickly, but the way I look at this is secure bread and butter, so find one channel, learn how to sell ten units per day first. If you can learn how to sell ten units per day, that’s great, because now we have a stable engine, and we have a stable growth engine that’s considered [inaudible] revenue.

如果你有了,那么现在你可以花10% extra budget to try and find another channel. And then when you find another channel that works, then you can also scale it, and now your [inaudible] looks like … Landing doesn’t look like ten units per day, it looks more like 15 units per day, or 20 units per day. So now you have a bigger pool of resource available to finance the basics of your business, and you have a bread and butter that’s growing.

And a growing brad and butter, can I just use the word bread and butter because everyone gets it? A growing bread and butter part of your business allows you to finance other growth opportunities. But trying to go off the growth not understanding what the core of your business is and what the core revenue driver is, is going to be difficult, because at some point, growth would plateau, but your growth would not sustain, because you are not able to add more customers anymore.

Felix:What was that for you guys? What was your bread and butter in terms of the sales channel for you guys?

Gaurav:It was Facebook and search.

Felix:明白了。So when you are starting to expand out, so you figured out how to get ten units a day, whatever the threshold is for that, for first Facebook or search, and then when you start looking at other channels, I’m assuming you have to start staffing against the established bread and butter channels. How do you make sure that you’re able to scale safely? ’Cause I think a lot of times, people will be successful on one channel, and then focus on another and they kind of ignore what made them successful because they weren’t able to put a system or staff against it, and it kind of falls apart. How did you guys make sure to avoid that outcome?

Gaurav:We would begin with consultants first. So say Facebook began in house, Facebook was an exception, but say Brent. It began with consultants first, and we said, “We’ll spend a small amount of money, see if it works. If it works, great, wow, okay it worked.” So now we want to scale it, so we’ll keep scaling it with the consultant and focus our in house resource as much as possible on also holding our bread and butter.

Once Brent became to a size where this is doing really well, I hired someone to basically come in and start oiling the [inaudible] killing it. So there are only two ways in which you can do it. You have to put together a ton of processes and systems to ensure that the things that you deploy don’t break, that’s one, and second, bread and butter is as important as growth, so you cannot neglect bread and butter, because if the bread and butter and gone, then growth is useless.

Felix:I see, so you like to keep the bread and butter in house, and once you are looking to experiment, the [inaudible] that you took was to hire consultants to do the help with the experimentation, and then once they’re able to establish a foothold, then you bring that in house.

Gaurav:Absolutely, because if you are … Think about it this way, right, if you hire a resource in house to run an experiment, and if it doesn’t work, then there’s a problem, because now you have to fire that person, because the experiment that that person was running doesn’t work anymore, so it’s always easier to work with contractors and consultants first, see if that’s going to scale, and if that’s going to scale, then you can predict substantial revenue against that, and then bring it in house.

Felix:Hm. So I think you probably see this all the time and you might have experienced it too, where if you are a business, if you’re successful or not, you are kind of being circled by a lot of vultures that try to come in and say, “Hey we can help broaden your Facebook ads, or run whatever.” What do you recommend as a way to protect yourself from the … how do you determine who is a good consultant or contractor to hire, especially when it’s an experiment, meaning you probably don’t know much about it? How do you make sure you have the right hire in terms of hiring consultants?

Gaurav:That’s always a trap. That’s always a tricky one to explore or navigate, but these are the things I do. Even before … Say someone reaches out to me and say, “I can test this out for you,” instead of saying yes, I ask them to propose three scenarios. Present me your best guess scenario, present me your worst case scenario, present your most realistic scenario. Because I don’t want to do this based on your best case that you are promising me, I also need to understand what do you think is the worst case possible.

So knowing across different scenarios helps me understand what’s the lowest possible situation, and then what stand back that that lowest possible situation will have on me. That’s one.

The second I always ask for references. It’s also great to ask for references, because then you get connected to other companies that are non competing with you, but that are doing similar things, and you learn a lot from expanding your peer network, so I always ask for two or three references whom I can talk about that experience working with the particular contractor or consultant.

第三,我还读了很多abo血型ut what that consultant is trying to do, and then challenge them on their first principles. For example, if someone comes in and says, the click through rate on my ad is going to be 5%, then you ask them to basically lay out the plan and break it down into building blocks, and then you challenge them on building blocks, You can say, “Why do you think your click through rate is going to be 5% when my bread and butter click through rate is 1%?” And that way you can find bullshit, and often I’ve found that as a very interesting way, because they might come and say that, “With our process you can acquire customers for half the cost.” So then you should ask them, “Okay, walk me through your process, what’s your audience, what’s your CPM, what’s your click through rate, what’s your conversion rate,” and then you can call out bullshit when you see it. Oftentimes in your gut you will know that this is going to be BS.

Felix:明白了。So obviously, lots of things going on in terms of your marketing mix, you guys are running a lot of experiments. How do you spend your days to make sure that you’re able to stay on top of it as the person that is the VP of that entire department?

Gaurav:So I … To answer your question, I won’t say I am doing everything right, but what I do is I try and watch the metrics very, very closely, so I watch CPA, I watch net sales every day, I look at conversion rate on the website, I look at traffic on our website, I look at channel split, and then that allows me to understand how are we doing and which are the channels that are struggling, which are the channels that are not struggling. And then we have weekly meetings set up so every Wednesday and Thursday we have interdepartmental meetings just to look at different channels, just to look at performance, just to look at website conversion rate, just to look at AB testing, and so on.

So I would say a lot of team meetings, but dedicated slots, so like, Wednesday and Thursdays are good for team meetings on these topics, and then I have … We use a tool called Grow. Grow already connects to Facebook, Google, Shopify and so on, so you can create dynamic reports on Grow, and I use Grow quite a bit, just to understand, just to keep a pulse on what’s happening on a daily basis.

Felix:明白了。Now where do you wanna see the business go over the nest year? Where do you wanna see Molekule focus its attention?

Gaurav:I think we are at that stage right now where we are rapidly expanding, so the goal next year is going to be how do we hit a 2.5x growth number. So if we are doing a few million dollars this year, how do we expand that and how do we double that?

Felix:明白了。Thank you so much for your time, Gaurav, so Molekule.com, M-O-L-E-K-U-L-E.com as a website, thank you so much for coming on and sharing all of that.

Gaurav:The pleasure was all mine Felix, thank you so much, I hope this was helpful to your audience.

Felix:Thanks for tuning in to another episode of Shopify Masters, the e-commerce podcast for ambitious entrepreneurs powered by Shopify. To get your exclusive 30 day extended trial, visit shopify.com/masters.