Cracking the Code of Data Mastery with Feifan Wang of SourceMedium

[00:01] all right folks we’re back for the double header I believe the game is going on right now so we might be on a short clock because John has some money and he wants to go see it work lot of money but but we have one of the smartest humans on the planet in data why wouldn’t we do a potty so Fay go

[00:19] ahead introduce yourself let us know what’s up hey thank you Roba thank you John and it’s an honor to be on this podcast my name is Fay I’m the founder of source medium sour medium.com you you heard it you know if you ever use GA you probably know what it is uh we do data infrastructure for Brands we provide

[00:38] them with very high quality data to make decisions and uh everybody has Source medium codes oh yeah so tell me about your background take us through how Source medium came to be how you got your start in e-commerce and data thank you yeah so I actually started my career uh I was a business major coming out of

[00:59] college failed all my computer science classes where’ you go school University of Illinois oh Midwest boy I mean you guys there’s so much that’s exactly right but there’s so much alignment with like the Bloomington like everything the stadiums are called the same it’s so interesting that’s a really interesting campus actually I pared there a few times it’s

[01:18] it’s a very interesting campus okay so you’re like partying then yeah oh dude we were a party school that’s I rage my face off in mind yeah can see different life old life old life yeah yeah yeah a few a few rings around the tree trunk there theyy you know I I was not one to shy away from a good time but that

[01:37] doesn’t imped me from loving why D is so important thank you thank you yeah so I know you know that better than anyone to be honest uh yeah so I actually started my career at Google I was a analytical lead slac account strategist whatever the title is now um you know for midmarket Business Industrial sector so I looked at you know a lot of Google ads

[01:59] data uh you know for companies that were spending a lot of money on Google but wasting a third of it generally speaking so that’s where I had my first experience with data and understanding all of the opportunities that is within the ability to understand and look at data 100% um and then after that I actually went to the very first coding

[02:24] boot camp in San Francisco called uh Dev boot camp let’s go what was the language what were you coding Ruby on Rails wow it was uh what year was this this would have been like TW I I I left Google after a year and a half which was a little bit regrettable that was before the stock split sometimes you get the bear

[02:47] sometimes the bear gets you it is what it is exactly but this would have been in 20 probably uh 13 or 14ish back in those years yeah so learn how to code in 12 weeks supposedly and uh became technical but my so you were not originally technical no you weren’t nerdy at all or nothing business student yeah I could barely do sequel to

[03:12] be honest with you yeah wow yeah yeah selftaught dude those those are the G’s yeah yeah I can only self te uh school never work I’m the same way I would never go to class I would just study on my own I’m the ex like not to derail things but I’m the exact same way where like everything I went to school for

[03:29] [ __ ] econom like I wanted to be an investment banker and now I’m running marketing so it’s you’re buying calls don’t even don’t even get I lost on that back end of that didn’t take any gme action but no it was the same because I got in a photography I got in the AWS bills I got in a super nerdy

[03:45] stuff but to your point it was just all selftaught and I think there’s there’s a certain magic to that because you you learn what you need to learn yeah yeah and you go through all like the mental gymnastics to figure [ __ ] out 100% man I mean this that’s how you learn right it’s exactly how you learn and like you

[04:02] know again not to derail things but I think this is the best time to be extraordinary and the worst time to be average so true like if you you want to you know back in the day you could just go to your job do the factory get the house the the kid the wife the the car yeah now it’s like man dude you can just

[04:18] go on YouTube and pretty much get to a really high level of something or like a maven course or whatever so and sorry back to you f i got no kids no cars and no wife but incredible taste pants yeah thank you yeah my social anxiety pants uh yeah anyway so you know I didn’t know anything about Ecom you know

[04:39] don’t ask me what conversion rate was but uh my best friend um started working with this mattress company you know and they were scaling and they were scaling very fast and uh they needed somebody that could work with an Israeli engineering team and for some reason I guess um I’m able to work with them for whatever reason not easy not an easy

[05:06] task there’s a joke about his raes there’s two israelies and four opinions yeah it’s so might be eight it might be exctly exactly but they do respect competency which I 100% very meritocratic 100% yeah so next thing you know I I became VP uh of product and analytics uh over there while we were essentially scaling from seven to nine

[05:27] figures in run rate in 2018 we went from 7 to 9 one year hold on so you go from non-technical to VP yeah that’s a [ __ ] acceler this little mattress company that many of you might know yeah yeah we just exited for a billion dollars billion exit cool flew under the radar too like how did that F that’s

[05:49] incredible yeah yeah yeah and this company’s name was and we’re not allow about it look at that right who want a th yeah it’s called resident home that’s incredible dude thank you thank you yeah so in so you literally just did the boot camp and then just started Skilling up no I did a bunch of other things in

[06:04] between and I won’t waste your yeah we don’t have to get into it but that’s still incredible though so the boot camp laid the foundation and then that sprung board you into well the foundation was actually working at a Innovation agency where we were essentially working with Fortune 500s to help them validate new ideas using the Lean Startup method if

[06:26] anybody still knows what that is uh Eric Reese right Eric Reese was on our board so uh you know the one thing that I could claim is this same best friend Scott McLoud shout out he’s a he’s the G he yeah or not not periphery like I know of him but not him personally you got to bring him on the P

[06:47] next you’re he’s introverted these days after the exit you know do that I’m actually visiting I’m visiting him in uh Seattle he he got a spot in Seattle so Miami I’m going next week no he’s going to still stay got the Seattle pad and the Miami oh say you do the co but he’s he’s I don’t I don’t want to I

[07:10] don’t he’s a private person I don’t want to uh put up his story yeah yeah to one4 yeah uh so yeah so we actually met through this Innovation agency where you know somebody like a product manager that went to Stanford or whatever that works at Adobe this is the literal project he came to us and said I want to

[07:32] make an infographics maker this is going to be the future of adobe you know using my dchool experiences or whatever and we actually would go out interview people understand the demand so the very first project that we did together was we ran ads uh that prompted people to give us what they wanted to create from an asset

[07:55] generation perspective okay and Scott would actually make them in SK if anybody knows what that is the kids won’t get it but we get it and then we would zapier to Trello to Google Sheets fourways to heaven and come back send it back to them the next day so you know we made it essentially asset generation AI

[08:15] Bop do things that don’t scale yeah yeah so from there we uh you know discovered upon something in terms of social media image generation and that eventually turned into Adobe spark which is their most popular mobile app [ __ ] yeah no credit given but that’s okay but I’m going to say it wow you know yeah Bad

[08:37] Boys moving silence um so anyway so after that um so I’ve done we’ve we did a lot of these kind of projects right 0o to one 0 to one 0 to1 and that kind of got my muscle going around that um and he developed a particular um expertise around growth you know I used to call him the uh the dark wizard it of growth

[09:01] and uh so he started working with the co-founders at resident he’s considered a co-founder there and uh he you know he needed someone that could actually understood technology so I was just looking for honestly like something to do um so I job just over broke where he just like hey let me do something baby cuz I was traveling out of a duffel bag

[09:25] for like two years I didn’t have any money left so I was down to do anything anything uh but you know I obviously I I uh I got lucky so in 2018 I joined in 2017 2018 we uh scaled to n figures in run rating one year love that we were the fastest growing online mattress we till America you came in and what were

[09:45] you charged with so I was charged with uh initiating our AB testing program okay with a uh AB testing agency and I didn’t know anything about you know St got sick or you know really anything at all but um the business was growing like wildfire and we never spent a penny at that time on awareness campaigns while our competitors like

[10:12] caspers of the world were essentially like subway ads that’s Telltale sign I’m like out of home have you yeah so um you know so I had the opportunity to to essentially learn just because I could actually Implement yeah what was being done because one thing about Resident is that we built our own Shopify we built our own source

[10:37] everything was custom there everything was built in house and uh it takes a lot of technical knoow to really make anything happen right so of course like with formont like that’s the whole point of not needing that but at this beautiful human give me the shame of course man I got you all day so uh but you know if you wanted to make a landing

[10:58] page like we had to code it you know and really understand the intricacies around the performance hits and everything else I mean I had to you know hand Implement like the L our blot out yeah of the world you know and um so anyway so but we were just scaling really really really really fast and we went uh

[11:21] multibrand Omni Channel and multi- region at the same time uh in one year so we launched in the UK we launched dreamcloud at that time so we have nectar dreamcloud aora you know um and many many other brands and we went uh Amazon and Retail like so everything was happening and of course not having good data and making decisions on bad data

[11:46] was left and death yeah there was really no room for at that level of AD yeah so at one point we were in the top five Spenders on Facebook like globally not nothing no nothing so how did this transition into Source medium Yeah so basically you know we had this uh Infamous spreadsheet Google sheet dashboard called Master Dash and it was

[12:12] all you know just whatever right Google Sheets you know how it is but of course uh you start breaking and people start accidentally hitting enter on a cell and mess up that one formul breaks whole is gone braa you did what and ref and ref on days and uh you know but also there’s there’s a lot of different things you

[12:34] know there’s like the data isn’t deep enough for you to actually understand and explain what’s happening right and you don’t trust the data and all kinds of things so we knew that we needed to upgrade our data story there and um so then all we knew was like oh looker like that’s legit that’s what legit people use so what do we need to do to have

[12:56] looker then of course like I went down this rabbit hole and next thing you know is like four vendors that’s like you know 50 to 100K each uh you know we hired a data uh agency that was like $2,000 a day jeez you know day a day yeah it’s real money yeah yeah that’s how much they cost that’s how much they cost so

[13:21] after two months we had some basic aov dashboard on looker and like nobody trusted the data nobody knows how to use looker if you don’t trust the data the data is meaningless do no and get bad data is worse than no data exactly right yeah so you know long story short like basically I just got to see how um

[13:42] expensive time consuming and felt the pain yeah and difficult it was and risky really overall you know I try to hire our first head of data from Amazon which who didn’t work out and of course like how do you reverse that back that takes time as well so how do you take all the time back yes that you have already

[14:02] wasted in terms of AD spend you know so uh but what I also do understand is what that journey is like from Seven all the way to 10 figures in sales in terms of what infest dat a proper data infrastructure can do what the investment model is like and what the value it can provide you you know and I

[14:22] think to this day uh my last Shameless plug about Resident is I do believe we are the textbook of DTC like in Harvard Business School there has been a case that follows us till this day powerful HP for scaling uh consumer Ventures so you know yeah I think that’s uh that’s pretty good proof that definitely a feather in the Hat so who

[14:46] is sour medium right for right yeah great question so after so anyways before we get to that give us like the oneliner elevator pitch of what source medium does for people that haven’t followed along so far sorry so yeah Source medium you know we do two things we help you save money yeah by essentially gaining access to

[15:09] the most advanced data infrastructure that has supported the 10 Xing journey of Brands like ult overnight ollipop element proven skin care you know I can keep going yeah and um it doesn’t take a year it doesn’t take three months it takes a couple weeks yep right and you don’t need to have a whole bi team in house you don’t need to pay for five

[15:32] Trend to snowflake a do it for you solution right for data you’re a brand you want to extract all the data from the different platforms Facebook Google analytics Google ads Tik Tok right you have a firmat Commerce has an integration integ so you’re extracting the data but then you’re transforming it you’re some engineering you’re cleaning the data

[15:52] getting rid of all the draft orders everything that’s conflating potential average order value or lifetime value and then you’re going visualize it in a looker right in a yeah so I like to say that I believe we have the highest quality data in e-commerce and I have a open Pepsi challenge for anybody that want to challenge me on that kids won’t

[16:11] get that but I did I’m ready to go uh so you know but that’s not enough you know so what do you do with that data so what we focus a lot on is really embedding with our customers gaining empathy understanding their underlying business objectives and making sure that we’re scaling with them you know so um you

[16:32] know so ultimately the way that would help you make money is the ability to make very very important decisions confidently yeah 100% now we use Source medium for our agency we literally built our whole dashboarding infrastructure off of source medium it um and it’s been a game changer the second I saw his data he came into my office introduced by

[16:55] Mike true we just had on and he pulled up this dashbo and I was my mind was blown I started my career buying my own media and this was the exact solution that I needed that I wish existed and didn’t even know it exist until he pulled out this dashboard and I was you could look at LTV from organic Facebook

[17:13] traffic you could LTV from Facebook paid traffic LTV from Google search you know branded non-branded it was wild if you used a discount code XYZ you said I heard it from you from TV based on the discount code you can see which sales are driving low quality high value customers right and you can iterate LTV based on the cohort month they were

[17:33] Acquired and so essentially Source medium we’ve used as a centralized dashboard that every team in our business can log into and see what’s going on on every single client the email marketing team can go in and see oh how much is AD spend this month right the creative team we have a nice creative performance dashboard across all channels we can see which creatives

[17:52] are performing best and profitability so finan goes in there marketing teams go in there retention acquisition and it’s just that One-Stop shop that I’m a business I don’t know how you could operate out of solely Facebook Google clay dashboards and everyone is totally siloed and nothing connects yeah yeah thank you for that I appreciate that yeah yep so who’s the ICP for

[18:14] Source medium yeah so we typically um so we have gone through a pretty major transition recently you know we started out as a dashboard company we worked with uh you know typically smaller clients but now that we’ transition into a full infrastructure company we typically work with brands that are making between 30 million up to a billion dollars in

[18:36] annual sales across other channels um you know typically the main reason people come to us is because there is a level of complexity in the business that actually justifies somewhat of a more Advanced Data infrastructure setup um so you know if you have uh Shopify Amazon right if you have some retail data and you need to be able to look at that in one UniFi

[19:01] metrics framework in one unified decision- making framework like that’s where uh we come in I love it we’re getting heckled by the peanut gallery so let’s maybe do one or two more questions John and let’s wrap it up I think the the Celtics game is on and people are getting angry yeah yeah yeah for sure yeah stics are blowing them out though

[19:19] that game but I got I got plus seven on the Maverick still never miss I mean if you had one prediction for the rest of the year where you think actually before we get into the prediction maybe there’s two questions what is the difference and this is like common sense to us but I can’t tell you the diff every time I

[19:36] talk to a new client or a new brand that’s you know a prospect like what do you have for your data and they throw at me an attribution solution what is the difference between attribution software and data and analytics software yeah I mean I think a data software really is about having visibility on numbers that you can trust to make decisions as an

[20:00] organization not just as an individualized team you know so in that sense you it’s like looking at your speedometer on your car yeah right if you can’t trust that you’re actually going 80 miles an hour you could die right right so that’s what I typically say is you know if you’re going 20 miles an hour your numbers are 20% off it’s

[20:22] really not that big of a deal stick with what you’re doing but if you’re going 150 20% off either direction could be very dangerous uh but you know I think attribution is one of those things that is very useful from a data point perspective but it’s very very specific to a uh function so understanding where orders are coming from understanding um

[20:48] you know how you’re acquiring users and and growing your business is very important but that’s not the whole picture right how do you how can you make sure that you’re allocating your inventory correctly across Amazon and Shopify right how can you actually understand that at a blended basis when you have a big Good Morning America appearance that your Rass is actually

[21:08] increasing exactly right you know and those are not the kind of things that an attribution software could solve um but you know obviously attribution has a lot of using and of itself yeah right yeah I mean with the data you could just cut the proper analytics solution your proper name convention we slice and dice the data in ways like I remember there

[21:28] was for some reason Blended CAC just skyrocketed on this one huge cent of ours and we had to look back at all the data thank God we had proper naming conventions we could slice and dice to that and we realized we had a major shift in spend towards Advantage Plus shopping campaigns a frequency skyrocketed and so return on ad spend

[21:47] and attribution looked great but because we had that stacked bar chart on our spend allocation on the campaign tactic we identified as soon as we went from like 25% ASC to 50 % ASC that’s when Blended CAC went up attribution doesn’t do that right but if you have that that was our aha moment where it was like

[22:07] this is so useful I couldn’t have think of another way that we would have identified that without all the different views that we built that’s why we do what we do yeah another example that I could bring is with uh ults overnight this case study is actually live on our website you know they probably have one of the biggest

[22:22] subscription programs that I’ve ever seen yep and they had a big introductory skew that was uh you know a single flavor and they needed to understand but there was signals that uh Subs new new subscribers prefer more of a sample pack right so they needed to make sure that they had the data points that they could trust to to to be able to transition

[22:46] that introductory skew over to more of a sample pack and like that’s where the data really becomes uh you know that that’s where it really matters because you know if you make the wrong choice there uh you know you could tank the business very quickly it’s business decisions Beyond just the ad platforms it’s a bar absolutely it’s that’s two

[23:04] bars now oh topping bars yeah we just came up with two on the all right F they’re going to kick us out because I’m not buying the $2 million baseball card but hey maybe next year Fay how can people sign up for Source medium where can they follow you this time’s yours my friend yeah yeah sourc medium.com and uh follow me on Twitter I

[23:26] guess I’m not really on social we got to get him we got smart people out there and no one knows where to find them so they got to get active get the word out there somebody who is active on Twitter is my man Jon Snow Jon Snow times yours Dr Jonathan snow on Twitter follow me on LinkedIn as well and I got

[23:46] the Weekly Newsletter now just weekly Insight talking about great folks like social media thank thanate that all right folks we’re fir Out commerce.com book a demo we also have a great newsletter goes out every Monday called the geomet your growth go grab it right on our website you can subscribe there F dude thank you for coming on you crushed

[24:04] it you’re the you’re the second seat for you’re just the best all right folks that’s all we got for you today in New York City we’re going to go hit the town so thanks for join in bye