BFCM Retention Testing with Fil Pejic

Retention AI Challenge

0:00So, I challenged you guys as like retention marketers, is how do I take it to the next level? How do I use AI to just create it? You get that context, but then you should go off and test six different things in your key points in your funnel. And make sure that you can go back to your to your founder, or if you're the founder, you can say, "Hey, we know that this is what makes us the most money." And that's the mentality that I want you guys to make. So, a control could be like a pool you have today, running and then you're testing against that. But a holdout is like you run a test with nothing against nothing.

Engagement Survey Hacks

0:27One like hack that we do with all of our brands is just try to keep top of mind. Max had that really cool survey example, but every month we try to like spike like sprinkle in a few opportunities for customers to actually engage with the brand. It could be like a really simple like, "Hey, what do you think about this? Like the product that just sold out. What color should it come in next?" Super simple. Great performance marketing tactics and strategies. A lot of the stuff Max talked about, and we implement similar things like that. And just trying to make sure we go over to be on some of our clients and make sure that creatives are always super super disruptive.

Networking And Community

1:10before we start, my biggest secret, my alpha that I'll give you guys just to start, I want everyone to put their hands up. Put your hands up. Come on, guys. Everybody, high five people next to you. High Give them a high five. If you can, I'll see you. so, this is a community event. I highly recommend everyone in the room get a contact and meet someone next to you. Make a little Whatsapper. So, most of my alpha, most insight that I get from being a marketer, is talking to people in this room, talking to similar people who are obsessed. Jimmy and I were exchanging like the after the event about deliverability and content and future copyrighted. That's not the stuff you get in everyday. It's stuff you get in events like this. And in little group chats that we'll be up, so we'll talk to you. So, highly recommend involved.

Ecom Trends Retention Wins

1:55what I'm seeing with eight and nine-figure brands, I literally refreshed the slide this morning just so you guys can see on the graph. This is an eight figure brand we work with. Last 30 days, new customer revenue coming out of Shopify down 3%, returning customer rate up 46%. This is an eight figure brand we work with. new customer revenue got virgin sent here over year, returning customer rate up 16%. Who's seeing a similar trend right now with new customer revenue?

2:23A lot of folks. A lot of them might have bought something else. That's totally fine. this is a a trend we're seeing across email marketing across CommuniGator. If you can make card abandonment products to acquire customers, but it's becoming more challenging, it's becoming more complex, there's all sorts of other headwinds coming along. the reason we're having this talk today, the reason I came is to help you guys figure out on the retention side, you're going to struggle with your your brands, whether you write for someone else's.

2:49It's super important that the new customer numbers obviously going up. But it's also important that if you try to carry this trend out customer revenue automation the eight and nine figure brands that we work with today, there's essentially three very high-level kind of things that they're seeing. In these three places, I would make sure that your brand is dialed in up, and if they're not, if they're not, you need to kind of take a step back and think those strategies before you tactics. the first one is strong releases unlocking new levels. I have strong underline here. I'm seeing a lot of brands kind of seeking like product led growth, but not creating kind of being fully different or being fully unique innovation side of a brand. so if you do promotions and offer adjustments, this is where brands are having huge wins right now is just figuring out what's our promotional payments, what's our promos throughout the year, what's our offer top of funnel, what's our offer later in the funnel, what's

3:40our offer boomerang. That's what That's where brands are winning big time. And the third piece is obviously on retention, and that's where our type of our brand agency comes in. this is a meme. I love memes. if you guys haven't used memes in your emails, I highly recommend you give it a try. This is obviously relevant during during

BFCM Testing Roadmap

3:58break New York. But BSC is 6 months away. And here is going by, we're already halfway through. And kind of what I'm going to talk to you guys today is right now is the time when you should be setting up all the tasks and try to figure out what are we going to do 6 months from now. You should not come into October thinking, oh what should our pop-up be? What should our offer be? That's the question you should be thinking about right in this moment. And that is literally the presentation of everything I'm going to say right now. So, over the next 20-30 minutes, I'm hoping that you guys can take at least seven levers, write them down, and then go back to your teams and say, hey guys, these are the seven most important things. You need to sit down and make sure that we have the right tasks, the right strategy, and the tactics in place that we're testing it today, so that a month from now we get some initial results, 2

4:41months from now we get a bunch of results, and 3 months from now we're really clear on, like, hey, this is what we should be doing for BSC going into time cross that season. Okay.

Lift Not Bias Testing

4:51The first one is solving for lift, not bias. This is the most important slide I think that you'll get in my full brain today. Man, this is the place that I think probably 95% of brands I work with go wrong. And it's solving for lift, not bias. Brands often come to me with ideas. They ask me, hey, what's your take on this? What do you think about that? What do you think about all this? Should I do this with the blog?

5:17I can generally give a pretty good idea, but on the back end, there needs to be testing being done. All right? We have way too much data in e-commerce. And for us not to be actually testing things, and actually getting real, like, concrete, and not just revenue, click, open data, we need to get concrete profit data that proves out any test that we're doing is actually a good idea. So, should it be a text-based or an image-based email or first email for our welcome flow? That's a super important test that everybody should be running through. What is the messaging my first local email? That is a really good test. You shouldn't be looking to the guy next to you saying, "Okay, what are you doing?" You're asking a girl, "What works for you?" You should get that context, but then you should go off and test six different things in your key points in your funnel. And make sure that you can go back to your to your founder, or if you're the founder, you

6:03can say, "Hey, we know that this is what makes us the most money." And that's the mentality I want you guys to take across my whole presentation today. and these things are not actually that complicated to do. I know it sounds kind of in the world of meta, like testing is very complicated because of meta pixel and things get really weird really quickly. You know, you have KB tests and holdout tests. these are relatively simple, whether you're using Adobe Sensei or another ESP, relatively simple to set up. And we have our best friend Claude or ChatGPT, whatever you use. This is literally a prompt that I have my team use and it works. literally all we do is when a holdout test are put in different parts of the account, we take though that email list, we dump them in Claude and say, "Hey, here's a here's two email lists. Here's my Shopify. Here's my Shopify data for the the duration of this test. Tell me which of these two won."

6:52All right, that's super easy. for all the Excel nerds or you know it's not that complicated, but for a lot of folks, that could be daunting. Just let AI do it for you. Do you have a special? >> Yeah, what is a holdout test? >> So, a holdout test is essentially when you run a test, you hold 10% of the audience out of doing that. So, let's say you want to add one more email to the end of your flow, and then you say, "Hey, let's let's let's let's send that extra email, but for 10% of the audience, let's not send it." So, and then a month later, you pull that list of 10% who didn't get the email, you pull the 90% who did, and then you do the same Claude analysis, and Claude will tell you, "Hey, your audience you didn't send it to actually like there was no difference. Like they didn't spend more or less money, and you didn't make more or less profit." Right? And like Joel, sorry.

7:33>> Think of it as a control. >> Exactly. Yeah, technically it's a control, but the control might be you might be running something else. So, a control could be like what you have today running and then you're testing against that where a holdout is like you run a test with nothing against nothing. You basically try to figure out what happens if if if if I didn't have this at all. So, there's two different types of tests, but functionally they're kind of in theory like similar. and these are kind of what you should be doing ESPs are doing their best to give you guys context on their revenue, clicks, open, things like that, but they're not going to give you the like profit level of data you guys need really smart decisions as a brand. And as I talk today, every single thing that I say, everything personally we talked to. This is where your brain should go back to this number one of like, okay, like how do I test this and prove whether

8:22it's right or wrong? And then the second question you should ask is is it even worth testing? Like if this wins, like how much revenue will actually bring in for our business? Cuz a lot of like tactics are great, but if it takes a little bit of time to implement them, you got to always ask yourself, is it worth it? Cuz you only have $500 to play with. Okay, I want I want everybody to

Popup Testing Playbook

8:41raise their hand if they've run a pop-up test in the last 6 months. Okay, what about a pop-up Who's run five pop-up tests in the last 6 months? Three Who's run 10 pop-up tests in the last 6 months? Nobody. I guess me. Yeah. Yeah, me. This is a This is a huge miss, in my opinion. Pop-ups are on everyone's website. Max Max already talked about it, so I won't explain to you guys why it's important, but you guys are Sorry, you guys here are spending tens of billions of dollars, maybe spending millions or so like what's $1,000? This is traffic you are sending to your website. If you don't have an idea what the best profit per visitor outcome is of your different pop-ups, you are literally holding back your paid marketing team and you are holding back the growth of your business.

9:26And I highly highly encourage you guys to think beyond retention for a second and think as founders and ask yourselves, is our pop-up holding us back, yes no? How many people in this room could tell me if no pop-up versus a pop-up brings in more profit for the business or does profit or does pop-up A, B, C, or D bring more profit for the business? I'm assuming almost nobody in this room has that number. That's really freaking important, all right? People here are like there's probably a billion dollars of fine ass banging this room. And you guys don't have like if if you're you're spending irresponsibly, that's place that spell a lot of time. these are like just as agency different things we test with our brands. you go down the line, there's a bunch of different options. question test, offer test, format testing, full screen versus standard, creative testing, tying, repeat, and then sorry, we repeat back. there's all sorts of niche tests.

10:16There's no one like path of what tests we do. We just try to evaluate and create a plan for brands of what you think these things are going to work. And we kind of modify as we go. offer pop-ups, if you're not getting 10 to 20% on your offers, you might you you probably have a lot of opportunity on the table for options. if you have no offer, 5 to 10% is like healthy. I was here in the supplement space for some weird places where offers are operating.

10:39I'm not sure if we talked about all those different types of designs. Sorry. I'm not going to bore you guys with

Email Cadence Profit Tests

10:43that. the next thing I'm going to talk about is optimizing send cadence. There's this common myth in like email marketing that you should just send more emails or if you work with an agency, they'll probably tell you, "Hey, like like the next 3 months, let's just send one more email. Like let's just see how the revenue looks." And the revenue is always going to look green and awesome. similar to what I was saying back back in in number one. when you get an agency comes to you with that or somebody on your team comes to you or you're thinking about it, your question should be how much more profit does an extra email per week bring for our business? Okay? So, there's two really simple tests that come to mind with this today that I highly suggest everybody in this room does. Add one more text-based email per week or one more image-based or whatever you want. And and run that for 2 months. You want to have a holdout that we were talking

11:27about. Depending on how big your audience is, you might want to hold out 10% or 50%. I don't know how big your brand is. And then you just want to pull your Shopify data within those 2 months and see which which audience your brand brought more profit per visit. Because if it's very tight or there's almost no incremental profit, you're probably pissing off a lot of customers. On the second one I would test, and this is one that I I see so many brands miss out on it and it really pains me is instead of just sending more non-promotional emails, send more promotional emails. During your sales, AB test or hold out test, does more promo email during promo bring us more profit?

12:01Okay? And test that and test it aggressively for your next big promotion. And then wow, 3 4 months later come back and look at that audience and see how they're behaving after that. There's a lot of like I think today what we're talking a lot about is like point in time wins, but as you guys grow up this 8 809 figure brands in the room, there's like repercussions to being very aggressive with frequency of emails, but not you don't see that until 6 9 12 months later where customers are coming to you and say ads, not business.

12:30through the emails. And that's the kind of figure you're possibly fighting with like us optimizers and didn't. but I

Pressure Test Offers

12:35want to make sure you guys are all sincere in this. the next one, this is one that I think everybody in the room needs to test right away is pressure testing your offers. Okay? We're 6 months away from BFCM. That Black Friday offer will like make your break your year. If you're not testing that right now, you're you're dropping the ball. Okay, you probably got two maybe three big promotions that you can you test with flash sales and stuff before BFCM. Use those moments, split out 10% 20% 30% of audience and just test different offers on them. And the same way I told you earlier, figure out what offer brings the most profit into is the business and and just test that for the next few sales. Because you want to figure out a lot of you guys might be doing 30% off, maybe you're doing a bundle, maybe you're doing a free gift.

13:19I don't know what it is, but right now is the moment to figure out what that is cuz we see like insane wins clients where one client we went from 40% off to buy one get one 8% off which is like function the same offer and they drove 50% more revenue on that on that offer. It was awesome and they're super happy and they crushed it. You guys You guys have a question? >> Yeah, I was just going to say how long are you running these tests for typically?

13:43>> so for for a sale test you just run it for the duration of the sale. And I'll show you. If you don't have enough data you can just you can repeat it multiple times. >> Yeah. >> Or you can like and you can just see like okay like how does kind of over come like you'll see a trend usually just in one sale. There is like a free like a recency bias so I suggest doing multiple offers even the same test multiple times cuz the first time people see something usually they're excited and then it's like okay but if we do that again and again what happens? And you should be running these throughout the year. It shouldn't just be right now cuz I'm telling you you'll see it's time. Offers get tiring, people get tired, your product mix changes as well.

14:17offer testing, testing your offers in the pop-ups. This is the next place where brands have huge wins. Every year there's a welcome offer. I'm assuming for the most part you don't need to have a welcome offer. That's totally I'm sure that's going to be a test as well. But I don't know how many of you guys can tell me like this 10% off does $20 cash back, does buy one get one free gift. Which one of those brings the most profit into the business, which one of them brings the best conversion rate, etc. Like if you guys can't answer those questions you should be running pop-up tests focusing on offers. I look their big win is post purchase win back. This is super underrated but that like three to six day period where you know that customers are like at at a cold bird or customers comes back to your business. What are you putting in front of them? Is it no offer? Is it is it cross sell? Is it a special offer if they hit 30 days, 60

15:06days? These are the kind of questions we're asking with our clients to figure out what brings the most profit back into the business and kind of where is customer lapse where they wouldn't purchase and what is the right offer at that moment to bring them back into business? Or lapsed lapsed customers. I any questions on that? No? Okay, cool. this type here. Sorry. When was the

Segmentation 80 20 Rule

15:28last time you reassessed and updated the segments you send and made to? So, I'm assuming most people here are sending to like 30 days, 60 day, and day engaged. When was the last time you can just think of this disrespectfully, but when was the last time you actually went back and thought about what that campaign segment is? Who are you sending to? and how do you like what was the last time you decided about we should throw in people who haven't checked out in the last 90 days. We should throw in people who purchased three times. When was the last time you actually took a like a look back? Most brands don't intentionally think about this all the time and that's fine. but this is your moment right now that I'm calling it out, right? Got it?

16:05When I think about segmentation, and I look at audiences, I really think about like the 80/20 law. for the most part, these are like the big kind of ways you can segment your audience. So, you have this engaging segment which splits out into like purchase frequency, You can you can you can target people off last purchase date, target people who have shopped, AOV, how engaged they are with email. You want to like brand and demographic things. Product categories of shopping age, gender, region, subscription type. And most of these buckets actually break out of like 80/10/5, which basically just means like 80% of people probably should be hit one way, 10% be hit the other, and 5% be hit the other. And I slowly I want you guys to intentionally think about for all of your flows, all of your campaigns, what is like for each of these groups, what does that mean? 10? And if you were to show like custom journey for one of these groups, do you think you could like actually shift like click rates,

16:58open rates, and profit in a direction that could like make your business a lot of money? And like the easiest example of that is basically product category. If you guys sell I'll mention the guy selling at the honey the honey gummies brand. If he had if he had the four different products, protein powder, his honey gummies, and then like collagen powder, But my honey gummies are bringing 8% of this like this revenue. it's very likely that if he's just sending generic emails about his whole brand and all those products, it's not that interesting. But if now I create custom journeys cuz I know 8% of people are kind of worried about honey gummies, if I make the journey just about that, and then leave the rest of the 20% to generic, I've now created like this like 80/20 approach of a really custom journey for most of my audience. And everybody else for now going to have this like generic kind of experience.

17:47And later on I can make it more tailored to those other folks. And you got to go down this list and just identify, okay, what's a similar example to that? It could be gender-based. It could be like whether they're on a subscription a subscription customer or not. it could be like that they should do they usually shop like BSC or You can do different kind of experiences. And I I really encourage you guys to identify like within each of these, what are those big 80% segments?

18:10And now with your campaign planning, you can go back and say like, oh yeah, we have this huge audience of like gummy shoppers, kids, huge audience of like people in Florida. Like we could do all these cool campaigns and segment all these ways. but I I would also just put a really big asterisk here. If you're not doing like over 10 million like a year in your in your store, a lot of this will just be a waste of time.

18:31You're probably making most of your money off of one demographic and one product or maybe like one category of product. And you probably just want to directly dial in that 80% and not even worry about anything else. yeah.

Flow Optimization Priorities

18:45Okay, 80/20 flow optimization. this is like my object. I think brands spend way too much time like fantasizing about new flows. They're always saying, "Hey, how do you do an email that you see with TransPharm flows?" Or "We're missing all of these flows, the birthday flow. We're missing the winback flow. We're missing all of this." Max kind of touched on that, but like if you guys can cover the worth your time stuff, so the really important stuff, your campaigns will do the rest of the work for you for the most part. some of these things that I say, I'm not saying just don't do them, but I'm saying there are times that I think stuff like this, like unless there's branding together that eight figures, obviously nine figures, that's where like stuff like this brings back an ROI time. Most of your ROI time will be right here. A lot of tactics, strategies. I won't go too in depth as far as cuz like Max was like super in depth that I think you guys

19:35would learn a lot from what he said. but these are the ones I care the most about, like local flow, all stuff Max said if you go over it, everything else I'll say in a second, nation. then I'll check out product browser banner, most purchase reaching up sales. browser if it goes under that fashion, skin care, or if you get like really big product catalogs, browser banner, that's your area. I need to be super dialed in.

19:59People window shop a lot. subscription, a lot of I've heard a lot of subscription brand guru. I feel sorry if you guys email marketing is like 10 times harder for you guys than anybody else. You're welcome. Your subscription upsell needs to be super dialed in. Cuz once they're out of stock, they're basically dead. I have post purchase for a lot of folks where with single top product stores or with four to top product stores, that's where I've been spending so much time to spend your I would get buy again. I would frost sell though.

20:26Then you know, welcome them back and check out the next steps. Okay. go big with creative. This is

Go Big With Creative AI

20:34one of my favorite slides. I want you guys to make sure like as we kind of approach this like new AI era, kind of go big with ideas. You can use some really cool stuff, like I'm seeing all these brands like they're like, "Oh yeah, you can just pick our own emails AI." And it's just like it looks like an e- like an email that was designed like five years ago. It's a great that you can do that. But like what's the actual question? What I think like everybody needs to be thinking about is like we have all this awesome AI tech. You have all this creative like this brainstorming. How do we use AI to have like the best emails in the world.

21:07Like this is a great example of like high-performing like high grade emails. It's not something that AI has missed today. Today's day it will like generate for you. But it's different. It's unique. When people see it in their box and they open it, they feel the need to click. They feel the need to read. They feel the need to engage. Whereas like everyone else is like kind of moving towards swap. I want you guys to approach, you know, like you would social media. Like you would TikTok.

21:31Like you would Instagram. Every time you get to the inbox, it's it's like keep your repose on Instagram. That's the level of like caring you need to give it because it's the exact same customer who's looking at it. It's the exact same scrolling. It's just a different need. And if you're not interesting in that inbox, other brands will be way more interesting. They'll take the same principles of TikTok hooks hooks bodies etc. They'll just take that same email approach in email and they will beat you in the inbox. You way less worried about like deliverability and all these little tactics about like oh should we make this like do this and then do that.

22:03I want you guys to think a lot more about like how do we make sure this email is so damn good that they're looking forward to opening the next. And that's the kind of mentality I want you guys to constantly have. And it can be tricky in e-commerce because a lot of people want just product. They just want the offer. They want the new arrivals. So my challenge to you guys as like retention markers is how do I take it to the next level? I don't want to use AI to just create it.

22:23Super super interesting created. And I'll just like give you guys an example. Like this is a brand we work with. They're actually out of New York. I think everyone knows about them. Sunday. Super cool brand. They got apparel. They dropped they got sales on so huge last year. Great. But these are like all examples of emails we're creating them. We're going super out of the box. Like with these like crazy animations, teasers, fun like illustrations, I'm playing up sales in like fun ways. Like fun washes. Just trying to figure out like how do we take what like AI would give us? How do we just take 10 times further? How do we empower our designers to have insane design brainstorming tools? And Jimmy's as like AI I after. I don't know if you guys want this but we're arming our designers with like every possible AI tool. We're not I haven't let go of a single desire. We're just giving them way more and expecting way more of them to create. This is kind of far more than I think you know it

23:15don't look this good. Like we're not doing our brand justice. And the same thing goes copy and the same thing goes breast center parts. I'm going to just archive them for I really feel like I can go just like a few quick rapid fires saying what's the

Text Versus Image Debate

23:28question about. >> Have you tested like I guess like a green quality for textile or you guys mentioned like the promo stuff but like >> This kind of high quality you know have you tested for copy and rewards? >> Yeah. Yeah, it's it's a mixed bag. We've tested it and we've seen it just depends on what the content is. What is the actual text based email about and what is the image based email about? Every email is like it's own post that's kind of how I look at it. And we've tested like for some clients doing a bunch of text based and a bunch of image based and we've seen like varying results.

23:59The way I see emails like I guess if your brand is always sending text emails and people kind of get tired of it it's over. So you constantly have to like reinvent yourself and constantly test. So it's like rather than like text versus image I'm trying to figure out what are the best types of text emails I can send. What are the best types of image based emails I can send. Looking at what the client has coming and then figuring out which one of those can we slot. And then within that we're testing like text versus image as well. So it's it's like it's kind of a mixed bag.

24:26I don't think there's like a winner text versus image. It really depends. Every

Image Versus Text

24:31industry is different. Every customer is different. If you're like apparel or fashion or something like image is going to win. Why? Because people are quick shots as they scroll. Will you if you run a test once with a text based email is it going to win? Probably because you have recency bias that people haven't seen it but they collect copy. I've been if you really want I don't know what your brand is but if you really want to know and this is kind of going back to slide one and slide strategic testing is super important in the sense of like you need to like coming out of this talk ask yourself like what are the tests that I want to do and what's my priority of these tests? Cuz some of them will overlap and you got to figure out where is like where is the juice kind of worth the squeeze and like what do we actually care about?

25:08Just to re- I think that question is much less important than like what is the content that we're putting A lot of people are like does that matter or are you going to be testing what that content actually is? Yeah. >> They putting any white text in your image basically was that with like deliverability and AI's and >> No This is a hot hot hot topic. I'm going to go with the hot take and say no.

25:32We're for the most part doing image based. I haven't seen like a negative impact like our brands that like in apparel or like which is like apparel industry care and beauty. That's where you get your deliverability clicks. I don't see a negative impact on our other brands. The ones where we are max kind of touched on the split up which is a lot split up engagement sucks. That's where we're doing a lot more testing on that. Like supplements are not to like brag here on supplements marketing script but nobody wants to get supplement emails. Like I don't think anybody here like like casually reads the emails from supplement brands. It's you're fighting a really hard fight and like that's where we're doing like a lot more of this like more like direct response like I almost see like supplement emails as like info product you're like direct response. It's like not even the same as everything else. Like that's where we're doing a lot of tight space stuff like try to be like I don't know if you could

26:19test emails on my subdomains. You could do a bunch of weird stuff but I don't

Deliverability Best Practices

26:22see the the the the POE because I've spoken to some of my really nerdy deliverability people. Jimmy you can you can answer out if you disagree but like the folks at Gmail are young they're not idiots. Like they know what brands want like they want brands to help like drive engagement drive revenue because ultimately they're going to monetize that. They're able like they have tools to stand in boxes. They know more or less what we're doing right now. The thing that needs to be considered is like are you guys following best practices within that? If you guys are sending super bloated big images super long emails that people aren't engaging with, that's where you're going to get hammered. Google's dumped endless amounts of money into AI and a funny Gemini work. That is that is so valuable. Like where you will get screwed over is if you're sending big emails like that that break through like a certain size threshold, Google's just going to hate you in the inbox.

27:11They're just going to say, "Screw this guy. We need this data bandwidth for Gemini. We're just going to put his stuff in spam and like good luck." So if your emails are not like like compressed, like your your team isn't compressing, some of these gifts will automatically compress, which is nice. That's kind of where I'm a lot more worried. yeah, I don't know if you have a separate take or you want to do a hot take.

27:31>> I've seen a lot like honestly the first sub had a lot. >> Yeah. >> Paragraph versus >> I've seen a lot of text versus the image. Same thing. It makes sense. >> Did that You're saying almost >> Yeah, that that hap right now I'll tell

AI Reads Your Emails

27:45you is like two three years ago AI couldn't read images. Now they can. And you know what I tell my brands and people doing right now is literally take your email that you're creating, just shove it in Gemini and tell them, "What's the context of this email?" If it can't read it or it's giving you the wrong message, then you're probably not hitting it with the images. But if you literally go in there and tell it and it can read your email, it's going to read your email on the inbox side. It's really that simple. It takes them one extra step and just shoving it in Gemini and telling it, "Tell me what the heck this email is. Summarize this email for me."

28:16>> Cuz that's exactly what you >> That's all Gmail is doing right now. So you're using the same engine. >> Read it.

Engagement Signals Matter

28:21>> Yeah. >> One one new criteria which is not in my present past so that they they measure how much time actually the person spends reading it. So if your kind of image is something engaging and then people will be looking at it more. So this place So some makes similar to a website. So how much time you keep seeing it? >> To To to that even if your creative good and it's really eye-catching people forward it they save it. These are the things that you can't measure deliverability big huge impact.

28:56Ability that you'll never see especially on Gmail. >> Well, you know it's kind of that's correct. >> Yeah, forwarding saving if they open it and they save it or forwarding or like the two biggest like indicators that can't measure. Any will get any of you back in the scene. >> Yeah. Yeah, I mean it's a black box. It's just like metal like no. Most of us want to understand what people at those companies don't even understand how it works and like that's how you should be what you should be approaching is like how do I put the best content in front of my audience?

29:27That's why I had that slide earlier on a frequency testing because if more emails is not sending more profit into your business, what it is doing is pissing people off. They're probably engaging with your email less and you don't see the hurt now, but three months from now you'll be hitting a lot more promo and spam tabs than you want to be and something to be really like survey about thinking about. One like one like hack that we do with all of our brands is just try to keep top of mind. Like that's why I had that really cool survey example. But every month we try to like spike like sprinkle in a few opportunities for customers to actually engage with the brand. It could be like a really simple like hey, what did you think about this like product that just sold out? What color should come next? Super simple or hey like here's like our top five Instagram posts like this post just went viral. They need to get someone to just click on a

30:12higher clip than they usually would. Just to get them to engage because like you were saying like the like Gmail has all these crazy things they're doing to figure out are people engaged and if you can kind of add that to an extent and just prove to Gmail that you're interesting and you're relevant, that's great. But at the end of the day like just be interesting and relevant. >> Just to add to that Gmail now does prioritization based on your engagement at previous history. So if you aren't opening emails of a single behavior user isn't opening emails right now and they ignore your next seven to 10, they're going to deprioritize your email, put it down lower, push it to promotional, eventually to spam if you're not engaged because you're not that as a customer experience improvement for Gmail. You got to remember, Gmail doesn't care about your emails, they care about customer engagement, and they want you to use the product. So, you've got to think about it like that. And now with the AI prioritization that's happening

31:01in Gmail that's rolling out right now, it literally is engagement based. If you're sending a lot of emails, no one gives a crap about them, well, you're going to keep getting pushed until they just start putting you in spam cuz if it's important, you'll pull it out of spam. That's really That's really the fundamental answer there. >> Yeah.

Inbox Competition Reality

31:15Yeah, I I would I guess I'll just add to that like that's why I mentioned like some of the brands that you're fighting a tough fight. I look at like a lot of you guys' brands that sell like a one-time purchase product that like most people never purchase again. Like you're going to have a really hard time like making email work because they would like they already know what you're selling and that thing you need to bring kind of push this ads at them. But like the percentage of people who are going to open it is like less. You're going up against like Abercrombie. You're going up against Sephora in the inbox. Like good luck. You know what Sephora customers do? They open every single text message and every single email.

31:48They click every single one. Who do you think is going to win that fight? So, just like you forgot like why I said like you're you're fighting a tougher battle because you need to just be so much more advanced in that area, which like is also questionable marks for that. someone else had a question. Yeah.

Image Size Limits

32:03>> What's your max image height and image size you would allow you to have? >> that's a really interesting question. I would I go off of what's mobile friendly. So, all image size for GIFs, is GIFs are really big. We try to keep it under 200 KB. There's some armies that people are here, so actually what I choose what they'll say is this like the label. for regular images, I'll try to go under 100 kilobytes. I know the whole email like this amount loaded people that I sent my stuff on my phone. Height-wise, I I know if there's like a hard cap. What I tell my team is, send it to yourself on your iPhone and check out. above the fold is like all the messaging that you're trying to get across is a clear as your call to action. Does it have one area or is it social that created? If not, it's too long. People don't have time. Roll like that above the fold. That's your moment to sell what put put a put a

32:53message across. If you don't win there, then you're screwed. That's why Max said like put the offer at top because a lot of people just open it will be will just pass just like a website. So that's like your moment of shooting. Yeah, geo flips. Yeah, a lot of the ESPs will like tell you like when your butt's like when you're at risk of flipping. There's also a lot of like software you use that like also like warn you if your screen exercises too big or whatever.

33:18But yeah, I don't know if Army sent folks on your website. >> Gmail usually puts the app name monthly. >> Yeah. >> And that's we optimize those images. So basically the main thing is that's that's sent this email to ourself and I'm looking at all of this still okay. >> Yeah. >> Yeah. >> Yeah, it's like a limitation for that. >> Curious by the way what you But I did that that's something I can support. >> Yeah, the biggest my thing is like some that felt building an email and you know on the big monitor it's like 2,500 pixels long. I don't really need to delete the whole thing different places. I can just throw the whole super long image in that email.

33:56So I appreciate it like it all zoom to your CVs. And I was like has anybody ever just thrown the whole like 3,000 thing in there and >> Some people are like >> How did that happen? >> So I did chop it but I'm just curious >> anybody >> So it's bad thing about that is it it eats the whole of the whole thing. So if you put a big image and someone's in the subway, it's going to take a while to load that.

34:18>> Yeah. >> And and that that's what you need to worry about cuz it's going to load them in order unless you got like somebody who just because they're so big it'll eat prioritize them. But like you want it to like as quick as possible load. So that's why you break that into chunks. If you do it all like that, don't do that, please. >> I'm not sure. I'm just curious. >> Yeah, oh fair enough. Fair enough. No, I wouldn't There is a brand, I don't know if you guys know Harry, they're like huge mistakes, but I don't know how they do it. Their emails are like so long. like like scary like if you guys have watched Neil Patel or whatever you used to know like Neil's like it's like the same how long their emails are. I'm pretty sure they'll get clipped. I don't suggest that strategy. One matter of what's your brand, what you sell.

34:57>> maybe >> Oh, you have an e-commerce store? Okay, cool. There. what I would challenge a

Email Length Testing

35:01lot of brands on is like AB testing for some of the bigger brands who are AB testing the length of your email based on your niche. If you're like in apparel, fashion, beauty, stuff like that, I think length matters, but I'll tell you for a lot of brands like most people will click that first kind of call to action. It might get through the first couple sections, but if your email is like a novel, you really got to ask like who's reading this. And there's a really simple way to figure this out.

35:25Just put beauty hyper links throughout your email and take an analysis of like, "Okay, how many how many people actually click?" It's very similar to like screen recording when you guys hear like CR alerts. Just looking at like where do people drop off? What ESPs will also just show you this visually. But just make sure you're not sending super long emails just in general. It's so much more work for the creators and a lot of times like the juice isn't worth the squeeze. There's like a famous like Steve Jobs quote where he went to like a marketing agency. He asked them like he asked them to like hold up like this part I can't page. I can't have it like some like simple like thank you for I don't know what he's asking was. He was like, "Yeah, we were messaging but this isn't enough." And like what the guy did is he he stretched out like 10 pieces of paper and threw them all at Steve Jobs. He said, "Catch this." Steve Jobs caught

36:11nothing. and then he threw one piece of paper he said, "Catch this." And threw one piece of paper at Steve Jobs. He caught the paper. He was like, "This is like marketing. This switch." The crowd is like, "People can't capture 10 things." Like they basically have too much going on in their day. So you you know you just have one much super clear message and like that's that's kind of my case for like like does this email need to be long?

36:33And she asked that question I ask in all your flows. I start short and then expand versus really around. >> I'm just looking at >> Yeah, I'll say one last thing. I think a lot of brands try to copy the bigger brands. And when you're Costco for example, I don't know if anyone subscribes to their Costco emails, it's ridiculous it gets clipped. But I'll tell you right now, most of the time I click on the view image and I look at it on the web just because I want to know about the information. But most brands that are not of the brain affinity to be able to produce things like that.

37:00>> And that's so you got to remember like I still kind of said like one focus on one thing, one message is really still still the king when it comes to e-commerce brand. >> Yeah. I I used to I used to sell to Costco a lot of brands there you know. but I I used to run like the biggest e-commerce account one of the biggest e-commerce accounts in Canada, but I can tell you that Costco has no idea what they're doing and they kind of works.

37:20They're shooting from the hip because they have a great product. There's I have no idea. If you copy that like and that's like a rule of thumb like big brands are doing so much wrong. When I look at them there's like so many issues. so like like I said like go back to my first point in this presentation like coolest brands doing what's tested. Let's see. Two weeks from now let's figure out what's right and what's wrong.

37:42>> Like this like this feature allows you to test >> You guess. >> You guess, yeah. >> Yeah, may lose your attention. And it's I agree with you what Phil said. >> And it's so easy kind of like with the XCPs and AI coming like you can literally just be like Claude like before I send this you use up-to-date URLs and like rebuild a framework or like the system and then a month later you could just say hey Claude go through all the emails and the framework you built and give me a report of like how deep people are going in my emails over time.

38:10That's that's where we all live. >> Okay, it's up there on the right of the screen. >> All right, any other questions? No. Okay, cool. random bonus content.

AI Workflow Automation

38:20I You guys are going to cover a lot of this, but this is just my take from like that agency standpoint on how we're using AI and how I suggest everybody here QA sends right before they send out. You can set up like a con- cover You can set up a flow bot. You can set up an automation to just like hit the API or with table or draw plan and just QA sends before they go up. You can I don't know if you guys use Notion or whatever.

38:46You create a table or you sort of just offer the status and just have it double-check. I'm sure everybody here has messed up at least one or two like big sends. It happens to the best of us. AI lets it never happen again. It just kind of also I also would reduce the load for your virtual assistants by the hours whatever. Pre- pre-populating content for your visuals. I don't know who said this earlier, but you can now have like Claw just like builds out an entire flow and sort of fill out all campaigns and just have it like pre-populate a bunch of content so that your team goes into like updating an email. It's like already kind of done. You're not spending time like doing a drag and drop or with the right stuff. That's really interesting.

39:25Building dashboards and reporting. We could build like some really cool like dashboards internally. Stuff that like companies would charge just like hundreds of thousands of dollars to build in the past. We're building it all now for our clients in real time. If you're not doing that, highly recommend it. And like marrying up with Shopify, especially the holdout test setup I said earlier, but marrying up on Shopify data with your ESP data and like doing those holdout analysis is great. Exploring brainstorming. You can do a ton of I'm sure I don't have to explain that.

39:54Building a structure dimension. Although this isn't even working out, but like just like WhatsApp emails like cuz I think buyer Google Illustrator or whatever just auto uploading that all into whatever ESP you're using. Analyzing past emails for expirations. I shared this on my Facebook group on the NCP to say, "Hey, what did we do last July? What was our further July promotion? Like what made us opt in the most?" Just have it, don't you? Just instead of having it do all the creative work solid for you, I should just do the initial data to bring sorry.

40:25Then this is one that I'm like going super deep in. This is like crazy alpha. And I don't remember here I was saying what I was doing, but chat GPT is a model that's like awesome. Like like it's like the craziest work that's like I've been all of it every time I model code I test but you know what I'm saying? Chat GPT is a model if you feed it the right about here, its ability to create like really good looking like almost production ready like designs looks really good.

40:54And so we've been like building up tools for our designers there. I can see like the copywriter can hand off copy and my designer where they open up that copy to design and pull out 20 like fully built designs that they could go I like this one, I like the hero on this one, the collection on this one, I like the call to action and like they could go and like figure out tweaks and stuff. All right? It's not like fully like being Figma or the product circle tab, but that's that's where we're going like very quickly and especially some of the agencies in the room like if you're not already thinking about this, I would say don't sleep on subject

Subject Lines Are War

41:30lines. This is like when I look at email like for me it's war. Like in the in the inbox like I've got to beat out other beings for with my subject lines. These are just random examples I pulled up earlier today staring at for a fashion brand. I've got your you said you were going to shopping and your favorites are back in stock. This also for fashion brand. Fresh. This holiday's banned in China. Just out of curiosity but all of that right like that. I'm going to be scared of that.

41:56This is borderline like shady so I don't want like necessarily go to the back but the premise of this is like subject like email inbox like actual inbox is war and copyright is what wins you that war. And like everyone's using AI now so everyone's suffering clients are okay, but there's still so much like wrong. And if you're not like it's I see it just like Twitter. Like, if you have like six or seven words and you don't like optimize those words to like the best you can, possible, support is going to get that one.

42:27That's what you should all originate on. It's like I don't I don't I don't use support. I don't I don't need I don't need to call me or hour or two or whatever because that's who people want.

Offer Testing Playbook

42:38Okay, offers I like this is like very generic, so I won't go too deep on it. what I was talking about testing offers. We We test all of these at once. Every brand we look at we have quite different offers that we think might work. I just It really costs more. Like, it really depends on what you want cost percentages. We've been playing like really crazy like buy five get four free like crazy things like that. They even offer in the software roll out to there like buy one get five free gifts for all like like that for more like the slice subscription people would probably get a lot of order This will crush you. We should both go deep or something. So, buy one get one free or something. So, buy one The question is if you're AOB is sales and the force tier discounts buy or save more. gift with purchase. I'm going to put a really big call out for all the retention marketers in here. If your

43:24boss tells you, "Hey, let's run a gift with purchase on a lot of product that's that we have too much of that's moving slow, you should tell your boss like absolutely not. Like, this is a waste. This isn't going to help our deliverability. It's crazy how many brands I see try to show a really hot product to get a purchase and think it's going to do something for the business. It never does. It just creates problems with the LTV because we're just giving away 1,200 droppers. gift card credit with purchase on the website, free shipping threshold. These are all work for different brands. there's like clever ones that more sneaky people get out. So, like refund one in five orders.

43:57They do like, "Hey, you won a lottery today." you like you're like one of what percent of people gets very tricked out. I don't know. subscription only offers of something higher to less frequent shipping, but I mean they will do programs as well. There's probably a lot of you guys too like rebuild or multi-programs. I'm going to say like eight isn't a brand you sign up for like ghost map programs. Just set up and never bring it up again.

44:23This is like a really simple hack to like do points drops, do occasional like deeper discounts if you use your points or just like like multiple points multipliers. These are all like your campaigns that if your brand is big enough to your VIP seconds, you start actually moving the needle to the bottom line for us. Let me see if I missed anything else.

Wrap Up And Q&A

44:44That's all I have. This is I don't know if if you're really into generic stuff, but this is just like I'll do like a anybody wants to stand so this is just my calendar, but the first five people who stand at I'll just stand on the phone with you guys for 30 minutes. Just show me your army sending account. Show me what you're doing and I'll just roast you for 30 minutes. Five times you do that seven times like you just go tell your boss yeah, it's cool what he does. And like I won't say anything or tell you anything. So feel free to use it.

45:15I put my with my brand's thing in my folder, so feel free to reach out. I'm here for Q&A, but that's all. Thank you guys for having me. Appreciate it.