Scaling Performance Marketing Programmes on Facebook | AWasia 2016


Welcome everyone and thank you for being so brave
as to attend this one. So I’m going to ask you to introduce yourselves. Well, I’m actually gonna ask you
to introduce yourselves in two parts because we have two representatives
from Facebook here and we have two people who are doing the impossible and making Facebook work at extremely large scale as affiliates or performance marketers. So Facebook people, please introduce yourselves. Hi guys! First off I’d love to say,
Hugh, your trousers are fabulous. Oh, thank you so much. Congratulations there. So I’m Maria Claudio, a client partner at Facebook
based in our Los Angeles office. I have done kind of the US tour, I like to say. Base in Menlo Park headquarters
for 18 months in New York a year prior to that. Work on performance marketing in the States
but also cover the AMEA region. And I will hand it off to Niket to introduce himself. Yeah, sure, so my name is Niket Shah.
I work out of the Facebook Canada office. I head up the performance marketing team out there. Similar Maria, we kind of focus
on the North American aspect as well as a little bit in Europe. I’ve been on Facebook for 5 years
as a client partner as well. Thanks for having me. Interesting fact about Niket and Maria is firstly, you are a successful entrepreneur as well, I believe, before Facebook? I was on Wall Street before Facebook. Trading options, arbitrage,
so lead gen, makes a lot of sense in my book. And numbers that make ours look small, yes. Mysteriously, I’ve been chatting to a lot of the
affiliates I know around the conference and they all seem to know you already. That’s a good thing, I hope? It’s a very good thing! I’m wondering if you’re the secret Facebook person on STM that we’ve heard about for so long. No, I don’t think that’s me but I think I’ve done a good job of just being the human aspect like what you mentioned earlier is that
Facebook’s up in this mysterious portal for a lot of people and I think Maria and myself have brought that human aspect to what Facebook can talk about. Cool. So, our affiliate performance marketers,
please introduce yourselves. I’m Jason Kryski. I have a performance agency venture fund
called Strawhouse and we’ve been working with Facebook
for the last, almost 2 years. Cool. How’s it going everybody? Big up. I’m Alex Tshering, business consultant. I’ve been in affiliate space since 2008,
running Facebook Ads since then. If anybody’s old school, Tracking202.
I used to work with that team. These days I work with Fortune 500,
Fortune 1000 companies and teach them direct response techniques
on Facebook. Yeah, I was amazed to hear that Alex is one of the people
who I can blame for Prosper202 which many of us will have used. Okay, let’s get right to it! So I have initially a question for the Facebook team and it’s a question that I think
we’ve all asked an awful lot and often found that we couldn’t really get an
easy answer to, so seriously guys, what can we run on Facebook and what can’t we run on Facebook without the ban hammer
descending like the wrath of God? So we’ll definitely go through a little bit about what you can and cannot run and I think what you cannot run is the things that you’re already knowing about, like the weapons, the prohibited items
that are already on there on Facebook. For the rest of the stuff,
it’s not so much about what you’re running, it’s about how it’s being run. I think how it’s being marketed
and how it’s being advertised on the platform is what separates a really good partner,
a really good marketer versus an affiliate that is getting accounts bad all the time. Right. Yeah. I had a feeling
this question would come up so I brought my phone with me to dive in some detail for you guys, because I know it’s painful when you go to our policies page and try to navigate that because there are some gray areas which we should do a better job addressing
and we’re working on that, that’s one of our big initiatives of 2017. So we have our absolutely banned content, of course. Weapons, drugs, casual dating,
short-term loans, payday loans, all of that we don’t want out on the platform
because it provides a terrible user experience. At the end of the day, Facebook wants to have the best user experience because we’re such a personalised platform at scale that, when you’re looking at content
that’s next to your friends and family whether it be positive stuff they’re posting, or negative, death, birth, its life events. We wanna make sure that the advertiser experience fits in with that, and there’s nothing that basically promotes something misleading
or something a user would not trust not only us, but our advertisers at the end of the day. So these kind of difficult areas,
and I’m sure a lot of people in this room have had issues with, for example, dating. Dating’s allowed via whitelist but casual dating is not and I know that’s kind of difficult to differentiate between the two. Health-related. Financial.
RMG, real money gambling. All needs to be whitelisted. So you’re saying whitelisted. Does that mean the offer needs to be whitelisted
or the affiliate needs to be whitelisted? So whoever is running that offer needs to be whitelisted. For RMG for example, you need to prove that you are working with a company that has a
licence to promote gambling in the UK or wherever it is allowed. Nevada, New Jersey in the States, for example. Elsewhere in the world the policies change by country which also makes it a little difficult. We’re trying to provide you
guys more transparency with that. I will ask you about that once you get
through the list. So, low quality products is something also that’s kind of new into our realm. We wanna make sure you’re getting what you’re actually buying on Facebook and it’s a high-quality product and I understand e-commerce is big,
especially in Affiliate World. So the biggest thing is like anything misleading. All capitals in your ad. We don’t like that. I’m sure you guys have gotten
disapprovals for that in the past. Online pharmacies? Allowed but you need to be very careful.
Make sure you’re being compliant with the rules and regulations of the particular
country that’s running that. For example, also things like alcohol in different countries varies. You need to be targeting 21+ in the States, Europe 18+ and areas like Malaysia or the Middle East,
it gets a little more strict. And then quizzes, sweepstakes,
background checks? You just have to be very careful
of how you’re positioning it. You wanna think about you as a user on Facebook. You wanna get an ad that’s relevant to you and serves you in the best way possible and you don’t want an experience
that you might think is creepy which is the term that we’ve been throwing around
in our discussions lately this week. I just wanted to build on one of the things you mentioned there
about user experience and why that’s so important to us at Facebook. Facebook’s a unique platform where
the newsfeed that you create is so personal to you. It’s a collection of
family, friends, pages and brands that you’ve selected to hear about and every time you log into Facebook, Facebook has the responsibility to show you the best
content that makes it the most relevant, engaging experience for you. And what happens in marketing as a whole, we don’t typically think about the sentiment that goes back to the platform that you’re advertising on. So if you’re thinking
about marketing in traditional digital, so you think about, you know,
the web properties as we know of them or even traditional offline media, things like television, billboards even, if you see a bad ad or irrelevant ad on any of those you don’t actually blame
the person who put that billboard there or you don’t blame the television company for showing you that ad, but on Facebook, because it’s so personal, you put that responsibility in Facebook’s hands and for us, we get that sentiment back
really quickly from users. User might not engage with ads
as much as they used to engage with ads. they may not come back as often as they used to, or they might not even purchase
another product off of Facebook, and that’s a behavior or an action
that we just can’t afford as a business because if that happens, there’s less and less users on the platform, it doesn’t become a marketing vehicle
for everyone in this room and everyone that’s been talking about it
at the conference and that’s just not something we’re willing to risk. And when we talk about ad accounts going down, it’s primarily to ensure that
we’re protecting the users of the platform. So fundamentally, your issue here
is that you wanna protect the users in quite a personal space, first and foremost. Okay. So as an affiliate, and I’ll come to you guys with a witness, I promise, as an affiliate, my response to that would be we’re happy to work, many of us are happy to work with the limitations of whatever traffic source we’re on but the issue with Facebook is that
it’s so opaque, you know? I am promoting a product so legitimate,
I’m perfectly happy to talk about it on this stage. It’s a VR game, One of the biggest and first VR games in existence. Still, when I’m putting up an ad on Facebook, every time I’m doing it I’m sitting there, going, “Is this gonna be the one? Is the banhammer gonna descend?” And how do we avoid having that? Are we gonna get hit by lightning experience? Yeah, I think that’s a really good question. I think it’s something that comes up often. The rules are opaque on purpose. We wanna make sure that it’s a platform that, our policy team can actually work on it and shut down bad advertisers
when that does happen. I think the way to look at it is, “Is this a misleading experience
for the people at Facebook or someone that’s using it
or seeing it?” A common example that we see, and something that runs
on every other platform compliantly, is a promising of an X result in a Y timeframe. So, you know, “Purchase this weight-loss pill
and lose 14 pounds in 2 weeks.” It’s something that we see all the time, or “Removing your wrinkles by using the skin cream.” That’s not a guarantee or a promise, it’s like a misleading experience but you’re using it to get the click
and you’re hoping that someone’s purchasing it. By doing simple nuances like, “Click here to learn how to manage your weight better.” Or “Click here to learn how to
reduce the visible signs of aging.” Those simple tweaks in your copy will make it compliant for Facebook because now you’re not promising a result
in a wide time-framing. It’s kind of goes back to
what Maria was saying earlier of how we’re trying to ensure that the experience is something that’s a high-quality
for everyone that uses it. So if you’re presenting a possibility
then that’s legit. So how we wanna position any ads is that, your product or service? That’s what’s being advertised, not how to get people
to use your product or service through click-baity marketing, and I know it works and it generates profits for you guys and that’s kind of this back and forth that we play with all the time,
especially in this space. A great example that I’d like to share is so user perception is very subjective. An ad that I might love, Niket might hate. So Teespring, for example. Years ago they used to run the name scraping campaigns where it would be like, “Oh it’s a Niket thing” or “It’s a Maria thing.” Niket might be like,
“Oh this shirt is awesome! It’s so highly-targeted to me,
Facebook knows me. This is awesome!” but I might be like,
“Okay, no. Too creepy. Don’t want to deal with it.” And so our policy team is faced with kind of that, the two
sides of the coin? and how we rule is like, we want to be as conservative as possible because we don’t want
bad user experience in the platform. If we can prevent that, completely. So that’s why we’re a little more stringent than a Taboola or a Bing just because it’s such a personalised area that we need to be very mindful of that, and that’s why there’s more rules and restrictions than other channels in the industry. Is there any likely that
we’ll get more clarification on those rules or even, what would be ideal would be, and this happens already
with affiliate networks and other traffic sources, is there any chance that we will at some point get the opportunity to just say,
“Okay, I’m going to submit this ad and then it will get looked at by a human being and if they say no, that’s too far, they’ll come back to me with feedback
and then I can improve on that”? It’s coming. Right. Any timeframe? Some point in 2017? I don’t want to overpromise and underdeliver. Some point in 2017
was better I was expecting. That’s cool. The great thing is we partner very closely
with our internal ads integrity team and they understand what a vast and great business you guys are running and so they want to be more client-facing and help you guys understand what’s okay, what’s not okay. And it’s something we’re working on
with them in collaboration. We will be for the next several quarters. Yeah just add on that like, it is, even when we go to human review, gonna be subjective. The thing to look at when you’re running any kind of campaign, any kind of ad
is ensuring that it’s not misleading. If that’s the simple guideline
that you follow or take from this, if you do that then you shouldn’t have any issues
running into any account bans or any policies. Even in some of the talks that we’ve heard over the last two days, I think several people are running 10-figure businesses with a single ad account, and that’s normal. That’s how we expect it. That’s how it should be. And I think it’s the people that create a misleading experience
that kind of ruin it for everyone and I think that’s what we’re trying to get away from. Just to add on to that from a digital marketing perspective. A rule of thumb I use with a lot of my clients or even in my own campaigns is avoid the perceived condition.
I think that Facebook can attest to this? The utilisation of the word “you” in your ad copy is assuming that the user has a particular behavior or does something. “You have terrible skin. You should click here.” And just trying to avoid that “you” at all cost. It helps get away from the tactics that people think as devious or misleading. You have answered the question
I was about to ask which is, how do you guys avoid
the thunderbolt of God? The general idea of trying to exploit a user base to try and reap profit from that,
you’re gonna have a bad time. It’s just not gonna work out well. Typically affiliates always try to optimise towards ROI as being the only KPI that really matters and by optimising towards ROI, you’re going to do things in your ad copy and your creative and your landing page
and all these things are going to cause these systemic challenges to happen. I mean Facebook,
like Niket and Maria are saying, they’re trying to protect their users If think about it, if the users have a bad time and they leave Facebook, Facebook doesn’t have a great business. So being able to give the users the best experience is in everybody’s best interest. So “How can you achieve that
by still driving results?” is the question. Can both of you maybe give us some examples of how you would craft specific copy of that is both compliant with these guidelines and also actually sells? Because obviously you can come up with
a very, very compliant copy that then sits there like a dead cat but how do you craft copy that sells
and fits in these guidelines? I think the best way is to look at other kinds of really successful content
that goes viral that other people share willingly
or comment willingly, like willingly. I mean people like that kind of content and how can you take that kind of content in your vertical, your niche and apply that to your advertisement? So if it’s a recipe thing
or if it’s like a nutraceutical-oriented thing rather than using a picture of a fruit, maybe you’re using a Tasty-style video or something. Right. Alex? I really like to run aggressive ad copy so there’s a thin line
where aggressive can be tiptoeing the edge. I stick to that rule of thumb of really
avoiding the perceived condition of “you”. I utilise examples of of people who are
benefiting from a product as opposed to saying, “You can benefit from or product”,
“You’re gonna lose 2 pounds in 2 days.” That’s not that much weight actually,
2 pounds in 2 days but Instead say, “Many people who used our product
have lost this amount of weight.” Maybe that’s a little bit aggressive
but you could still kind of tiptoe the line and still come within there without being really dodgy. And I think, as an affiliate,
we come from a place where it’s like you just make up stuff. You know, Oprah endorses my stuff,
Dr. Oz endorses my stuff. And if you’re truthful to a point, I mean, we’re crap whole marketers. If we can build ad copy that’s truthful yet deceiving,
if that makes any sense, I think that there’s stuff in there
that we can really use on Facebook. So I’m kind of feeling like
I wanna sort of throw a rule of thumb at the Facebook fellows, Facebook people, and see whether this holds true. If we basically have the rule of
“Don’t directly pick out the user” and “Do tell the truth” is that like 90% of the law? Not necessarily. For financial services, health is like
calling out someone’s physical attribute. That’s where the “you” comes into play
that’s really bad. Actually when Alex and I first started working together, he was running ads with all caps and he thought I hated him for a solid 2 months because I kept yelling at him for it. I love caps. I love you, though. You know that. But it works. That’s the thing. So we understand that works for you guys and we want to not go over the line? Get maybe about 10 meters to the line and help you guys drive your business
at the end of the day while preserving the user experience. But I think the “you” really sticks out more for these products that touch onto personal, identifiable information, so health, financial services, anything that you’re exposing data that you might not certainly want to do online. That’s kind of how we look at it. But I know for dating, for example, like, “Meet new singles in your area”? That copy is fine, but if you’re like, “Oh you have low credit.
Click here to get a subprime loan.” That’s not going to fly. So it’s definitely you want to think about what the offer is but the degree of personal information that you’re giving and then that way you can kind of judge. We also have obviously a machine learning system of how we policy ads and you will get picked up and then go from auto review to manual
review, even if it’s not for necessarily financial services or healthcare or
whatnot. Another great example of things that
have gotten shut down in the past, an ad for a pharmacy. “Get your free flu shots on Tuesday.” The word shots and Tuesday got flagged for alcohol and it was targeting under 21 in the States? So stuff like that will trip it. We know that our system is not perfect and we wanna be your guys’ partner
rather than your enemy and help you guys at the end of the day. If you don’t mind, what happened with the ad in the end? It got approved eventually, but it took, you know. Manual review and all that fun stuff. So we’ve talked a little bit here about stuff we find difficult. I’d like to flip this over for everyone now and talk about the stuff, the potential because we’re all aware Facebook
has an unmatched gigantic reach and scale. So what is working well? I know there’s a seasonality thing you want to talk about. What works well on Facebook in your experience, as the affiliates in the crowd here, what should we be looking at
if we want to make a lot of money on Facebook without getting into any trouble? Cloaking. Drop mic? Do you wanna just drop the mic now? Not endorsing that. We did promise this would be an awkward talk. It doesn’t exist. I think a lot of different offers can work well and I think it’s what you typically see, so there’s health and skin and beauty products that work really well on
Facebook and they pay a really high margin so it works really well,
especially in the affiliate model. There’s e-commerce products
that work really well. I know we’ve all seen or run flashlight offers
in the last 12 months and I think that works really well because it’s at a certain CPA and you’re working on driving a conversion that someone really wants and you’ve got lead generation
which also varies in CPAs as well but it works really well. I don’t think there’s a specific vertical that we would say like,
“Double down on this in 2017.” If you can change the way
you perceive your product or perceive what you’re advertising on Facebook
and make it look like a real brand you can really run anything, and the success of Facebook’s been that for the last 5 years in terms of the
advertisers that we’ve brought onto the platform and obviously you’ve seen the success. It’s because if you think of it from a brand perspective, you can sell anything. Some of our best friends are the guys at
Dollar Shave Club, Dollar Beard Club. That’s kind of like
what the holy grail of any affiliate advertisers would be. They’re white-label dropship products that they’ve created and put a brand around it and Dollar Shave Club
just got sold for a billion dollars which I’m sure anyone in
this room or any affiliate marketer would love to have their product sell
for that much. So just one quick question following up on that. We mentioned product quality
and products experiments earlier. Do you as an affiliate have to worry about
the quality of the offer? Yeah I think you do. I think a lot of online marketers hide behind that laptop and I think it’s like,
“Well I never have to actually see what the product is or see what’s happening
with the customer.” But you wouldn’t treat a customer
that walked into a physical store and sell them a product because the assumption is or the fears is that
they’re gonna walk back into that store. I think you look at it the same way online. I think what we’re trying to crack down on is you’re purchasing a dropship product
that no one’s ever seen and you’re buying you
know a really nice silk or cotton dress and you’re getting like, tissue paper in the mail which is something that does happen and it’s something that we need
to control a little bit more as well. The same way that Amazon
and other partners do quality control. Cool. I have to put my Wall Street hat back on. Seasonality is something I look at, I’m interested in just because, coming from the trading world, I get that there’s ebbs and flows and trends. This is not a silver bullet by any chance or any means but things that I’m mindful of is, for my lead gen advertisers, how do we hedge our portfolios? For example, Q1 financial services-related products are great we just spent a tonne of money during Christmas. We want to maybe take out a personal loan or refinance our house or
one of those. Those tend to perform really well in Q1. Then Q4 you get to the e-comm holiday. Those offers are great to hedge when the
financial service lead gen stuff is doing really poorly but then you wanna keep in mind, and this is something
anyone in this room can look up, like temples, so the Emmys, the Oscars,
Superbowl. Things tend to get more pricey during those periods in general, across the board unless you’re a very big brand advertiser which I don’t think this conference is about whatsoever. So it’s just things you wanna be mindful of, like the cyclical nature of the year. The tax guys tend to spend really big
in the States, for example right after your W2’s drop, up until Super Bowl. That’s their big activation, where a lot of people think it’s
around the 14th of April when you file, but that’s not really the case so like, as much research as you can do around the different activations and what’s going on, obviously the big brand guys just have
tonnes of money to drop, don’t really care about performance. So whenever there’s something
like these big shows in the States or FIFA or whatnot that’s when things are gonna tend to get pricier so I think it’s something to be mindful of and come up with this seasonality matrix of whatever offers you guys are running, I think it’s just something that interests me and I’ve seen the data behind it and it’s true . Prices do fluctuate. But I’d love to get Jason and Alex’s opinion. This whole like, what verticals work and seasonality and what you guys have seen. Yeah if we talk about traffic flow, a good example, just really quick
not so much in the offer realm but you know, the elections were in November. On a lot of the stuff we were running,
we pulled back on advertising in the swing states. It was a handful of states but what we saw
is that we got a gained efficiency from just pulling back there because now we’re not bidding and trying to compete with
overinflated CPM, so that’s kind of an interesting strategy and we saw a direct correlation with costs
coming down with doing that. As far as products and offers, you know, for me, it’s always going to be either
lead gen or e-commerce. E-commerce is really the place to be right now. I mean everybody’s in e-comm but if you have a solid product, if you market it well, position it well, use retargeting, really build out that funnel strong and retarget multiple points in that funnel,
you’re gonna have a great time on Facebook. With lead gen, if you’re able to control the flow and you’re able to retarget on that same level, you’re gonna be able
to really hit that user on multiple points and if you’re placing that pixel on your page, people are hitting that page
from other traffic sources. You’re just gonna be able
to constantly reach out to them, so if you can have an offer, ideally and in my world
I’d like to have a broad audience and a talk, earlier talks, about that. Broad reach. I like to work with CPA’s around $25+. Broad is the way to go if you have a broad understanding of your product,
it’s gonna work. Yeah, same. E-comm, finance, nutra. Those are great staples. Those are what we do. Those are the kinds of companies
even we’re reaching out to outside of the affiliates space that we’re building
performance campaigns for or even verticals are around there. Okay so it’s an interesting thing
that I hear from both of you is that you’re kind of targeting, much like Oded was saying in the last panel, you’re kind of targeting higher payouts here, $25+. Is that just because that the
cleaner verticals tend to be in that range or is it a specific economic thing
based around the bid verses. So I have a few lead gen campaigns
that really worked in that $1-3 range. I think that what you’re gonna see on
Facebook for the most part, you’re gonna be paying CPM’s of probably on the low end, like $12 especially if you’re running
a conversion targeted campaign. Right now I have some advertisers that are running
an $80-90 CPM’s, which is ridiculous but if your CTR’s are high
and if you can make anything work that $25 gives you, I can get a 1% CTR
and probably still try to make this work. If you’re trying to make something
on the lower end of the CPA range, $1-10, you’re gonna want to get CTR’s, link-click CTR’s, through your website. Anywhere from like 2-5%, maybe upwards of 7-10%. Those are high numbers so your ads have to be very relevant. They can’t be link-baity and they
have to be aggressive to a certain extent. It’s a harder game. Also I mean one of the real beautiful things we’ve been doing lately is it takes a lot more effort but more personalised creative to the targeted audience? So segmenting your audience down
and then building creative for that segment of the audience
that’s more relevant to them. It’s gonna produce lower negative feedback scores and better engagement, and you can perform as well without getting banned or having the
negative effects of accounts. Okay, interesting. So the next thing I want to ask
which is actually a question that one of you suggested and it’s a really, really good one, so I’m stealing it. The aha moment, because it’s becoming obvious to me that the way that you think about Facebook advertising is fundamentally different to the way that you, as an affiliate,
are used to thinking about pop advertising or display or whatever. So can I ask each of you, for our affiliates, what were your aha moments and for our Facebook representatives,
what aha moments you have seen? Yeah so the people
clicking on the ads are people, I’ve been doing this since 2008. I came in through email. It was all too easy to look at
clicks and conversion rates and forget that there’s actual human beings
taking activity behind this and once you start realising that, “Holy fuck, there’s people
doing this shit and we need to like, treat them with some modicum of respect” to be able to understand that they should be getting what they’re looking for. When we started doing that, All of our policy issues went away and we were able to maintain performance
where we were running profitably. Well that’s certainly blown my little affiliate mind. Next? So I could spend the next hour
talking about this, but I won’t. So over the course of the last three years
I’ve been at the company we’ve been testing a tonne of stuff. Micro-targeted, basically people who have checked into a home-improvement store for refinance. It was great. It worked, but it plateaued. And so we realised what Alex said earlier is broad works best. And so the better and the higher-quality
your custom audience seed is, the better your lookalikes are gonna perform and the more scale you’re gonna drive,
and the more efficiencies. Anywhere from 1-5,000 in your custom audience. I know a lot of people in this room
are doing bigger. Don’t do it. I’m telling you right now, you just want to focus on
really high quality different seeds and test them, and then scale up and like a 10% might not work for everyone. You kind of have to find where
a 3% might work better for your offer. If you work with an FMP
you can go up to 20% which we found great success with and it’s kind of part of our playbook and what we found to be great
in the lead gen space, for example. But the biggest aha moment was how important it is
to test, iterate and deprecate because that’s the only way you’re gonna be able
to be successful in the platform. One thing to add on that. If you’re getting those
really high quality users from the advertiser to build those custom audiences, the quality of the users and
the leads that are being generated are better than average so beat those guys up for better payouts. Totally. Second that. I get so many aha moments. I think I’ve been advertising on Facebook
since 2009 and I think that back then the aha moment was
“Target Tyler Perry for scholarships for moms.” That really worked. It was awesome. I think Angry Russian actually taught
that technique in a webinar in 2002. I was one of the beta testers on newsfeed ads and so I think there was only 50 or 100 advertisers on it at the time and that was a time when
we were trying to figure out if Facebook was a viable traffic source to run at scale and we were seeing ROI’s 1000-5000% because there’s nobody competing and they were inundating these feeds
with those ads. That was very, very,
“Okay this is gonna be something huge.” And I think these days,
what I’m constantly preaching is that really build out that quality,
really build up that funnel flow. And realising that when you control your funnel,
you control the flow, you control your upsells and you have good positioning
and good marketing, truthful marketing, you could really make huge strides. For the most part these days, actually 100% of the stuff that we do is stock photos, you know? We’re not even using Google images. We’re clever with it. We do split images. We do videos, real videos. But we’re able to make that work and I think we come from a space, the Affiliate World. In 2009, I got a cease and desist from Harpo for using Oprah’s images in my ads. We’re so used to that but I think that you could figure out how to make stock work even so that’s kind of interesting. I’m actually, if you don’t mind,
gonna ask a quick follow on there which is, okay. Can you give us one specific example
of how you took a stock photo, which I think a lot of us are used to seeing
die on the page with a CTR of nothing, and turn that into something that worked? So I think that, in images in general, that selfie style works. If you have a face shot, and it’s of a person, that’s good. Stuff that stands out. A shocked face. Stuff like that. I have in the past made the quality look a little, sorry, I almost choked on my gum. Please stop, we don’t want you to die. Wait a minute. Spiro gum. The quality of the photo
look a little bit more granular but I think that if you have an image of a person, straight on, you know? Females tend to work really well. They click well. Sometimes we use filter,
sometimes we put stuff on top of them. A split image is good. just trying to figure out a clever way
to use those stock images and yeah. We’ve used stock video,
like stock video animation, to great success with health. I have actually a lot of success with stock video. It’s nowhere near as played out
as stock photo at the moment. Yeah I think one of the biggest aha moments, or at least coming from the brand side, is working with a lot of the big brands. The buys that they do on the platform are blocks of media, so $100 to $1 million of a week’s spend
or something like that. I was actually working with Jason
and his team at Strawhouse and it was very simple tweaks
that we were doing to the creative and to the landing pages
that actually helped them. I don’t wanna call you guys out but like going from 4-figures spend, 5-figures spend
to 6-figures spend a day in a matter of 2 weeks. So it was simple nuances like that
and logically it makes sense. Of course, if you’re gonna invest 6-figures,
you’re gonna make six-and-a-half. I do that every day. And you double down and I think that was the moment that, for us, we started really doubling down on the performance vertical for Facebook and understanding, not just how Jason makes money, but understand how the whole ecosystem works, understanding the advertiser model, understanding how mids work,
how manufacturing works, how distribution works, the networks,
how they fit into the entire ecosystem. I think understanding all of that just made it a really easy comparison
to a born offline world I think even as Facebook, we’re oriented around working
with born offline advertisers, your typical big box retailers,
the typical brands that we’re all used to but there’s an entire world that’s come up
in the last 15 years that’s been born online, and I think for us,
to understand what that world looks like and it doesn’t have to be
in a retail store down the street, it can be online and I think that’s that’s
the big aha moment for us. Right so one thing that’s coming up here quite a
lot is that you work very closely with these guys. So particularly for newer affiliates
who maybe aren’t doing that kind of big spend, is there any way to get that personalised help to work through these guidelines,
to maybe be able to say, “Okay I’m thinking of running this. Is it too aggressive? and get the answer, “Hell yes, don’t do that”? Yeah I think our Help Center is a lot better than it was 2 years ago. I think even our contact team that works with us
responds typically within 24 hours so I think the support that you’re
gonna get from them is fantastic as it is. Beyond that, as you continue to scale,
as you continue to grow, there are limitations in terms of
what you can do on the platform and I think, naturally you’re gonna
flow into a space where you’re gonna be working with Maria, myself
or anyone on our team to help scale your business
to where you can get to 6- or 7-figures a day in spend, which would be fantastic. I think it’s just a matter of making sure
you’re doing the right checks and balances, to make sure that you can go from 3-figures a day to 4-figures a day. I think that’s the hardest thing that
we’ve found for any advertiser is going from $200 a day to $2,000 a day. The challenges that happen, even within
a cash flow management system, those are the things that,
once someone has that figured out, I think the scale is infinite and I think we’ve got a support team
that’s gonna continue to build and evolve around this vertical and I think you’re gonna see it
get a lot better, like Maria said, in 2017. Okay, so an interesting question here is obviously you’re absolutely right. The 4-figure jump is something
I get a lot of questions about on STM. How to jump to 4-figures a day, 5-figures a day. At what point, blunt questions I did say, at what point is an affiliate making
enough money that he’s worth your time to speak to personally? Or she’s worth your time to speak to personally? To be completely honest the spend is arbitrary.
It’s more about the growth. That’s what we’re looking at. And so I know it’s very difficult for a lot of people in this room
who wanna spend and wanna scale and feel like they’re being blocked and don’t have a way to do so. I know in APAC, for example, it’s very challenging. So we have taken the initiative as this team, and my first AWA
was a year ago to this day. I’m still fighting the fight internally
for everyone in this room because we understand
the value you guys can drive to Facebook and so with time and effort we wanna be able to scale out a vertical
that supports you specifically. That’s our big goal. That’s our meaty project that Niket, myself
and some other colleagues have been taking on. We are a team of, what, six people globally scaled with this billion dollar revenue opportunity and we wanna make sure
we’re doing it the right way, and we’re working on it
and going to enlist support. How you want to look at it is like, the education piece where you’re on board with maybe a small-medium business rep
within market and then kind of scale up your spend, get educated, get to that point
where you have the pixel implemented, you either are tapping into our API or running really sophisticated campaigns, kind of scale up
to the upper global tier of service and that would fall into what we’re doing. But the reason we keep
coming to these things and partner with Jordan and Lorenzo and Hugh
and everyone here is because we understand the value and at the end of the day, we wanna be your partner and not your enemy and I know it feels like that a lot of the time. Unfortunately the six of us can’t
service the entire affiliate market but at the end of the day, that’s our long-term goal and vision. In the short term, as Niket said,
our Help Center has great resources. If you are spending within the
several thousand dollars a month, reach out to the help team and they will triage within
our different areas of support in market and they can help you out. I will definitely be sending more resources
towards STM for you guys to access just because we know how valuable you are and we want to continue the relationship. As Maria said, we’ve got a lot of things
in the work for 2017. I know that sounds like we keep saying 2017. It’s only 24 days away, if I did my math right But you know there’s things we’re working on,
stay close to the STM forums. Webinars, they’re gonna be
specifically geared towards this vertical. The webinars are gonna
help you scale from $100 a day to $1,000 a day. And we’re gonna have
one-on-one support available and we’re building all this stuff out. 2017 is gonna be a big year not just for us but hopefully for the whole community
as well. Cool. I know some of the stuff
were discussed privately and it’s sounding very, very exciting. And you also know that
if it doesn’t turn out for 2017, my first question next year is gonna be, “Where the hell was all that stuff you promised?” Totally fair. I’m sure you’ll hunt me down, Hugh. Okay so I wanted to talk a little bit more, technical now, one of the things that confuses an awful lot of people
on Facebook is targeting. There are so many different ways to target. There are so many ways to target
that will just not work and fail dismally. I’ve tried most of them, I think. There are so many ways that work
astonishingly, ridiculously well when you do them and you suddenly go,
“Why the hell wasn’t I doing this before?” I’ve running into a fair few of those as well. Can you give us a guideline on how to
avoid category A and implement category B? What’s the stuff that really works? And I’ll start with Jason. So we almost always target completely broad. We just do broad and we run website conversion but before we run website conversion,
we run post engagement to give some juice to that post ID, then take the same post ID and run that post ID through website conversion. Okay. So when you say targeting pool,
do you literally like, US 18 to 65? I mean, based on the product? We do our due diligence on our product and that’s actually a big thing
that most people should do but they don’t do. We take a product or an offer, whatever, and we break it down, do all the demographic research and we say, “Okay this is appealing to
people 35 to 50” or whatever and then we target that. We don’t bother with 18 to 65. We use our due diligence to determine the appropriate demographic and
then we’ve run broad on that demographic. Okay. Alright. So, you guys know your CRM data
better than we do. Let’s be honest. You know how to spice it up. We did an ads engineering study
about a year ago of the different targeting features on Facebook. Interest targeting, partner categories,
broad, custom audience, custom audience lookalike. And custom audience lookalike
outperformed all of the other ones by about 30%. In terms of lower CPA’s, higher scale. So knowing your own
CRM data and customer base and then building lookalikes off of that has been the strongest thing
we’ve found over the past few years and I highly recommend that
being a strategy you test. Just out of curiosity, show of hands, who’s doing niche targeting
and finding it successful in this room? Define niche targeting in this case? Like targeting grandmothers who like motorcycles. Interest targeting. Okay. Okay. And then lookalikes, how is that working for you guys?
Show of hands. Alright. Okay. A little more. I think you guys should
continue down the lookalike path just because how our model works is you take your seed, we’re mashing over
tens of thousands of characteristics of similar people within that seed
to find them. And what’s great is it doesn’t necessarily
need to be country specific, so say you have a CRM from the US and want to test it out in Portugal or whatnot. You can run that lookalike in the other country and we’re just looking at the characteristics,
not necessarily the people within it, so I would play around with that. The one thing you wanna be mindful of is do not layer on
additional interest targeting. You can exclude recent converters. Sometimes it works, sometimes it doesn’t, but because our model’s dynamic in real time, the more you throw into it,
the more confused it gets. So the cleaner the better. Simple seed audience with the lookalikes and then just test different percentages. We’ve just seen that to be really successful. That one I just want to highlight
because that one that blew my mind when Maria mentioned it the other day. You actually recommend, rather than just going highest possible lookalike, you actually go like 1% versus 3% versus 5% with a split test
and see which works best. Exactly, yeah. A good business phasing thing to do for people who want to
build those custom audiences is spend $500, have your lawyer
do up a data management agreement and send that to your advertiser
and be like, “I need 3,000 fucking emails.
Here’s the paper. Sign this.” And it makes them feel better and it’s more likely
you’ll get that data right away. So I’m big on custom audiences
and lookalikes as well, but I come from a train of thought,
or school of thought of letting the data speak. So when we launch campaigns, we’ll go abroad, we’ll do interest targeting. I’m very good about interest targeting. I try to find one or two interests
that make up that group and try to have that audience
larger than 700,000, let’s say. I think that we see
success with niche targeting. We saw a few hands go up because at a niche-small level,
you can make that successful but if you’re looking to build the scale you’re gonna want to go broad. So being able to test broad
and having success with that, being able to test custom audiences,
1%, 3%, 5%, 10% and having success as you scale that up and then working with those
larger interest groups, I mean. At the end of the day, if you’re trying to
get to a goal CPA at a blended place maybe the quality is good with the lookalikes but you can drive good volume
from the interest target so just trying to figure out
where that position is and testing within everything. Yeah you can do that by looking at the
custom audience and audience insights to be able to pull out those interests. That’s a really easy way to do it. One of the things, on top of targeting,
that comes up a lot is the ad account health. A lot of people feel like that certain ad accounts, no matter what you do,
they’re always gonna perform poorly regardless how good your targeting is, and I think that’s a misconception. It’s not really true, but it does matter how
the ad account is performing to start. I think if you’ve got a lot of misleading ads, a lot of bad things that
we’ve been talking about earlier on an ad account, then that ad account
automatically can’t convert to go and be the best ad account
that you’re gonna perform, It’s really a matter about understanding
and starting from a really clean place, and just running really good advertising
and learning from there and iterating. You can only go up, and I think there is a misconception that certain ad accounts just naturally, no matter what
you do where they come from, have different performances. It’s not really true. They all start from the middle. They all start from the same spot. So this is gonna be a case
of people seeing random data and coming up with explanations,
as humans do. Okay. So a follow-up question that, I suspect that certain affiliates
in the audience, their Facebook accounts may not be 100% clean right now, so what should they do? Should they just be like,
“Okay, I need a new account.” Or should they be,
“Let’s find a way to clean up.” Is it possible to clean up
a low-performing account? Yeah, there’s definitely a way
to clean up a low-performing account. I think running very good content. I think running good ads,
good media that’s not misleading to really good, high-quality landing pages. We do track not just a click on Facebook but where
it’s going. The post click conversion. That does really matter to us and we track that,
not just for performance marketing but for content publishing to see, “Is it click-baity?
“Is it just a misleading experience?” and we do the same across both. I think other thing to do is
just create a business manager and really build one or two ad accounts
in there that are really clean and start and really build that business. This conception over buying ad accounts? It’s not something that is
gonna last for a long time. It’s something that’s gonna go away. I know it’s a churn-and-burn game and if you’re in it for the next 2-3 months, that’s great, go do what you gotta do, but we’re probably not gonna see you here
in 2017, at Affiliate World. Yeah and I can build on that a little bit. Something actually Marilla, I hope everyone caught her talk today
cause she crushed it, about having enough conversions
when you start off a new campaign, so even if your account might be within our realm of
not wanna get you delivery, a lot of advertisers
not necessarily game the system, but you need at least 30 to 50 conversions
per adset on the first day. So what we find really successful
if you want to launch a new campaign is bid a little higher or let our auto bid mechanism
go into play and that way you gather enough
conversions so on day 2, you can start adjusting the bids
where you feel comfortable. You’re going to spend a little bit more
like day one, but you need those conversions to have your campaigns calibrate correctly. That’s the biggest thing and the biggest misnomer that
advertisers don’t really think about. It’s like, “Oh yeah, great, I got 10 conversions today and I’m very profitable.”
$100+ payout CPL ad, but those 10 conversions
aren’t going to work for our system and then your campaigns are going to start to die out every day. So the more you can get on day one,
the better off you are for performance and it’s just something to keep in mind, just going really hot
and then dial it back down and you’ll see success that way
and I think Alex and Jason? Yeah. We tested that. We test adsets at $1000 an adset. Yeah I think that’s the big thing. Being comfortable with testing. You’re not gonna get your ROI on your first dollar.
It’s just not going to happen. There’s gonna be an investment period. Our algorithm, as it is, takes 3 to 4 days to learn on what works, what doesn’t work so I think, if you’re thinking that you’re gonna put $100 in and make a profit
on the first hundred dollars, it’s just not true. I think any business doesn’t work like that and I think you’ve got to be comfortable with investing money
to figure out what works, what doesn’t work. There’s a lot of great content
from our friends at Teespring, Shopify, on how to do this profitably. Just learn and test and learn from that. That’s why you’re seeing all this variability. So you’d run a little campaign like,
“Oh this is great!” You run a different adset, like,
“Oh my god I did terrible!” But it’s just like 5 conversions here,
10 conversions there. It’s just not enough data. So this brings me, sorry. You go first. One more thing, because I know a lot of you guys, you’re interest targeting in the room, a lot of the problem we find is like, say you want to target people who are
interested in like Wall Street Journal, New York Times. They tend to break out their adets
so they’re individually targeting all these publications so they’re getting less conversions but it’s technically the same sort of people
who want to consume content, if you think about it? So group them together
just to make the audience broader, and that way you’re likely
to have more conversions than really being granular and it will just
feed our system so much better. Are there, I know some of the audience are
going to be new affiliates who maybe can’t just spend, you know,
“Ok, new campaign. $1,000.” Are there any hacks
for smaller affiliates who are looking to get started but want to achieve
statistically significant results? And I see we have a
new member of the panel here. Thank you. Good to meet you. An interesting hack
that came out of Dublin actually is running page post engagement
with the pixel attached to it. So you basically are going after how we look at different users on the platform. Start really small and get all those conversions
within the page post engagement and then switch over to oCPM
for conversions or website conversions just because you don’t want to start fresh with the website
conversions unless you have a really big budget you can push, so start with that and then scale it up. That’s something I can totally
share with the STM forum. I think it would be really helpful for you guys. So just just to clarify, so you’re starting looking for engagement? Like brand awareness basically. And then once you have the conversions? You have the pixel tied to it
so you’re still tracking conversions and once you get enough in there, switch over and we’ve seen 300% increase in conversions on a very small scale, like spending $100 a day
going up to like yeah. It’s all about the social clout
on those posts, right? Ensuring that those posts do have a high traction,
high engagement right off the bat and then converting it to, like what Maria
was saying. Yeah, I think that pixeling every single page and what Maria’s saying, and Niket’s saying is that Facebook wants to see
30 to 50 conversions fire on that first day. If it’s not a full conversion
and it’s on page number three where you’re collecting first name and last name
and that’s a quality indicator for you and you fire a pixel there and tell Facebook that’s your conversion
that you’re optimising to because you know you’re gonna get
30 to 50 pixel fires there, I mean, that’s a good way to
slightly game the system. It’s still a valuable metric for you
and you’re gonna be able to push traffic there and actually send that signal. Might be at a cheaper cost for you as well. That actually brings me to one very specific point that we’ll talk about quickly
because we’re running out of time. Squeeze pages on Facebook. If you’re wanting to build an email list the popular perception is that you
just can’t run a campaign with a squeeze page. Is that accurate or can you actually run ads that are basically sending people to a,
“Please sign up to my email list” page. So there’s nothing wrong with
signing or doing a lead generation campaign on Facebook. We have our own product
called lead ads product that is fantastic at capturing leads off of
Facebook because we’ve already got the emails and phone numbers and names of a
lot of people that are already on the platform. Nothing wrong with it. It’s how you set it up. It’s ensuring that,
if your asking them to sign up for a lead, they know what they’re signing up for, if there is a trial aspect to what it is
that they’re signing up for, that there’s a clear checkbox
that they’re opting into to understand that they are gonna
be receiving a trial and after 14 days or 31 days,
they’re gonna be charged $99. That’s what we look at
from a misleading perspective. Capturing leads, capturing emails,
even capturing credit card information, there is nothing wrong with that. Trials work great on Facebook. I think these guys
will talk about it that as
well. Did you mean just email submits? Put a nav on the page. It can’t be a single page but let them be able to go somewhere else. I think just being as
transparent as possible of what the offer subscription services, terms and conditions upfront, We won’t have a problem with it. That’s the biggest thing that trips up advertisers. So I have a big question which I have to ask. Okay say that you’ve done something wrong, your ad’s a little bit too affiliate-y or you don’t know what happened but suddenly you come in one day,
your Facebook account’s banned. What are the steps that you should
take at that point to get your account back and not have to do anything shady? So I think right off the bat is really questioning yourself like, “Was there something misleading on my landing page, either on the landing page
or even on the ad?” We consider both the ad on Facebook
and the experience post-click. If you do feel that it was a false positive
or something that shouldn’t have been disapproved, Definitely send a task via the help at form. There is a form on
everyone’s business manager ad account. That does get looked at. I know sometimes it seems like it goes
into a black box, but it really doesn’t. There is a team that looks at that stuff. Send it through there and they will look at it to try and get the ad account back up. If not they will give you a reason
as to why it was disabled. If you, even beyond that,
still feel like there was something misleading or incorrect disapproval,
you can send it to our general inbox, which is [email protected] which goes to Maria, myself
and a couple others on our team and we will take a look at it. But again, if it’s something that you’ve already got
a response back from our team saying that this was shut down
for a valid reason, it’s not much we can do on that side. What do we do if, yeah, we screwed up and it was genuinely, you know, we put it in all honesty but we were perhaps reading the warrior forum a few too
many times that day and yeah we look at it we go, “Yeah, no, that wasn’t legit.” What do we do then? Make a new ad account. It kind of depends on how bad the violation was. There’s definitely different degrees of it. A lot of the times, some of our advertisers have old stuff in the pipeline that accidentally
goes live and we advocate to our policy team explaining, “A new campaign manager came on.
Didn’t know. So these ads went live. We’re gonna kill them”
And make that argument. I know not everyone has a rep in this room, but if you do, take advantage of them and explain that might be the situation. If you were trying to cross that gray line, you’re probably gonna need
a new ad account. Okay. Don’t go buy one. Honestly, don’t buy ad accounts. I just want to reiterate that a thousand times. We’re cracking down on it. I know it may not seem like it, but if any any true affiliate marketer
will tell you from 12 months ago to today, it’s a much different world. It’s only gonna get more strict. Create a business manager. Everyone has the ability to create
three different business managers so go to business.facebook.com, create one and start fresh. That is the official Facebook-approved way
to create new Facebook accounts. Yes. Within reason. Okay unfortunately, much as I would love
to keep doing this for hours, we have completely run out of time. I think we ran out of time
quite some time ago. So I would just like to once again thank our panel so much for having the
patience and the tolerance to listen as I badger you with questions. Thank you very much to Jason!
Thank you very much to Maria! Thank you very much, Alex!
Thank you very much, Niket! For your time. Thank you, Hugh. It’s our pleasure here, as always.

One comment on “Scaling Performance Marketing Programmes on Facebook | AWasia 2016”

  1. Ari Gold says:

    Well its 2019 and still we are facing alot of policy issues even when running compliantly on Facebook

Leave a Reply

Your email address will not be published. Required fields are marked *