Facebook ads course: How to use automation to scale with Revealbot →

How to use Revealbot for agencies

Share

Transcript

Brenna: Welcome everybody. Welcome if you're in the live chat and hello if you're watching this in the VOD, really appreciate your time. Just to give you a little bit of info about me. My name is Brenna Kleiman. I am one of the senior media managers for paid social at Hawke. I am also the training manager for the media buying department as well as one of the VIP account specialists for media buying in general. So what I specialize in is special accounts that either wanna grow very quickly or have extremely high budgets and need a slightly different approach. Some of the clients that I've worked with in the past include like Lenovo, OPI, Milani Cosmetics, LA Lakers. I've done strategy for them, and so on and so on. That's sort of my background.

I've been at Hawke for almost a year and a half now. And it's been a wild ride. So I'm excited to show you guys what I've been working on lately. The first thing I wanna go over is just like to give you guys a basic overview of what we're gonna be looking at, how I'm currently using Revealbot, how my clients are using it, sort of, how my team uses it and then how to start using it yourself. And then we can get into some questions and have a little bit of a back and forth with Reggie over some of the strategies and approaches.

Reggie: Yeah. And if you guys wanna put in your questions down, go ahead and we'll get to it at the end.

Brenna: Yup. Please do. So wanted to give you guys the results first, show you guys the good stuff. So the main account I want to show you guys is a case study for this. They came to Hawke in I believe August, so Q3 of 2019, and they said, "We wanna grow absurdly fast for Black Friday, Cyber Monday, and for the holiday season," because they have a very seasonal business. They sell jewelry. And we said, "Okay, let's figure out how to do that."

So basically we ended up implementing a strategy towards the beginning of...I guess, beginning mid-October and early November to really scale as quickly as possible. So this is the short of the ROAS, the revenue and the spend for this particular account over the course of Q4. So you can see at the beginning of October, we started at a fairly low spend. Over time we scaled to spending over, you know, 20K in a day in the course of about 2 weeks. And then from there we brought it back down manually to give it a little bit of a respite before Black Friday, spiked again for Black Friday, and then started to bring it back down as the shipping deadlines passed for the holiday.

So couple things I'm happy about with this account in particular. The fact that we managed to grow their spend by over 600% in 3 months is pretty fantastic, quarter over quarter. We also increased our purchase volume by almost that same amount as well. And then also increased our ROAS while increasing that spend by over 50%. So, all this was done pretty much just with Revealbot and very minimal input from me. I set up a structure and Revealbot really took care of the rest and really managed to continue to grow ROAS. And we can see that the line is actually just trending upwards. So it is continuing to get better and better as we go along despite us increasing spend on a regular basis.

So here are the rules that I used to do that. There's a couple ones that I really enjoy using. The first one that I think is really important is ad rotation. So the first thing that I wanna know is, are my ads working? And if they're not, I want them to stop. I want them to stop. So I figured out for this particular client when we should start to see particular benchmarks. So by the time an ad has spent $50, I should start to see it get a 0.8% return or a 0.8 ROAS. By the time it hits 100, it should have a 1.1. By 150, it should be 1.4 and so on and so forth. By the time it hits $300, it should be at that 1.8 that we're aiming for. So our goal for this particular account is if we hit 1.8 we can scale. So anything above 1.8 is considered profitable and worth expanding on.

So set up the ads there. If it's paused then don't worry about it. And then if it actually reaches that threshold, then you can actually set the name to add it as #winner. And then I know that that's actually a winning ad and I get an export of those every day in Excel. So I actually am able to track what the winning ads are over time and see if I can see patterns. I also included the schedule of these over on the left side because I think they're important. There's a couple of them that are running on a particular schedule. This one I have set to run every 15 minutes and we can get into a little bit more about why I love that as opposed to Facebook rules, in a little bit later. But for right now, the main point is I want to know as soon as an ad isn't profitable and then turn it off to minimize losses. Because when you're spending 20K in a day, that $50 goes fast. So if you're not paying attention or if you're only checking your account every couple hours, you might actually lose a few thousand. And that's really difficult to recover from. So highly recommend having us run as continually as possible.

The next one is the start. So Facebook has a weird way of doing attribution which means that if an ad is paused, it might still get sales attributed to it within that 28-day click, one day view timeframe. So we call those zombie ads. They come back from the dead. So what we do is basically institute a reverse rule from the pause one that says, well, if it gets over that threshold, turn it back on.

Another safeguard I have here is to make sure that we're not turning on so many ads that Facebook can't spend enough on any one of them. You know, we wanna make sure that we only have about six ads running in an ad set at a time. And that's kind of the maximum you wanna be using to make sure that Facebook can allocate a significant amount of spend to your ads so you can rotate through them quickly.

So I have basically a rule in here that says to start an ad, if there are fewer than six active ad sets...or ads in the ad set and the lifetime spend of that ad is at the bottom. So I loved that Reveal actually implemented rankings. There's a whole bunch of cool stuff you can do with that. The way I use this is really simple. It's just saying if there's an ad that hasn't had any spend yet, let's test it out in this ad set and give it a shot. And then the zombie ads also only turn on if there are fewer than six active ads in the ad set. Again, I have this running every 15 minutes because I want it to be able to counterbalance that pause. And I wanna make sure that all of my ad sets are constantly running at their capacity and running at maximum efficiency.

Reggie: And Brenna, I remember we were talking about this before we hit record, if you could go back to that slide really quick. And Justin had this question too. Can you explain a little bit more how you determined the spend amounts and the ROAS thresholds that you use for each of these stages?

Brenna: Yeah, absolutely. So a lot of this came from past consumer data and a lot of analytics. So we had a lot of history in the account before Hawke started working in it. So we had a lot of back data to work with. We also had a lot of information about how the site itself worked and what the purchase funnel was like. So for this particular account, we know that a lot of the purchases are impulse buys, so we know that they happen fairly quickly. And we also know that some ads take a little bit longer than others to make a return. So we wanna make sure that we're actually giving them a fair chance to get that. But the thresholds were set from information we got from the analytics of the account. So your mileage may vary. I recommend like figuring out these numbers individually on account by account basis because this is not a one size fits all solution.

Reggie: Got it. Thank you.

Brenna: The next thing is the scaling part, which is the fun part. The first part is making sure you don't lose money. This is the, how do you make more money part. So there's a couple different sets of rules that I have for budget increases and decreases. The main two that I have are differentiating between a 5K or less per day budget and a 5K or more per day budget. And the reason for that is once you get over 5K and you're dealing with percentage lift, percentage raises, you're getting into the thousands of dollars and it iterates and rises pretty darn quickly and you might want to be a little bit more cautious there.

So, for this one, I have increased the budget for any campaign by 10% once a day, if the campaign has spent more than 80% of its daily budget yesterday. The reason for that is there are some campaigns that due to deliverability issues or the fact that they haven't been running very long might not be actually up to the point where they're spending most of their daily budget. So I don't want Revealbot to go in and say, "Hey, this account or this campaign that only spent $50 and made a $200 sale has a 4x ROAS, so I'm going to dump a bunch more money into it." I wanna make sure that it's actually using what it currently has before I give it more.

And then I have my thresholds for ROAS. Again, your mileage may vary here, but since I know that my ROAS that I'm aiming for is 1.8, once I hit 1.9, I know that I'm kind of in a comfortable place to maybe start pushing the budget a little bit higher. And so that's my 10% increase. But if it's over 2, I can actually increase it by 20%. So again, this is if the daily budget is under $5,000. And you'll notice in the bottom left here that I actually only have these set to run once a day. And the main reason for that is I don't want my campaigns changing their budgets constantly because that's a really easy way to send your campaigns back into learning phase which is a money pit essentially. The less time you can spend in learning phase, the better. So making sure that you're setting your budgets at a consistent time and at a consistent pace and no more than once per day.

I also recommend setting your budgets early in the morning. If you set them after noon, Facebook does some wonky stuff with daily budgets and it can cause spikes and weird spending habits for a couple days. So always set your budgets in the morning and make sure that it's only once per day.

Reggie: So I had a question about this actually. So if the ROAS is between 1.9 and 2, it's raised 10% once a day, and if it's over 2, it's 20%. What was like the reasoning that you found to have such like, I guess a small window for the 10% budget increase.

Brenna: So the reason for that is we wanted to scale fast and see what happens. You can change these percentages. And honestly, once you're...if you're under 5K in terms of daily budget, 20%, if you're looking to scale rapidly isn't a huge change. Facebook recommends depending on who you ask, it's anywhere between 10% and 30% as the threshold for sending your ad sets back in the learning phase. Twenty percent is kind of the safe number. So 20% is the maximum that I'll ever increase my budgets by per day. So that's kind of the maximum that I can push it. And since we have a pretty small window there, usually I don't see my ads like my campaigns going over to per day, day over day because that's a pretty serious increase. Usually, it'll be a 10%. And in very few cases it'll be 20.

Reggie: Okay, cool. Cool.

Brenna: The next part of this is the over 5K per day, very similar. It's pretty much the exact same rule except the percentages have changed. Because once you get over that 5K like I said before, you're increasing extremely rapidly, so you maybe wanna mitigate that a little bit, even out that curve slightly. And again, that runs at 8 a.m., same as all the others. So I actually get a nice Slack message in the morning that says what all my budgets have done over the morning. So when I get into work, I can see sort of where I'm at for the day. It's a really good way to get started.

The next part is the budget decrease. So what happens if you put a ton of money into your campaign and it tanks? So this is the fail-safe for that. So the important part of this, and I think this is where Revealbot really excels is being a dispassionate observer. There's a very human instinct when you see things start to decrease in terms of performance, where you go, "Oh God, I need to fix all of this right away." And you cut the budget in half or you, you know, make it 10% of what it was, it was $10,000 and now it's $1,000. And while that might seem like the right move, because you want to, you know, protect your losses as much as possible, in terms of the algorithm, it actually takes much longer for it to even out if you make that drastic of a change.

So what Revealbot does is it actually just reverses that. So if I have a campaign that has less than a 1.8 but greater than a 1.6, I just decrease the budget by 10% once a day. Just bring it back down a little bit, see how it performs, let it even out, we'll see how it goes. And then if it's less than 1.6, decrease it by 20%. So I wanna make sure, like, okay, I'm not super positive that that was a good move. If I'm under 1.6, that means that there's a pretty significant difference between performance from the day previous. So let's even that out a little bit. So again, 20% rule here, don't change your budgets by more than 20% per day and you'll generally keep your campaigns pretty happy and out of learning phase.

So then we get into some report stuff. So I said earlier that I get a message every morning about my budgets. This is another one of the reports that I get. So I get a morning update. And basically what this tells me is what my 30-day spend and ROAS is. So I can look at sort of what the overall performance is. And then I can also get yesterday's performance. So I get a snapshot and I get a big picture. And this is just delivered to my Slack inbox every morning. So I can see what my ROAS has been over the past 30 days, and it gives me a nice chart showing me what the progress has been.

And then I also just get a readout of the average website purchase value, spend, website purchases and ROAS. So I can see, you know, if our AOV has increased, if our spend has increased or decreased compared to the day before, same with purchases and same with ROAS. That particular portion of the report is more likely to change significantly than the chart. And that's kind of the goal is to have the micro and the macro given to you at the same time so you can sort of figure out exactly how to respond and how to react.

And then the next part I actually use...so I didn't realize until Reggie asked me to put this together, how many different Revealbot tools that I use in the platform. I also use the bulk creation and the post ID export. So bulk creation I use to create post IDs and source ads. So this is an example of the volume of ads that I create. This was just over, I think the past week. I took a snapshot of all the different ads that I created. So I've made over 200 ads, maybe 250 over the past week.

So it's really easy to put all this together. I cannot tell you how many hours the bulk creation has saved me. And it also makes it much, much easier to AB test things. It makes it easier to know exactly what piece of creative, what copy is doing really well because all of those things are uniform across the board. Facebook is using the same image hash for everything. So when you export that information, you're actually able to gauge your performance much better and iterate on that success. Yeah. Bulk creation saved my life so many times.

Reggie: That's awesome. I love it.

Brenna: Yeah. That's kind of the breakdown of all of the different technical aspects of it. Yeah, I think, Reggie, you had a few questions.

Reggie: Yeah. Thank you so much for sharing your rules with us, sharing your secret sauce. Let's see. Let's get into gallery view here. Awesome. Well, yeah. That's super cool. Thanks for sharing. And I think what we're gonna do is after the webinar is over, we're going to turn those strategies into like a Revealbot strategy. So you guys listening, you can actually download them and use them on your own account, tweak them, and see how they work for your ad accounts.

Brenna: Yup. And I'll make sure there's documentation so you know what variables to put where and sort of how to consider those different pieces.

Reggie: Sweet. Thank you. That's super valuable. Thank you. Yes. So I had some questions here for you. And everyone listening, please feel free to type in any questions that you might have for Brenna. And we'll get to those if we have time. So one of the most obvious questions that I always get is like, what are some of the benefits you've seen of using Revealbot over Facebook's native automated rules?

Brenna: Oh, man, so many. Facebook's automated rules are pretty terrible for a couple different reasons. So number one, they don't always fire. So I've had multiple instances where I've set a Facebook rule, the condition has been fulfilled and then Facebook hasn't fired the rule for like three hours and it ends up spending like a couple hundred dollars. I've had clients come to me and say, "This rule didn't go off, you know, I lost $500." You know, that's what happens when you use Facebook rules. They just don't work sometimes.

There's way fewer parameters and metrics that you can track. So Facebook really doesn't give you a lot of options when it comes to things like ROAS, website purchases. You don't have things like custom metrics or custom timeframes, which has been a major savior. I actually use custom timeframes in another one of my rules to track CPM changes week over week.

So I can see if my CPM is increasing for a particular audience, which means I might need to switch some stuff up. And then the 15 minute versus 30 minute rule. Fifteen minutes makes such a huge difference. Like it really does, it can mean like hundreds of dollars saved. And yeah, I really, really like the 15 minutes because I get it as soon as possible. And it's way easier to get that reporting as well. Facebook is terrible at telling when its rules have run. But for mine I can say, all right, send me an email, send me a Slack, let me know if it doesn't, if it doesn't go off for some reason and why. And then I can check the logs really easily for each individual rule and be able to see how Revealbot was measuring those parameters and implementing those rules.

Reggie: And another thing too, just thinking about the AB rotation rule that you were sharing, with Facebook's native rules, you would have to create a different rule for each one of those things, right? Is that correct?

Brenna: Yeah. I don't even think that you could do it with Facebook. Like, I don't think you could do it with Facebook rules because it has ranking stuff. And it also requires you to use different timeframes within the same rule, which Facebook doesn't allow you to do either. So there's a 28-day and then there's the lifetime in the same rule that are required. But yeah, you can't do that with Facebook rules.

Reggie: Yeah. And on top of that, when you're managing so many different ad accounts, it's so much easier to manage everything inside one tool, Revealbot, where you could work with all of your ad accounts at once.

Brenna: And it's much easier to duplicate those rules over and to track with Revealbot. And thanks for adding folders. It made a huge difference.

Reggie: Awesome. Glad to hear it. So yeah, so you're using Revealbot for most of your Facebook ad clients if their spend allows it. And we'll get to questions about who or what are the right type of ad accounts to use Revealbot for. But I wanted to ask you first, with using Revealbot for so many of your different clients, how has it improved your agency's overall efficiency?

Brenna: Oh man. In some ways it's hard to quantify and in some ways it's very easy to. Really what Revealbot does for me and for my team is it saves time. It saves time and it also makes sure that you're able to implement your vision as perfectly as possible. I don't have to be in my accounts on a half-hour, 15 minute basis. I don't have to be watching everything constantly which gives me more time to train my team. I have a bunch of direct reports that I work with. You know, I've got meetings to go to. And that's a really big part of it.

And also just being able to spend more time on strategy to be able to look at my rules and go, okay, how did this affect this thing? Maybe I can tweak this little piece and set it forth and see what happens. And having that space, that mental space to be able to work in a proactive way rather than a reactive way, makes a world of difference for myself and for my clients.

Reggie: Awesome. That's really good. So when you take on a new client at your agency, what is kind of like the process for implementing Revealbot for a new ad account?

Brenna: Sure. So there's a couple considerations for implementing Reveal. The first thing really is a function of spend. You can implement Reveal at any level of spend. It really doesn't matter. The difference is in how minute you can be in your rules and implementation. So for instance, if you're spending, you know, $100 on your ads, you're not really gonna have a whole lot of rules to implement, but if you're spending 50,000, 100,000, 200,000, that's a lot of variables to keep in check. And a lot of quickly moving objects. So you need to be able to track those and be able to manage them efficiently.

So that's one of the first things I take into consideration. Also, I need to know if their pixel is seasoned. And by that I mean does it have a significant number of events in it? From an optimization perspective, that's really important because Facebook tends to flounder a little bit when it doesn't have events that it's being told to optimize for. So having a little bit of foundation laid makes it much easier to implement Reveal because you have more consistency in that performance.

And then from there, it's really what the overall strategy is. So, you know, do you want to...do you want to do a bunch of tiny audiences? Do you wanna do like a few big ones, and sort of what is the vision for the company moving forward and what are those business goals? But yeah, from there, it's pretty much implementing a couple fail-safes and then based on those business goals, honestly, a lot of it is bespoke. There are some ideas that I approach it with, but every client's different. Everybody has a slightly different need and a different way of doing things. So it's a different challenge every time. But it's cool.

Reggie: And earlier when we were talking, you were mentioning something about how not only were you gonna take spend into consideration, but also making sure that the client knows like their business metrics really well. Can you talk a little bit more about that?

Brenna: Yeah. So it doesn't matter if you have a great strategy. If you have a strategy and no goals or the wrong goals, you're going to end up in the wrong place faster. So I really do a lot of foundation work with my clients to make sure that we have established goals. I'm one of those people who really love smart goals. I always make sure that my clients, you know, when we establish that 1.8, I was like, all right, just very clear. If we hit 1.8, we can scale. And we had that...like we had it written down, like that was our goal. That's what we were gonna hit. And then we can go from there. And it becomes much easier to measure your success.

And then once you hit that goal, you can start to reevaluate and maybe set a higher one and you can start to see your marginal improvement. And that's really the best way to do it. Because, A, you know that you're accomplishing what you need to accomplish rather than just sort of sending money out into the ether in the hope that it's going to come back to you. And then it also gives you, you know, a sense of sureness that you're kind of moving in the right direction. Like there's no sort of ambiguity there. But yeah, having goals, having very clearly defined goals is really important especially when you're dealing with automation because robots don't do qualitative information very well. They're not particularly good at feelings. So if you can give them numbers, they're very good at giving you numbers. If you give them a, "I don't know, I'd like to make money," that's not gonna fly.

Reggie: Yeah. What was your saying that you were mentioning before?

Brenna: Oh, robots are...I mean, automation is only as good as the buyer that made it, and robots aren't, you know...what was it? One of my coworker says that humans are bags of meat with feelings and robots don't sleep and don't have any feelings at all. So if you set them forth and tell them to do a task, they will do nothing else. So if you have a strategy that you wanna put in place and you know that you aren't going to need to go to bed at some point, it's really helpful to know that you have some kind of assistance along that way or some kind of augmentation.

Reggie: Yeah. And that could be a good thing and a bad thing. I think you were telling me too, that's one of the things that you always do when you're setting up new rules for an account is always check the locks to make sure that they're actually set up right in the first place, right?

Brenna: Yeah. I've definitely had some mishaps. I've accidentally deleted. I've been a little bit too cavalier with my deletion rules and accidentally deleted like a bunch of campaigns. I've accidentally increased budgets by way more than I needed to. There's been a lot of trial and error, so really do recommend doing test runs, doing trial runs. I also recommend writing out your rules by hand. And that might seem kind of archaic, but I really do recommend it. Because sometimes when you're inputting things and typing stuff in, it doesn't flow through your brain in exactly the same way. So if you're trying to really parse through something that's kind of complicated, I really do recommend writing it down and talking yourself through it because it'll help you find those steps along the way where you're like, hang on, I think I need to add another layer here. But yeah, check those logs. Make sure you're actually getting notifications through Slack or email at least in the beginning for the first couple weeks to make sure that everything's working properly. Yeah, definitely, definitely check your work with these because it can go south if you don't give it the right instructions.

Reggie: Yeah. And do you typically use like the same exact rules for every account, beyond just slightly adjusting the numbers and the thresholds?

Brenna: There are a couple different ones that I use for all my accounts. So with clients that have a fixed monthly spend, I have a fail-safe that'll shut everything off if we hit that. Or it'll give me an alert and I can send an email to my client that says, "Hey, we're already...you know, we're hitting our goals but we're at our threshold. Do you wanna increase the budget so we can make that decision?" Some of my clients do e-commerce. That's a lot of what we do here at Hawke. But we also have some B2B clients. We have some lead gen clients. And the funnels for those are slightly different.

So sometimes what you're gonna be doing is optimizing for micro conversions. So for instance, the way that I buy a bracelet is going to be very different from the way I...even Revealbot, from the way I come to Revealbot, I go, "All right, like, let me learn more about this, you know, let me talk to somebody on the team, figure out how it all works." So it's a much longer process. And if that's the case, then I need to be able to determine markers along the way that are going to indicate success. So it does take a little bit more complicated route. But the overall goal is essentially the same. It's really the question of the steps and how you get there.

Reggie: Right. And so I heard you mention a few different verticals there. For your clients that are B2B or doing lead gen, do you use Revealbot too for those or is it more so strictly for e-commerce?

Brenna: You can use it for anything. I really do think that it works for pretty much everything. The question is how well have you mapped out your strategy. Automation can be used for any vertical for any industry. It really is about understanding what your goals are, understanding what the steps are and then having an augmentation that helps you like fulfill those, I guess. So yeah, anybody can use it. It's maybe the easiest for e-com because e-com is a pretty simple funnel overall. But if you're looking at lead gen for B2B or like SaaS, it's slightly longer, but you can do pretty much the same thing.

Reggie: Right. Cool. And I know you mentioned...I mean, Hawke is a pretty big agency and I know you mentioned that you have a team and all that stuff. You were telling me earlier before we hit record that you actually found Revealbot and then implemented at Hawke with how you guys manage your clients. So my question is how do you train other members of your team to use Revealbot? Do you just set up everything for them and be like, "Hey, this is how it works?" Or do they create their own rules for their own ad accounts?

Brenna: Yes, both of those. Generally the way that it goes is introducing Revealbot in a way that's accessible, setting up some very early rules. I kind of approach it the same way that you might approach like coding training. So you start out with the building blocks and you say, "Hey, here's a simple thing that you can build right now. Here's a small project, here's the project that you can layer on top of that." And then you kind of go from there. And it enables people to do some exploration within the platform because there's so much that you can do with it.

I usually tell my trainees to approach it like you're asking a question, so what do you wanna find out? Or, what action are you trying to take? And then what rules do you have to put in place to make that happen? And that's generally been a pretty effective way of doing it. And I really do recommend doing that for like just anything really, of just exploring how you can use a tool to accomplish a task that you're trying to do. Yeah, it's worked pretty well so far. It's pretty cool.

Reggie: Right. And so has using Revealbot and automation helped, I would say...or I'll structure this question, more so with helping you guys free up time to do better on each ad account or do you guys use that extra time to take on more clients per account rep?

Brenna: Yeah. Again, yes. It really is both of those things. Some of us want to focus on having fewer clients and going in, you know, doing crazy strategy stuff. Some people want to get more efficient. It really depends on how the buyer wants to use it. And I've seen people use it in different ways. Some people just, you know, have a bunch of alerts that go off. Some people just get reports, some people just use it for bulk creation. So it really is kind of all over the map depending on how you do end up using it. But overall, it is definitely a time saver and it is definitely an efficiency creator for sure.

Reggie: So that's actually an interesting thing that you mentioned there. And maybe this is a topic for another discussion completely, but it sounds like you let the account managers at Hawke have lots of autonomy with how they do their strategy. So it isn't like, okay, this is what works generally across all accounts. We'll do something similar for all of our accounts. And so all you account managers pretty much follow this. It's not like that at Hawke.

Brenna: So at Hawke, we...actually, one of the main reasons I really like working here is the autonomy. We regularly have team trainings, we talk about best practices. We really do a lot of knowledge building. So a lot of us are working with sort of the same pool of knowledge, but the fact that we have so many different kinds of clients in so many different kinds of verticals with all different kinds of budgets and needs, it really means that there isn't a one size fits all solution. So when it comes to actual strategy and implementation, it's gonna be different no matter what. So really we do rely a lot on those account managers and buyers to sort of use their best judgment and use the knowledge that they have. And we also work with each other constantly.

We're always zooming over to each other's desks going, "Hey, does this look right to you? Is there a thing that you think I could do? Could I add an audience here?" And it's really helpful to have those kinds of workshops. So in a way, it's kind of everybody working on every account with one per person. But yeah, it is very much an autonomous sort of space. But I think that's what makes us so agile and effective in a lot of ways because we don't have like that one singular approach.

Reggie: Yeah, that's actually really interesting. And I'm not super familiar with how agencies are being ran today, but it kind of seems to me like most agencies do the opposite approach. They try and package what they do into one single thing and process and nail that down and then just try to scale that process. That's interesting.

Brenna: Yeah, it's definitely not the traditional agency model. I've worked at mid-size agencies in the past and it's very, very structured. There's a very defined flow of how things work. In the current day and age, it really doesn't make a ton of sense for a lot of solutions. Some people just really wanna do some Facebook ads and it's really not that difficult to do. So a lot of it is bringing those pieces together. So a lot of what we do is also project management. So it's a little bit of everything.

Reggie: Okay. Awesome. So we're coming up on the end of our time here. We got some questions about will there be a recording? Yes. We'll send out a recording probably tomorrow. So everyone who is here and registered will get that recording. And this is your last chance to ask a question for Brenna. And other than that, you always have the recording, you still have, so if you have a question enter it in now. And Brenna, if you could tell us a little bit more maybe about Hawke. Yeah, why don't you tell us about Hawke and what you guys do there specifically, and all that stuff.

Brenna: Sure. So at Hawke, our mission is to make great advertising accessible to everyone. And really, we really do stand by that. A lot of our clients are small to medium growth stage businesses. I've sold horse shampoo, I've sold like flashlights, like anything that you can think of. It really is a lot of different projects and passions that people bring to us. And our goal at Hawke is to help them realize those dreams. And it's really neat. We have essentially an a la carte sort of structure. So you can sign on for one service, so you can just get like media buying. You can sign on for strategy, you can sign on for email or a combination of any of those things. And it's month to month so you don't have to stick with something forever.

You don't have to sign like AOR contracts. There's nothing like that. It's much more and as you need it type thing which is also really neat. It means that, you know, if you need to take some time to build your website and then maybe you need some time to build out your email and then maybe you need some time to, you know, work on your organic social, you can do that and not have to pay for all of those services through that entire time. It makes the strategy of it a lot easier and it also makes the entry and the actual planning phase much more structured. So it's definitely a different way of doing it. And it's more modular in a sense than a traditional agency. But everybody that I've worked with is really passionate about what they do. And there's a lot of really great expertise on the team. A lot of us have worked in agencies before, so worked with a bunch of different, you know, sizes of clients. We've had a few people who ran their own e-commerce stores and just designed all those structures. So there's a lot of different points of perspective that we try to take into account and really try to rely on each other for.

Reggie: And kind of in relation to that. We got a question from anonymous. Can you give an example of how this growth has helped reach an overall business goal for a client in the bigger scope for the company for the client?

Brenna: Yeah. So I guess, one of the main things is that this growth really helped us scale our efforts. It helped us, you know, handle more clients. I think it made it so that we are able to, you know, have a little bit more time to breathe. You know, there's always going to be more people coming through the door and we wanna be able to help them out as much as possible. So as far as like, OKRs, I think really it's about, you know, being able to grow our team. So I think a year ago we were at maybe 14 people and we're now at about 22. And then also just being able to grow profitably. As far as on the agency side, we were actually in models, so we are able to scale and be profitable while, you know, adding more people to our team, which is really great. And that's really kind of our main goal is to keep our people happy and here and to help grow the agency as a whole. So it's really been cool to be able to be part of that over the past year and a half.

Reggie: Awesome. I've got one more question from anonymous. Is there a minimum budget you would recommend or that Facebook likes to see to properly AB test?

Brenna: It really depends. It depends on if you wanna do the quick and dirty version or if you wanna do the test and learn version. The test and learn version is going to give you an actual recommended budget. Usually, those tests take about four days. It also depends on what your KPIs are. So if you're going after purchase, you might take a little bit more time and a little bit more money. But if you're going after view contents or clicks, that's gonna go way faster. It also depends on the purchase cycle.

So like I said earlier, if it takes four weeks for somebody to decide if they want to purchase your product or not, they're gonna have to be testing during that full time, or you're at least going to have to wait a significant period of time to understand what works and what doesn't. So unfortunately it really does vary, but the main levers that you can pull are time, money. And then I guess the one constant would be the purchase cycle and the price point.

Reggie: And when you do AB tests, do you use like the Facebook split test feature or do you make your own tests?

Brenna: It depends. Again, the test and learn is really helpful if you have very clearly defined variables that you want to test. If you wanna do something that's more like, "Oh, I wanna see if this ad works well," and you're really just focusing on the ROAS generating like direct response driving component of it, then it's not as super necessary. It also is a function of budget. So whenever you're doing testing, you have to go in with the knowledge that it might be a loss for you. So if you have, you know, a 5K month budget and you're spending 20% of that on testing, that's a pretty significant amount that the other 4 grand is going to have to make up for. So you have to take that into account as well.

Reggie: Have you ever used Revealbot to help you with doing AB tests?

Brenna: I kind of am with the rotator rules. That's kind of what it is with the bulk creation. So as something works and something doesn't, I actually started to notice a couple days ago that there are a few different combinations that are rising to the top and keep getting tagged as winners. So that's definitely interesting in like an anecdotal sense, but as far as like statistical significance and getting super into the nitty-gritty with that, it's a little bit more complicated to do with Reveal. But it's definitely doable. Get those custom metrics going.

Reggie: Yeah, Facebook has a lot of good tools to do AB testing on like the creative side, especially with like its dynamic text options and all that stuff. We're actually coming out with a blog post very soon by our head of product Masha, who will explain how she uses Revealbot rules to test. Not like things on the creative side per se, but like how increasing budgets or decreasing budgets affect ROAS and AB testing that, and it's pretty cool. So yup. We'll share that when that comes out too.

Brenna: Yeah, it sounds awesome. Cool. "When I work with small audiences, how do I use Revealbot to detect or manage audience saturation?" That's a good question. Honestly, the main way I do it is checking CPMs. So I think I mentioned this a little bit earlier where I checked my CPMs week over week. And that's really the best way that I do that. So if I start to see my CPMs drastically increase week over week, that's an indicator to me that it's getting more expensive to reach that audience, which means I'm losing out at auction or...so there's more competition maybe that might happen depending on the season. But more likely it's because people aren't going to be as responsive to those ads. So it means that either I need to bring the budget down, maybe I need to expand the audience. Maybe I need to refresh the creative. So there's a few different things that I can do if I see that happening. But CPM is generally the Canary in the coal mine when it comes to that.

Reggie: Awesome.

Brenna: And then good creative and copy. Good creative and copy is everything. Really, if you have a good audience and Facebook is getting better and better at helping with audience selection. Really, it's about the creative and the copy. Creative more than the copy, sorry copywriters out there. But it really is a lot about the creative. Making sure that you have a very clear idea of what you want the person to do when they see your ad. It's extremely effective. So if you want them to click, tell them to click. If you want them to shop, tell them to shop. If you want them to just remember your brand, make sure your logo is in there somewhere. If you want them to know that there's a sale going on, better plaster that in every possible place that you can do it.

So there's a lot of importance and honestly I think a lot of buyers don't put enough weight on the importance of good creative. But really the principles are very simple. And Facebook has a ton of really fantastic resources on how to create good creative for direct response, lead generation, brand awareness. But it really is a lot about the creative.

Reggie: Yeah. I can imagine a lot of media buyers not like having creative as such an important thing because it's almost like everything.

Brenna: Well, a lot of them like it depends on if you find somebody who's more of an analyst versus a creator. There are a lot of people who think, oh, I wanna make really beautiful ads and I want them to, you know, work this specific way. And it's kind of a mix of the two because at a certain point, you know, if you're dealing with a copywriter who's like, I just wanna make this really amazing brand-focused ad. It's like, cool, but we need to drive sales. That is a completely different objective.

Reggie: Yeah. They are a little bit different skillsets. And that's one of my favorite things about Facebook ads in general is because you can, by just running a successful campaign, you kind of use all different aspects of skills in marketing to achieve that. So it's kind of a cool platform to learn because you're not just learning a tool but you're learning marketing as a whole.

Brenna: Yeah, definitely.

Reggie: One follow-up question to that real quick and I know we're almost out of time here, is, how do you kind of get creatives from your clients? Do you guys create creatives in-house or do you just kind of consult with the client to get their own and what happens if you just can't get good creative from your client?

Brenna: Very good question. We do kind of a mix of those things. So there are a few different options if you're working with Hawke on creative. So, number one, we either...we can get creative directly from the client, so if they have a designer, if they have some product photos, whatever, we can work with that. We'll do like some cropping on maybe like a text overlay just to try to drive the point home. We also have an in-house creative team, so we are able to develop creative in-house if that's something that you need. We can do product photos, any stuff like that. I've worked with our creative department a few times and they've made some really fantastic work.

We also are a Facebook marketing partner for agencies. So we get access to things like creative consultations where we can actually talk directly with a creative expert at Facebook and talk to them about the creative that we have and see how we can use it and how we can leverage it to the best of our ability. And then we also have access to a couple really fantastic programs where Facebook has partnerships with a few different content creators, video creators, specifically agencies, places like VidMob and Shuttlerock. They'll actually foot the bill for developing video creative. So you get a consultation, they develop a few different ads that are optimized for each of the different placements. And then you actually have some really great creative based off of your brand. So there's a few different ways that you can do that.

As far as bad creative or just creative that isn't optimized, there are always going to be a few things that you can do. Sometimes it's just as important as...it's just a simple matter of showing them how...showing your client how important good creative is. So if you have, you know, a piece of good creative that you can show them, or one fantastic image and you just AB test that with another one, that's just a simple product shot on a white background, and say, "Hey, there's literally no difference between these two except this creative." That's a really great argument for putting some money behind some design.

Reggie: Yeah, that's one thing I've struggled with before when working with clients is they'll have...they'll be very adamant about what their brand is like, and they have a creative team. They wanna use their creative team, but the things that they produce are not really focused on direct conversion or driving sales, which is their business call. And that always makes it so difficult.

Brenna: Yeah, I actually have two slides that I put in every single one of my executive summaries and my execution plans for my clients, a full breakdown of all of the asset requirements. So if they have images, here are the things that the images need to fulfill in order to be optimized. And here are the things that videos need to be in order to be optimized. So I generally have my clients send those to the creative team before they produce anything and see if they can either retrofit some stuff. And it also helps them pick out selects when they're doing shoots and things like that so they can keep those elements in mind when they're making new creative.

Reggie: Okay. Cool. All right, so we have two questions left. We only have one minute left until our time is up. I don't wanna take too much of your time, Brenna. Maybe in a yes or no question answer. Do you utilize the audience expansion feature on Facebook?

Brenna: Yes. But you should be using lookalike audiences instead.

Reggie: And do you use auto bidding as your default bid option?

Brenna: I do actually. The only time I use bid capping or cost capping is when I have a completely...when I know that I'm able to meet that bid, otherwise my delivery is gonna go through the floor. And also it messes with Facebook's ability to optimize correctly. So generally auto bidding is gonna be the way to go unless you have a very, very specific reason for using it.

Reggie: Right. Yeah. Okay, cool. Yeah, let's just end it there. We covered a lot. We're getting a lot of great feedback from people and that's awesome to see real expertise, bookmarked, love to see it. So yeah, we'll take the recording and send it out to everyone and so you as well have access to it to watch it again. Anything, any last words, Brenna?

Brenna: Awesome. If you wanna reach out to us and work with Hawke, we're more than happy to give you a consultation and talk to one of our business development people. Just send an email to hello@hawkemedia.com and we'll get in touch.

Reggie: And mention this webinar.

Brenna: Mention this webinar.

Reggie: And, you guys take e-commerce and B2B companies, are there any types of companies you don't take?

Brenna: No, we work with CBD clients. You know, our roster pretty much runs the gamut. Yeah. Talk to one of our people and we can tell you where we should get you started.

Reggie: All right, cool. So that's, hello@hawkemedia.com.

Brenna: That's correct.

Reggie: Send your email there if you're interested. Okay. Thanks, Brenna. I really appreciate it.