Pearmill

pearmill

Part of Pearmill’s Facebook Advertising Playbook Before we get into the CTR rule, it’s important to talk about the lens we use when it comes to Facebook Advertising. There are plenty of different ways to use Facebook Advertising for growth, however, we’re specifically looking at it as a — meaning all the campaigns we run, and all the rules in our playbook series are meant to have the user perform a specific action (e.g. make a purchase, install an app, etc.) performance marketing channel This is an important distinction, because every metric becomes related to its positive or negative effect on helping us reach the ultimate performance goal. With that in mind, let’s take a look at one of the first ratio-based metrics in Facebook Advertising: Click Through Rate. Baseline CTR Rule The main reason we look at the Click-Through-Rate metric is to help us judge the relevance of the Ad Unit and Creative to the audience that our Campaign and/or Ad Set is targeting. Our rule is based on campaigns we’ve run from budgets started at $5 / day to $50,0000+ / day — so the rule can be applied to a wide range. However, a good rule of thumb is that the less money you’re spending on a campaign, the higher your CTRs should be. Rule If a link ad has a link CTR less than 0.75%, you should consider swapping it out with a new ad unit. You should de-prioritize changing the Creative or Ad Unit if the CTR is above 0.8% and close to 1% or higher. Caveats You should only judge the CTR of an Ad Unit or Creative once it has passed the . Enough Impressions Rule You should be comfortable with lower CTRs in the case that the post-click Conversion Rate is higher than the average you see across your campaigns and when costs are sustainable. For example, an ad unit may not be the most engaging — but it gives a great deal of context. In this case, when someone does eventually click on it they’re more likely to convert than people that came from other ads. It’s also okay to be at the lower-end of this range if you’re using self-selective Creative (which we’ll talk more about in the future). Thank you for reading our audience selection mini-guide (in our series ). If you’ve landed here and would like to get the rest of the series, (or below). There’s also a mini-glossary at the bottom of the post for clarity on terms we’ve used! Facebook Advertising series please subscribe here Mini Glossary — the number of times an ad has been seen (counts multiple views by the same person) Impressions — the number of people that are being shown an ad Reach Size — the grouping of Ad Sets, usually indicates targeting and/or Creative direction Campaign — the grouping of Ad Units within a Campaign, usually indicates a sub-target within the Campaign and/or Creative direction Ad Set — the image, copy, and/or video combination that’s displayed as an individual Ad Unit Ad Unit — Individual pieces of an Ad Unit (i.e. image, copy, and/or video) Creative — the number of clicks over the number of unique impressions CTR

Facebook Ads Baseline Click Through Rate Rule

Facebook Ads Playbook: Enough Impressions Rule Part of Pearmill’s Facebook Advertising Playbook The most common mistake we see advertisers make is not having a clear framework for knowing when to make a decision regarding an Ad Unit. A significant part of our rules series will focus on giving you clear guidelines on how and when to make decisions. We want to be clear enough so that you never have to ask yourself if you’re waiting too long. Or are you not waiting long enough? Is this Ad Unit doing well after a day? Two days? 10 Clicks? 100 Conversions? 2k Impressions? Note: there’s a mini glossary at the end of this post for all the capitalized words and phrases we’re using as a way to keep everything consistent across our rules series. Enough Impressions Rule The main purpose of this rule is to make sure that we aren’t making changes to an Ad Unit, Ad Set, or Campaign too early or too late. Our ruling here is based on years of experience, though it’s not at all statistically significant. As with all things around Facebook Ads, you should take it with a grain of salt. However, we can confidently say that the number below is the amount of Impressions you should be comfortable with. minimum Rule Before making a decision about any Ad Unit, Creative, Ad Set, or Campaign — you should have AT LEAST received over 2,000 Impressions on it. This assumes that you only have one Creative, one Ad Unit, one Ad Set within the Campaign. Caveats This number doesn’t linearly scale across Campaigns that have multiple Ad Units or Ad Sets, so we do recommend looking at the data on an Ad Unit basis to keep things simple. To reiterate — this is the minimum and not a maximum. If you generally have low CTRs or Conversion Rates, we recommend a higher minimum to ensure more confidence. If Facebook is in Learning Mode and you usually have high Conversion Rates, then we recommend waiting the couple of days for the Campaign to exit Learning Mode before making a decision. Though, we don’t stick to this too strictly. Unfortunately, as with all things advertising, it’s all contextual. Thank you for reading our audience selection mini-guide (in our series ). If you’ve landed here and would like to get the rest of the series, (or below). As mentioned, there’s also a mini-glossary at the bottom of the post! Facebook Advertising series please subscribe here Mini Glossary — the number of times an ad has been seen (counts multiple views by the same person) Impressions — the number of people that are being shown an ad Reach Size — the grouping of Ad Sets, usually indicates targeting and/or Creative direction Campaign — the grouping of Ad Units within a Campaign, usually indicates a sub-target within the Campaign and/or Creative direction Ad Set — the image, copy, and/or video combination that’s displayed as an individual Ad Unit Ad Unit — Individual pieces of an Ad Unit (i.e. image, copy, and/or video) Creative — the number of clicks over the number of unique impressions CTR

Facebook Ads Enough Impressions Rule

Part of Pearmill’s Facebook Advertising Playbook In order to understand why we use certain strategies for Facebook Advertising — we think it’s important for a brief lesson on how the Facebook marketing engine works. Facebook Pricing and Pacing Dynamics The lovely people at Facebook have explaining how their system works but we’d like you to remember two simple rules that define their ecosystem: already done a great job Each impression (when an ad unit is shown to a user) is priced based on how many other advertisers are currently bidding on each specific user / audience. Facebook uses a discounted pacing algorithm, which means they’ll attempt to give you the users within the audience that will be the cheapest (i.e. are the least sought after) as much as possible. Because of these two dynamics, it then follows that the more you spend (and the larger the size of the people who you’d like to reach), the more expensive any campaign gets as you’ll be competing with more advertisers (#1) and you’ll slowly run out of the cheapest users within your audience (#2). As a marketer, your job is to go to battle against all the other marketers in the ecosystem to win against those market dynamics. Through a rigorous process and approach (at least as of 2018), this is still possible. The good news is that you don’t have to be the best copywriter, graphic designer, videographer, or even statistician to be a great Facebook Advertiser — you just have to be systematic, follow the data, and continuously double-down on whatever is working. Audience Selection In general, you should attempt to have a diverse audience strategy. This means you should: Go after multiple general audiences (i.e. if you’re going after millennials, attempt to go after specific segments of them) Find different ways of targeting the same audience (e.g. if you’re targeting teachers, target people who job titles of “teacher” in one campaign and in another campaign target people who have interests in teachers’ associations. Why? Because when you have diversity of audience — you’re doing as much as you can to take advantage of Facebook’s discounted pacing, and finding more specific ways of targeting your audience (usually) helps with conversion rates since you can provide more optimized creative for each segment. With that said, there are a few methods to use when it comes to coming up with targeting options: Segmented Targeting This is the most broad to target audiences. You should attempt to target your audiences based on interests they might have, behaviors they share, and/or their demographics. (e.g. age, location, etc.) Broad targeting solves for the fact that you’ll just have more people in the bucket, which means there are potentially lots of cheap users for Facebook to target for you. However, it also means that unless you’re selling Pepsi — there will likely be a lot of false positives. You’ll want to use “selective” creative (which we’ll describe more ) to filter folks out and other audience targeting options (described below) to help with this problem. in future series With all that said, consider the following when you’re creating audiences: — who are your competitors? Target people who are following them. Competition — target followers of other products that your users are potentially using or buying. Popular Market Products — target specific job titles, employers or industries that fall within the users/buyers of your product. Employment — target interests directly related to your product. Interests — use this to target the specific part of the market that you believe has the highest likelihood to buy your products (i.e. if you’re selling luxury goods, you should go after more affluent areas) Income levels Custom Audiences If you already have an existing set of potential users and customers (i.e. if you’ve had a newsletter) then you can use them on Facebook to create an audience. Facebook attempts to match emails, phone numbers, names, and addresses to their dataset and lets you target those who were matched. If you have such a list, this is likely the easiest way to get started and get easy wins. The main issue with this approach is that if you don’t have a substantially big list (the minimum is 100 on Facebook, but anything below 500 is rarely useful) then the upside is rather small. If you do have a large list, then we even recommend segmenting the list with any internal data you may have (or enrich it using tools like FullContact) and targeting each segment separately with relevant content. For example, you may want to segment your list based on activity, recency, or any other factor you think is relevant. Note that you can use custom audiences in Lookalike audiences described below as well. Lookalike Audiences Facebook gives marketers the ability to create audiences that look like other audiences (i.e. in their behaviors, interests, etc.) — which can be a great tool. We end up using Lookalike audiences almost as soon as we have some success with other campaigns, or if we have access to custom audiences. You can create Lookalike audiences based off: that have been fired before (i.e. Purchases, Add to Carts, Achievement Unlocked) App Events or Pixel Events — based on any internal lists / data you have. Custom audiences Consider the following when creating lookalikes: — we’ve rarely had success beyond 1–2% lookalikes. Keep it simple here (unless you’re spending hundreds of thousands of dollars a month, in which case we really hope you know what you’re doing by now) Percentile — if you’re scaling up your spend (or just have 10s of conversions per day), then we recommend limiting your lookalike window to the last 7, 14, 30 day increments (when based on App/Pixel events) Timeframe — you can narrow down your audience bu only targeting people who are interested in “X” within the lookalike to be even more specific. This usually helps increase the quality of users that come through the audience. Layer Audiences Retargeting You can retarget users who have engaged with your ads in some way, have been on your website and/or app, or are existing customers of your’s. You should retarget in scenarios where you’d like the user to take a specific action (usually the next step in your funnel, or a repeat purchase of some kind). Sometimes it makes more sense to think about your advertising strategy in multiple steps: — campaigns that make your potential user aware of your product / service. You just want them to take some simple action here (watch a video, like a post, etc.) Awareness — now that you have a set of users that are aware of your service, you’d want them to start engaging with you through a more complex action (e.g. signing up for a newsletter, downloading an e-book, getting into a trial) Engagement — with an engaged audience, you’d like to have them make the final action you’d like them to take (e.g. sign up) Purchase / Conversion For simple products or low-priced products, you can actually merge #2 and #3 when it comes to retargeting. But for complex products (like high-priced SAAS products), it may be worth using a more comprehensive strategy. For the first step of this multi-step funnel approach, you can use all the audience ideas above. However, for the second two steps you can target people based on: — retarget users who have done something that constitutes as “awareness” (e.g. read an article, landed on a specific landing page, etc.) App/Pixel Events — retarget users who have engaged with one of your awareness ads (e.g. commented, liked, shared, watched, etc.) Engaged with an Ad — you can use custom audiences whom you believe to be past step #1 already (or whom you’d like to have repeat an action) Custom Audiences Structure & Learning It’s important to keep track of audiences that have worked, ones you’ve already saturated (which we’ll explain in more detail later on), and ones that you’d like to test out. So we recommend the following: — either on Facebook and/or in a spreadsheet, keep track of all the audiences you’d like to test out each month. This helps existing and new members of your team learn about what you’ve done faster, and also take over the account easily if you need them to. Audience Lists — outside of what you know about your users, you can use Facebook’s Audience Insights Tool to get event more data about them after you get some conversions. Use it to come up with other sub-segments of your audience you think you should be testing. Generate Ideas Thank you for reading our audience selection mini-guide (in our series ). If you’ve landed here and would like to get the rest of the series, (or below). Facebook Advertising series please subscribe here

Techniques to build relevant Audiences for Facebook Advertising

It’s quite common for our clients and the larger community of growth professionals to come to us with this question: how well are doing in terms of performance on cost and creative? At — a tech-enabled ad agency — we’ve spent the last couple of years managing budgets from small to large for a variety of clients in different industries. From direct to consumer, to B2B — we have first-hand knowledge of looking at these numbers and believe they should be shared for the larger community to use as benchmarks when judging their internal team or agencies they’re working with. Pearmill That’s an easy thought to have, but it’s rather complicated — because these numbers fluctuate a lot on a per industry and company basis but I’ll do my best to provide a clear picture on how to measure the success of campaigns based on what we’ve learnt over the past couple of years. Creative Performance One of the main performance metrics we always keep an eye on (sometimes programmatically) is around how well the creative is performing. The main metric we take a look at here is Click Through Rate (or Tap Through Rate). This turns out to be one of the only metrics that isn’t very industry dependent, though we look at some secondary metrics to judge it. We are usually comfortable with creative that has a CTR of . That’s a pretty big range, so let’s set some ground rules: 0.7% — 3% First off, at a minimum — we should be aiming for creative that’s close to . Anything below that, we usually consider pausing the creative altogether or using it with another audience. 0.7%-1% Secondly, we’re okay with a higher CTR (or one that’s at the low-end) as post click. CTR is a function of the familiarity and relevancy of the creative to the audience that it’s being shown to — but if it’s not relevant enough to the product or service that’s being advertised, then higher click-through-rates aren’t very useful. long as the Conversion Rate holds So the simple rule here is to . aim for high CTRs that maintain or increase down-funnel conversion rates Click to Action Conversion Rate Another metric to always monitor is the post-click conversion rate of an ad. This is calculated by the total number of unique conversions divided by the total number of unique clicks. This metric is fairly industry dependent, and on top of that — very highly influenced by comment rates on the social post, the audience, and the relevancy of the creative that the user is clicking on. However, with all that — we’ve seen some trends in performance over the years. The metric turns out to defer widely based on the goal and down-funnel action you’d like the user to take: Lead / Form Completion For consumers, prosumer, or B2B companies that are attempting to get a user to fill out a simple form (simple input such as name, email, and company) — we’ve consistently seen an average of conversion rate from click to lead completion. 15%-30% As per all our benchmarks here, this is a rather wide margin — but usually with poorly set up campaigns, they turn out to be well below 15% or just roughly at the bottom of this range. So you’d like to be achieving around 20–30% conversion rate on average and above that on your best campaigns. Install For campaigns that are attempting to get people to have users install an app on their Android or iOS device (and not looking any further than that), we’ve seen an average of from a tap/click to an install. 20%-35% Note: installs are tracked as “opens”. So even if a higher % of people actually install the app, only a portion of those users will ever open the app — which is how this metric is tracked. Product Sign Up For campaigns that are working towards getting the user to go through a user on-boarding (either after an install or a landing page view), we’ve seen an average of completion. 20%-50% This range is higher than we’re comfortable with, but it’s very dependent on the relevancy of the audience and the medium. For mobile apps, the metric tends to be closer to 50% for well designed onboarding experiences, where for landing pages — it’s definitely on the lower end of this range. Purchase / Payment For campaigns that are attempting to get the user to enter any sort of payment information and/or complete a purchase, we’re seeing a very low range of . This fluctuates depending on the value-prop of the product, the relevancy of the product to the user, and the product’s price and commitment. 1%-3% For great e-commerce campaigns, this trends towards the higher side and usually, those are the brands you’ll hear all about in the press as well. Having some name recognition helps with increasing the conversion rate of a direct response ad. Cost of Install Performance For app install campaigns for iOS and Android apps, we’ve seen a different range of installs that are somewhat industry dependent — but we haven’t seen a consistent pattern to present it to you per-industry. A few years ago when Facebook was still in its early stages, we all remember getting very cheap installs — but currently, we’re seeing a range of per install. $5-$30 Android tends to always trend on the cheaper side (if all devices are included), but becomes more or less the same as iOS if you’re targeting high-value device types. The range is mostly dependent on scale of spend and how broad the app is for the general Facebook audience. Cost of Acquisition Performance Given the previous data we’ve provided, it’s somewhat possible to achieve the numbers below with some simple math of looking up average impression and click costs on Facebook — but to make it easier for you, these are the costs we’re seeing for different action goals. Business Products When it comes to advertising to businesses and selling them a SAAS product, we’ve seen a clear division of costs for the type of action we’d like them to take: Sign Ups If the main goal is to have the user to sign up and start using the product (usually a free trial of some sort), then we’ve seen a cost depending on the industry. $50-$200 For highly competitive industries, or very niche industries — the price tends to be closer to the high-end of the range. Leads If the goal of the campaign is to receive demo requests and bookings to show the product to slightly larger companies, we’ve seen a range of per lead. $30-$200 This has been a great way to help sales teams get an extra flow of leads to work with, and we’ve seen companies of any-size being prone to being able to be reached through this method. Consumer Products There’s a large array of actions consumer companies attempt to incentivize the user to do, though the most common of them are Sign Ups and Purchases: Sign Ups For sign-ups (i.e. creating an account for the product, signing up for a newsletter, etc.), we’ve seen a rather unhelpful range of depending on the strength of the brand and value-prop for the consumer. $2-$20 This is a useless benchmark for us to give out since it’s very dependent on the subsegment of the consumers you’re going after and how painful the problem you’re solving for them is. The more unique and differentiated the product, the cheaper this turns out to be! Purchases Getting a consumer to put their credit card somewhere and order something turns out to be very unpredictable as well. It’s a function of the product price, market, and the strength of the brand. We’ve seen a range of in the past year on variety of products and services being sold to consumers. Unfortunately, the range here also varies widely due to the same factors — unique product differentiation, product’s price, and competition in the market for the product! $6-$120 What is Pearmill? is a digital advertising performance agency focused on user acquisition for technology, and consumer product companies. Pearmill We’ve worked with a large range of clients of different types and believe in our internal process that helps us achieve great results. Started by and — our team has been working on growth for 7 years on average! me (Nima Gardideh) Karim El Rabiey If you’re looking for help growing through paid acquisition, feel free to . reach out

Benchmarks for CTRs, CPIs, CPAs, and Conversion Rates for Facebook Ads

Understanding why CPA-pricing is sometimes used and why you should likely avoid it. At — a tech-enabled performance agency — we’ve dealt with many scenarios where our client wanted to use a Cost Per Acquisition (CPA) pricing model. (Note: if you don’t know what CPA pricing is, there’s a brief explanation at the end of the article) Pearmill Having been on the brand side (when I was the Head of Growth at ), and now being on the agency side — where we’ve helped brands like , , and many more — we’ve come to realize that CPA pricing is almost never the right choice. Checkout 51 Frank & Oak Cover Why you shouldn’t use CPA pricing. The main argument against CPA pricing is the alignment of incentives — the advertiser and the agency’s goals aren’t fully aligned in the long-term, and this leads to miscommunication and wasted performance that we believe can be avoided with better pricing models. Ineffective Long-Term Spend With CPA pricing, the advertiser’s spend does not get more effective over time — which is typically the goal of performance marketing. If the does a great job at getting valuable users for cheap, the efficiencies aren’t passed on to the advertiser. As the cost of acquisition goes lower and lower, the agency makes more and more money, and you don’t see any of those performance gains on your end since you’re paying a CPA price. marketing agency Short-Term Focus Instead of the agency’s focus being purely on performance — it’s split between how much money it can make in relation to the performance they can get. The allure of more revenue or the concern of not making enough can be distracting for an agency. This sometimes results in efforts that garner short-term gains that aren’t necessarily sustainable in the medium/long-term. Minimal Experimentation CPA pricing penalizes experimentation. If there are campaigns that are already crushing the set goals — it’s unlikely the agency will want to continue setting up other experiments to maintain the good performance since it could reduce the margin. Eventually, well-performing campaigns will be depleted, and the cost of acquisition for those will go up. Without constant experimentation, the advertiser and agency will be caught scrambling trying to spin up new campaigns. I’ve personally experienced this when we agreed to a CPA pricing model with a client. We were $70 below their ceiling CPA per user. This meant we were making a great amount of money from the campaigns, but we became afraid to test other things because it would reduce the margin and result in lower fees for us. We realized that this was bad behaviour for us, and for our client, and we actually convinced them to switch to percent-of-ad-spend pricing — even though it meant less money for us in the short term. The ideal environment you want to set up for your agency is one that doesn’t penalize your them for trying new things out. Taking that penalty away fosters more initiative, and ambitious thinking — even if the campaigns are already performing well. What makes CPA pricing so attractive? Both sides of the table have some incentives to opt-in to CPA model, though we believe them to be short-term — they’re worth understanding and considering when hiring an agency. Advertiser’s incentives For the advertiser, it’s all about budgeting your growth in the most precise way possible and getting a strong assurance that the agency is able to keep you growing steadily at a stable CPA. Since you somewhat know what your maximum CPA will be, you can model out your product’s growth and what rate you’re able to sustain with your budget. And since the agency makes their money from arbitrage in the CPA price, you can assume they’re pressured to achieve the goal and maintain strong campaigns that meet the goals. Note: in practice, this could all get fuzzy as agencies aren’t always capable of maintaining a steady CPA as your budget scales. Agency’s incentives For the agency, there’s a huge potential revenue upside and cost savings on ad operations and creative work. Depending on the terms of the deal, the agency could be making a lot of money if there’s enough budget for it and they’re able to get very cheap users. For example, if the flat CPA is set to $100 and the agency figures out a way to acquire users at $40 — they make $60 / user! That’s a pretty good margin. After successful campaigns are achieved — the agency can scale down experiments and let the campaigns run on autopilot for a bit without having to assign too many internal resources for ad operations and creative production. Note: in practice, it’s not all sunshine here of course — if the performance is harder to achieve, then the upsides are harder to reach. Conclusion Although advertisers may find comfort in acquiring users for a fixed CPA, they’re likely missing out on higher quality users or gains on cost efficiencies that the agency is discovering. CPA-based pricing will reduce the focus put on the long-term value of a user, and focuses the teams to only meet short-term goals. We almost always recommend other pricing models. Either a flat monthly retainer (for smaller budgets), or a fair ad-spend percentage that can be set in stone based on the current budget and cost of acquiring customers. It’s paramount to align the agency and the advertiser’s incentives, and CPA pricing fails to deliver on that in the long-term. If you’re looking for advice, or are thinking about working with agencies — we’re happy to help. ! Just give us a shout What is CPA Pricing? For clarity’s sake let’s define Cost Per Acquisition pricing, how it’s defined and calculated. CPA pricing is when a marketing firm or advertising channel commits to a flat cost per acquisition cost for the advertiser. They either guarantee it (very rare) or structure the contract so that the advertiser doesn’t pay the agency if the average CPA is higher than the set limit — this is usually called an arbitrage model. For example, let’s say the brand and marketing firm agree to acquire iOS app installs at a $10 CPA pricing with an arbitrage model for the firm: this means the firm will do whatever they can to keep the CPA below $10, and the difference between the actual CPA and the flat CPA will go to the marketing firm. To put it in numbers, if the firm acquires 1000 app installs at $8, then they are paid $2,000. If they acquire 1000 app installs at $11, then they’re not entitled to any payment. Why I have an opinion For the past 7 years, I’ve been deploying ad budgets on digital platforms in some shape or form. Be it Facebook, Google Ads, Bing, Pinterest, Snap, or smaller networks. I started my career running growth at various startups, and eventually a company called , where I spent millions of dollars on ads and hired agencies to help us achieve our aggressive user acquisition goals. I left when the company was , to start a unique performance agency called Pearmill. Checkout 51 sold a few years later At — we help brands grow through advertising. We’ve spent millions of dollars in the past couple of years for brands like , , and many more — using our own technology for reporting and ad operations. Pearmill Frank & Oak Cover I’ve been on both sides of the table: hiring agencies as an advertiser, protecting my budget tightly in a competitive market, as well as working with our clients as an agency attempting to achieve high growth goals!

Stories, News & Mentions about PEARMILL

Pearmill’s Stories on HackerNoon