Acquisition Channels

This is page five of the Growth Marketing Guide.

Acquisition channels

This page shows you where to acquire customers from.

We'll cover paid channels (e.g. ads) and unpaid (e.g. content).

Before you read this page, read the bottom half of this handbook's first page to learn which channels are likely a fit for your company.

Channels covered

We'll cover:

Which channels should I prioritize?

I consider three factors when prioritizing acquisition channels:

1. Channel cost

If you have small margins—say you're selling a $10 toothbrush—you are unlikely to succeed with paid channels.

For reference, Facebook and Instagram (FB/IG), which are two of the cheapest paid ad channels, can't drive e-commerce sales within North America and Western Europe for less than ~$15. 

Not being able to make ads profitable is why many businesses rely on unpaid channels, such as SEO, referrals, product-led growth, and social content.

Ad channel costs

Here are rough costs per click for U.S. audiences:

These numbers vary wildly depending on the audience, ad quality, and more. I hesitated to include them, but at least they provide you a loose range.

Channel affordability

Here's how to calculate if a paid channel is affordable for your company:

Let's say our hypothetical paid channel is Facebook ads. If your Facebook click cost rises by just $1.50 to $4.50—with no increase in conversion rate—you're now paying $360 to acquire a customer. Suddenly, Facebook ads are unprofitable.

That's the volatility of ad channel revenue, and that's why ads are typically the domain of companies with expensive products whose sizable margins buffer against volatility.

If you can't afford a channel

If you've concluded that a channel isn't profitable, you have three paths:

2. Audience fit

One channel may be twice as expensive as another, but if its audience converts three times better for you, it's a more affordable channel.

How well a channel's audience fits your product depends on four factors:

To summarize: For ads to reach scale, you're at the mercy of whether your target audience exists on a channel, whether you can accurately target them, and whether you can make compelling ads that belong on that channel.

3. Channel volume

The last criterion for prioritizing channel tests is whether enough of your audience uses that channel.

Here's a comparison of audience volume across channels:

Note: This data is always out of date.

You might think, "Aren't all of these channels big enough to run ads successfully?"

Unfortunately, it doesn't matter how large a channel is. What matters is how large your audience is on that channel. If you sell to a very niche slice of it—say, senior HR leaders in Texas—you risk quick saturation and high CPAs.

Saturation occurs when a channel's audience is tired of seeing your ads. This happens because your audience is small relative to your ad spend, so they've seen your ads many times and they start ignoring them. When this happens, ad clickthrough rates plummet and cost per customer acquisition (CPA) rises. The channel is no longer affordable.

(The solution is to suppress your daily ad spend to keep ad impression frequency low. You trade short-term volume for long-term affordability. I talk more about saturation on the Running Ads page.)

How volume differs on ad channels

Every ad channel offers targeting capabilities that reflect how users engage on the platform. You use this behavior as a proxy for identifying who's most likely to buy from you.

For example, Facebook users list their marital status and work history on their profiles. They also Like artists and various recreational activities. Therefore, Facebook can let you advertise based on marital status, work history, and Likes.

Let's compare this to Google Ads: users enter search keywords. Therefore, Google lets you target searchers based on their keywords. Or, on Pinterest, you're targeting based on the image categories people search and bookmark.

As an advertiser, your goal is to identify which channels surface user behavior that  can act as a targeting proxy for their interest in your product.

Behavior targeting versus profile targeting

There's no more direct way to target a customer for a men’s fashion company than to show ads to people Googling right now for "men’s clothing."

And that's exactly why Google Ads performs so well.

I call this behavior targeting: someone is seeing an ad based on their real-time behavior. This contrasts with Facebook's profile targeting, which uses profile and past engagement as a proxy for their current interest.

This presents Facebook with a timing problem. On Facebook, you may have Liked something related to fashion, but it could have been two years ago. Or maybe it was this morning. Who knows. In the eyes of Facebook's targeting, they might be treated as the same.

While behavior targeting is higher-converting, it's not a panacea: You’re at the mercy of how many users take the time to proactively search for what they’re interested in.

Not everyone wanting to buy clothing necessarily Googles for it.

For example, some clothing buyers skip searches altogether and go directly to their favorite blogs. Or they walk into stores.

In other words, if you target exclusively via behavior channels, such as Google Ads, many potential customers will never see your ads.

Further, people who might buy from you but for whom buying is not currently top of mind won't be searching the web at all.

That's why profile targeting is a necessary complement to behavioral: With Facebook, you can target someone likely to be a good fit for your product then bring them up the Ladder of Product Awareness. In other words, profile targeting gives you the ability to expand your immediate market.

Channel targeting and saturation

Let's revisit the idea of channel saturation. Here's how it applies to targeting:

Where to start with ad channels

The more you charge for your product, the easier it'll be to get ads working. However, you should still set your expectations low: most companies never make ads profitable—and they switch to unpaid channels.

For those who do get ads working, the most reliably performant channels are:

If you want help running these ad channels, check out Demand Curve (link above).

Prioritizing channels

Don't test channels randomly. Have a process to increase ROI and move quickly. I use two techniques for prioritizing channel tests:

ICE framework

Score every possible channel on three variables each with a scale of 1-10:

  1. How much positive Impact the channel would have if successful — Further, if the test fails, can the labor at least be repurposed for another project?
  2. The Confidence you have that it will succeed if attempted — How de-risked is this channel? Has anyone gotten this to work before?
  3. How Easy it would be to try — How much time and resources would it take?

Add the numbers together then divide by three to get the average score for each channel. Repeat this exercise for every channel then rank them by their scores.

Channels at the top of the list have the highest expected value and should be prioritized.

Unpaid channels

I want to share a framework for prioritizing unpaid channels in particular: persistence versus hit-or-miss channels.

Twitter is what I call a persistence channel: If you post good tweets consistently, compounding follower growth is guaranteed. It's just a matter of persistence.

In contrast, Product Hunt, reddit, or Hacker News are examples of hit-or-miss channels. You're at the mercy of that day's competition for the front page, and you'll rarely hit.

This has significant implications on which unpaid channels we prioritize.

Persistence channels

Persistence channels are the unpaid channels you want to prioritize.

Hit-or-miss channels

You're probably wondering, Do big hits on hit-or-miss channels make up for all the misses? Not really. It's still better to prioritize persistence channels:

  1. Hit-or-miss growth doesn't compound — A hit-or-miss platform indexes content and de-emphasizes the content creators. This means that having a post go viral doesn't reliably increase the odds that your next post goes viral; you're not accruing compounding gains. Compare this to a persistence channel such as Twitter: If your tweet goes viral, you'll gain new followers who retweet your future tweets. That's a compounding flywheel.
  2. Hit-or-miss ROI trends toward zero — After you repeatedly hit the front page of a leaderboard channel like Reddit or Product Hunt, you tend to wear out your welcome.

This doesn't mean you shouldn't post on hit-or-miss channels. Instead, it means you should treat them as accelerators instead of originators: post content on persistence channels then you cross-post breakout hits to hit-or-miss channels to shoot for an extra boost.

Strong marketers recognize the unbounded upside and compounding returns of persistence channels. Weak marketers overvalue hit-or-miss channels.

Recap

Here are the criteria you're using to prioritize channel tests:

Then use ICE to assess:

For many companies, these exercises result in prioritizing Facebook ads, Instagram ads, Google Ads, SEO, and referrals. However, I recommend testing nearly everything over time. Leave no stone unturned.

A reminder that the last page of this handbook has a downloadable cheatsheet that handily recaps most of what you're about to learn.

Ad channels

Google Ads and Facebook/Instagram Ads

To learn Google Ads, I recommend starting with their official guide.

For Facebook and Instagram (FB/IG), I dedicate an upcoming page to that.

Instagram ads 👍+

Instagram (IG) is the best profile targeting channel. This mantel recently belonged to Facebook, but people are engaging less with Facebook ads.

IG users convert the best after an ad click; they tend to buy at the highest rate.

IG is an especially good fit for mobile apps and ecommerce goods. But it's generally a poor fit for B2B SaaS and enterprise companies if your site or app isn't mobile-friendly.

For many of my clients, Instagram is their only profitable channel. 

Instagram is part of Facebook

On Instagram, you target audiences and create ads just as you do on Facebook. In fact, you advertise to both via the same dashboard.

The difference is in their ad formats: Ads look different on Instagram, so you should rewrite and redesign them accordingly.

Large audience

Instagram has more users than Twitter, LinkedIn, and Pinterest combined. This means if you get Instagram ads working, they'll sustain a higher daily spend than other channels (except Facebook).

Further, there's a sizable 18-24 audience that's engaged with Instagram but not Facebook. So if you target this demographic and you're only running Facebook ads, you'll fail to fully reach your addressable audience.

That said, people of all ages use and post-click convert on Instagram.

Users are in the right mindset

Users open IG for the purpose of having their attention grabbed by something new. They're open to discovery and distraction.

What a lovely time to shove an ad in someone's face. Timing is key.

(In contrast, on Facebook, users are expecting familiar content from familiar friends. So, ads from companies they've never heard stand out like a sore thumb.)

For this reason, (good) Instagram ads significantly outperform Facebook mobile ads for ecommerce goods and app installs.

Subtle text means images must stand alone

Notice how in the image above, you have little room to pitch your product via text. The text component is buried at the bottom and is given little focus

This means ad performance hinges on how well your image or video grabs someone's attention for a reason related to your actual product. Clickbait for the sake of clickbait is not how you run Instagram ads.

If you're selling a product that is not self-evident through imagery alone (e.g. a SaaS tool), you need to put text on top of the image to describe what you're selling. Or use the IG video ad format to show your product in action. 

Whatever you do, clearly depict what you're selling with your IG ad.

Not a fit for desktop software

IG is almost entirely accessed via mobile. That makes it a good fit for ecommerce goods and mobile apps, but a poor fit for desktop SaaS software.

Even if a visitor can sign up for your service on mobile then return on a desktop via an email reminder, it's rarely cost-efficient to introduce this friction into a funnel. 

For desktop apps, stick with Facebook's desktop ads and Google Ads.

LinkedIn ads+

LinkedIn offers profile targeting. Like Facebook, several distinct ad units are offered. 

Some units are impossible to make work, and others perform okay at best. But if you're a B2B company earning thousands per customer, test LinkedIn ads.

LinkedIn offers a uniquely granular company and employee targeting capability. In contrast, FB/IG and all other social ad channels are terrible at targeting people based on their work profiles.

On LinkedIn, you can target by a company's industry, employee count, location, and more. Plus, you can target employees based on their role and seniority. You can even target just employees working at a whitelist of companies you supply.

If you're targeting niche industries or niche job titles, LinkedIn is your best channel.

LinkedIn economics

Unfortunately, LinkedIn ad clicks are so expensive ($12-$20 USD), that they're only affordable if you're selling high-priced products or services.

In fact, if you don't earn over $10,000 USD in the lifetime of a customer, you probably cannot afford LinkedIn ads. That's a conclusion from seeing millions of dollars spent on LinkedIn. 

This threshold unfortunately rules out most companies.

What works on LinkedIn

B2C companies selling financial products have a shot at making LinkedIn ads work.

However, LinkedIn ads are primarily for B2B companies. Those with the best results typically offer education- and job-related products—since that's largely why people visit LinkedIn in the first place.

On your landing pages, it's best to implement an aggressive email capture system. LinkedIn visits are so expensive, and they're so finely targeted, that—more than ever—you want to move the relationship forward via email and eventually to sales.

LinkedIn's ad units

LinkedIn offers many distinct ad units. Experiment with all of them:

Pinterest ads+

Pinterest ads don't work for most companies.

You have the greatest chance at making them work if:

Pinterest can work for some B2C ecommerce goods. 

However, Pinterest can be a broader fit if you use it for retargeting, which is the tactic of showing ads to uncoverted site visitors to get them to return and buy.  Retargeted visitors convert at a lesser cost than new visitors, so you can sometimes make weak ad channels viable through cross-channel retargeting.

Pinterest targeting

On Pinterest, you can target users based on the content topics they engage with (e.g. fashion, furniture) or the keywords they're searching.

So, like Google Ads, Pinterest has behavior targeting! And Pinterest's behavior targeting can refer a sizable amount of traffic. Double yay! We growth marketers need as many behavior ad channels as we can get. They're the saturation antidote.

Unfortunately, while Pinterest's cost-per-click is low, its users generally don't convert once they arrive on your site. They bounce without signing up.

I chalk this up to the way Pinterest users interact with "pins:" They rabidly open new tabs then later return to them while forgetting why they opened your random site. Then they leave instantly. Plus, some users click on pins so quickly that they'll sloppily click ads unknowingly.

Therefore, straddle the line between two objectives:

  1. Don't make your ad look so much like an ad that people reflexively ignore it.
  2. But, make your ad look enough like an ad that those who click do so purposefully.

Blend in, but don't mislead.

Twitter ads 👎+

Like Facebook and Instagram, Twitter is mostly a profile targeting channel. It saturates.

(While Twitter does offer behavior targeting, most Twitter users don't use Twitter search. They just skim their Following feed.)

Twitter ads are over twice as expensive per click than Facebook and Instagram, and they convert at a worse rate. Further, it has far fewer engaged users than FB/IG, so it saturates quickly. 

Twitter is a bad ad channel. 

However, it's a great brand marketing platform if you're an enterprise company wanting to stay top of mind with potential buyers. But, this handbook doesn't teach brand marketing, so I won't dive into that.

Twitter's redeeming feature: hyper-niches

Twitter has a huge array of niche interests you can target by. For every lesser-known company, interest, or person, you can target people who Follow them.

That's unique.

On Facebook and Instagram, you're unable to target cultural interests with this specificity: You can only target interests that have reached a threshold of public interest. For example, if a newspaper's Facebook Page has fewer than, say, 25,000 Likes, they won't appear as a targetable option when making ads. And even if they reach the threshold, Facebook might still not make them targetable.

So if you're selling to a niche audience that can only be accurately identified based on who they Follow on Twitter, then Twitter is worth testing. 

But it'll still only work if you earn a lot per customer. I'm talking thousands of dollars. 

As a reference point, in the last four years, the only time Twitter ads worked was for a client earning six figures annually per customer. This client had fewer than 500 companies in the world who were a fit for their product. And I noticed that employees at those 500 companies followed the same few influencers in the space. 

So I targeted those influencers' followers and got a few to turn into sales leads.

Reddit ads 👎+

Reddit ads are typically a waste of money. Their targeting granularity is too broad to be effective or too niche to be scalable.

I'd only consider reddit ads if a product appeals very broadly (e.g. underwear, credit cards) or if you earn a lot per customer and your niche is active in a subreddit.

How reddit advertising currently works

Like most modern ad channels, reddit ads are native: they look like organic posts and are differentiated by a subtle "promoted" label.

Targeting works on either an interest- or subreddit-basis. 

Interest categories include Travel and Technology, for example. These show your ads to redditors with browsing behavior that falls into these categories. 

Unfortunately, this targeting isn't narrow enough to home in on your best-performing audiences: Putting aside how ineffectively broad an interest like "Technology" is, you typically need to narrow B2C audiences via gender and/or age to make ads cost-efficient. And reddit's demographic targeting is lacking.

Your alternative is subreddit targeting: running ads within niche subforums, e.g. German films or StarCraft. But these saturate extremely quickly.

Snapchat ads 👎+

Snapchat ad clicks are cheap, but the on-site conversion can be 5-10x worse than equivalent campaigns on FB/IG.

Snapchat users rarely take out their credit cards. They're coming for free stuff. (Even the wealthy adults who use the platform.)

This is a shame because the ad creation experience on Snapchat is wonderful, their support is top-notch, and their targeting criteria is granular.

I've seen two types of companies be successful on Snapchat.

Consumer mobile

If you're a consumer-facing mobile app or game with a free trial, Snapchat has a shot at working within 50% of FB/IG's performance.

Geotargeting

Snapchat lets you target users within 0.3 miles of a physical location. If you're a retail business that lends itself to impulse purchases (e.g. restaurants, entertainment), Snapchat may be a fit. You can capture people as they walk by.

Display ads 👎+

Display ads are the banner ads you see across the web. 

They perform terribly. 

People don't click banner ads anymore. They haven't for years. And those who are weird enough to click banners aren't usually the people who then buy on your site.

However, display ads are still good for cheap brand marketing. As in, advertising not to get clicks, but to stay top of mind.

Consider how Verizon and T-Mobile perennially spam every site on the web with banner ads. They aren't getting a lot of clicks, but they want to stay top of mind above their competition for the next time you're considering a new provider.

Or consider how movie studios aggressively spam sites with banner ads before a film's release. You're not expected to click the ads and convert into ticket-holder on the spot. Studios are simply making sure you keep the film in mind when you're bored this weekend.

If you do run display ads, do it through Google Display Network. It provides the most granular targeting and is therefore the most cost-effective when run well.

Display ads work for retargeting

Display ads are modestly effective at retargeting. (Retargeting is showing ads to visitors who've been to your site—to get them back.)

For this purpose, they can complement your Facebook/Instagram retargeting by showing your retargeting ads across the web at-large.

How do I make ads on these channels?

I cover ad creation over the next two pages: Make Ads and Run Ads.

The Demand Curve growth program also provides in-depth guides to making ads. You can also work with our agency, Bell Curve, which will run your ads for you.

Unpaid channels

PR

Most companies never succeed at making PR a major, recurring acquisition channel. The few that do fall into one of three buckets:

If you’re successfully doing any of the above, reach out directly to the tips hotlines at the publications in your market. (Tips do get read.) Write a concise email that opens with one of the above hooks.

Twitter Organic

Twitter (next to SEO) is the best persistence channel. Meaning, if you consistently tweet with above-average quality content, you're guaranteed to grow. It's also the most de-risked path to building influence and authority in your space.

To grow an audience on Twitter, you only need to know two things:

  1. How to post tweets that appeal to your target audience.
  2. How to post tweets that get retweeted.

On Twitter, you get new followers by being retweeted. Being retweeted is how content gets exposure outside of your existing follower base. Some percentage of the new eyeballs then follow you.

I polled my Twitter followers to ask why they retweet. There are four key reasons:

1. To bookmark a tweet for later.

2. To give kudos to a novel idea that was elegantly articulated.

3. To signal which side of a debate someone's on.

  1. In their words — "It helps me reinforce the narrative I project for myself.""It's a reflection of the image I'm portraying on my Twitter account.""I want my followers to know I relate to this statement.""See? I'm not the only one who thinks like this."
  2. Implications — Choose a side in an argument. State a novel justification for your side being better. Say it in a way that cuts like a knife. You're looking for a mixture of substance and bite.

Try to have each of your tweets fall into one of these categories. Throughout, focus on being concise, witty, and novel.

The following tweet types rarely lead to new followers: polls, memes, and videos. Your focus should be on written actionable insights.

LinkedIn Organic

LinkedIn is a phenomenal alternative to Twitter for companies wanting to build an organic B2B audience that they can later convert into customers.

Here's a thorough guide to building an audience on LinkedIn.

Product Hunt

If you're a tech startup, Product Hunt is one of the easier hit-or-miss channels to appear on. It's good for a one-time boost of 1,000-10,000 tech-savvy visitors.

Here's a thorough guide to launching your business on Product Hunt.

Content marketing and SEO

Content is appropriate for every business. In a landscape of sales fatigue, authentic content stands out and helps you build trust. Always test content.

Content and SEO are covered on the last page of this handbook.

All other channels

There are many more channels, but few are as big and effective as those covered.

But, at some point, everything is worth testing. It’s the job of a growth marketer to test new channels and run experiments on a recurring basis. Be in a position to discover pleasant surprises. Ultimately, finding just one scalable, profitable channel can make or break a business.

Next: Making ads

How to write and design ads that get people to click.

Next page →