Ad Channels
Header image featuring woman eating, sleeping, and working out

This is page five of a handbook on Startup Growth. Begin here.


  1. Content
  2. Content
  3. Content
  4. Content
  5. Content
  6. Content

Marketing channels

This page compares the popular ad channels against each other — from Facebook and Instagram to AdWords and Pinterest.

This page is part of a three-page arc. On the next page, I cover how to write ads. After that, I walk you through how to setup a Facebook Ads campaign from A to Z. 

Once you've read all three pages, you'll know how to run ads on any channel.

First up, a primer on terminology: A marketing channel is any place where customers are sourced. A marketing channel can either be paid (e.g. ads, sponsorship) or unpaid (e.g. content marketing, SEO, virality, PR, and sales).

Of the paid channels, this handbook's emphasis is on ad channels. Because they're typically the largest source of user acquisition.

And we wanna get big.

Let's start by learning how to determine which channels are appropriate for you.

Selecting ad channels

In general, when selling a product, there are three types of fit you must achieve:

Market fit is not this handbook's focus — it's not about helping you build a product people love. Audience fit and channel fit, however, are our focus.

I'll discuss audience fit on the next page when I cover how to write compelling ads.

So let's start by breaking down channel fit. That's the hardest to wrap your head around — unless you have experience running a lot of ads like I've done.

Four factors will determine your channel fit:

1. Channel cost

Let's start with the basics before I get into the nitty gritty.

The higher your margins, the more ad channels you can afford to experiment with. Because some channels are quite expensive.

LinkedIn, for example, sits at an expensive ~$7 per North American click. On Facebook, in contrast, you can easily get these clicks as low as $0.30.

So here's the issue: If you sell a product with a low absolute margin and only expensive ad channels (e.g. LinkedIn) work for you, say goodbye to ads.

Let's examine the simple math behind that statement.

Sample math

Let's say you're targeting an audience comprised of business executives. And let's say this audience is only targetable on LinkedIn (probably true). 

This means you will need to earn a high enough average revenue per user (ARPU) for a $7 click to be profitable. Because only a portion of those who click at $7 are going to convert into paying customers. 

For example, if your SaaS company's paid conversion rate is 2.5% (this is in fact the average), you're paying around $280 on LinkedIn to acquire a customer.

If you can't make $280 back per customer within the average timespan they use your product for, your paid ads won't have positive return on investment (ROI).

Consider this: If your product costs $50/mo and the average customer uses you for 6 months before leaving, that $300 in revenue is not nearly enough — just a $20 surplus — to make user acquisition viable via LinkedIn.

In fact, the rule of thumb is your cost per user acquisition (CPA) should be no more than 1/3rd your user's lifetime value. 

I'll talk a lot more about ad metrics soon. For now, I just want you to understand that every channel has an associated cost that determines its suitability for your product.

2. Channel audience targeting

Beyond margins, you're also facing the issue of audience targeting: Is your target audience even on the channel you're considering?

Consider how your B2B technology company's ideal customer may not be the type of person who spends her day on Pinterest.

Or how your wheelchair company may not be a good fit for Instagram Ads.

Fortunately, as we'll see, Facebook Ads and Google AdWords give us enough leeway to target everyone on the planet. But, certain demographic-constrained social networks, like Pinterest and Instagram, are not a good fit for most companies.

I'll talk more about this shortly.

3. Channel audience volume

Here's another criterion for assessing an ad channel's suitability for your growth efforts: Whether enough of your audience is on the channel.

I often find that a channel drives positive ROI customers — but at too low of a volume to warrant the time investment. So I double down my efforts on other channels.

Let's compare the major channels' audience volume:

Here's the key takeaway I want you to keep in the back of your head: The smaller a channel's audience, the quicker you'll saturate it. Saturation is when the channel's target audience gets tired of seeing your ads on repeat so your clickthrough rates plummet and your cost per user acquisition rises. 

I'll talk a lot more about this phenomenon soon.

4. Channel ad units

So far we've discussed a channel's marginaudience targeting, and audience volume as reasons for a channel not being an appropriate fit.

I've saved the sneakiest issue for last: a channel's ad units. Consider how a user's typical interaction with a channel may not be conducive to engagement on ads advertising your product type.

For example: Users open Instagram looking to lose themselves in beautiful imagery that inspires them to improve their lifestyle — or to admire others. So, an ad for a product that doesn’t lend itself to striking or aspirational imagery won’t perform well. 

Your dental office's ads aren't going to get gangbusters engagement on Instagram.

They'll be out of context and go unclicked.


Stay up to date on channels

All four per-channel issues I've just discussed are subject to change.

Channels evolve over time. They release new ad units and overhaul their interfaces.

The onus is therefore on growth marketers to monitor their channels and periodically re-assess product-channel fit:

Before I move forward

We're only getting started on the extremely convoluted dynamics of running ads.

There's a reason I started a growth agency: Companies oftentimes can't scale these logistics internally. Or, if they can, they don't know the nitty gritty of every ad channel.

But please don't get overwhelmed. I've written this page as clearly as I could. It's also a bit wordier than I typically write because I want to handhold you to a professional level of ad execution over the next three pages.

Good news is I know my shit, so you're about to learn a lot you haven't read elsewhere.

The major marketing channels

For nearly every business, our focus is on Google AdWords and Facebook Ads. 

Their massive audiences and highly precise targeting make them worthwhile for all businesses (including B2B).

The way you target audiences on these two platforms is very different. And, fortunately, very complimentary. Every business should pursue both to target the entirety of their online audience.

Instagram, Twitter, and LinkedIn may also play a part in your ad strategy. It depends on your product and margins.

Channel audience targeting

I'm going to emphasize channel targeting on this page because I'll be spending a lot of time on ad creation and strategy over the next two pages.

As a base-level means of targeting a particular audience, nearly every ad channel offers demographic criteria (e.g language, country, and gender) to target by. 

So that’s your first step toward weeding out ill-suited segments of the population for your product. 

But, it’s not enough to sufficiently narrow audiences so that your cost per acquisition is optimized. (The more precisely you target your ideal audience, the fewer non-converting clicks you'll pay for.)

To get truly narrow targeting, every channel also offers a unique set of criteria that reflects users’ engagement with the platform. For example, on Facebook, users sometimes list their marital status and work history.

Therefore, Facebook lets you target users based on marital status and work history.

That's where things get interesting. And, speaking of Facebook, let's start there.

Facebook targeting

On Facebook, consider how users A) share articles and B) Like businesses, products, activities, and people. 

Naturally, Facebook permits advertisers to target people with such interests.

Therefore, the task of a skilled Facebook advertiser is to find which of these infinite interests is most closely associated with their ideal customers. For example, people who like men’s fashion may Like a well-known men’s fashion blog on Facebook. 

Now let's compare Facebook's sharing-based and interest-based targeting with Google's AdWords targeting. The contrast is revealing.

Facebook Ads versus AdWords targeting

In contrast, Google users enter keywords to search for pages that satisfy their educational, entertainment, and ecommerce queries. 

So you target users by what they're currently searching for.

Certainly, there is no more direct way to target an ideal customer for your men’s fashion company than to show ads to people searching for "men’s jackets."

It's why AdWords performs so well: People are already expressing interest in your product, and it's up to you not to drop the ball pitching yourself to them.

Now consider how, on Facebook, you’re performing profile-based targeting. And how this contrasts the behavior-based targeting of Google, e.g. someone searching for what they want. 

With profile-based targeting, your potential audience size is limited by the number of Facebook users who proactively shared or Liked content related to your market. 

The problem is many people don’t bother Liking on Facebook. (I never do.)

Further, because people on Facebook may have Liked something related to, say, Cars, either two years ago or just yesterday (you don't know which), we're always guessing at people's continued interest in the topic associated with their profile.

So Facebook profile data is — at best — a proxy for someone's interests. It is neither exhaustive nor up to date.

In contrast, with search ads, people may express their desire to buy a car right now through a "buy Toyota vs. Honda" query. Doesn't get clearer than that.

But, AdWords is not a holy grail: You’re at the mercy of however many users took the time to proactively search for what they’re interested in. Some people might skip search altogether and instead peruse blogs for recommendations. Or ask their friends. 

Let's tie this audience targeting dichotomy back to the earlier concept of saturation:

Ultimately, Facebook and AdWords are very complementary ad channels: AdWords is better for targeting users who already want a product like yours, and Facebook is better at targeting people who aren't necessarily looking to buy now but are a perfect demographic fit that you can massage into buying your product over time.

AdWords isn't difficult to master, but Facebook is extremely complex. That's why I dedicate an entire upcoming page to it. For now, I'll run through how the remaining channels (e.g. Instagram, Twitter, LinkedIn) compare to one another.

This guide doesn't yet cover AdWords. But I'll soon be adding an advanced, extremely in-depth guide to running agency-level AdWords campaigns. To be alerted when it's out, subscribe for handbook updates:

Marketing channel: Instagram AdsExpand

While this handbook doesn't walk you through setting up campaigns for every channel, I want to you to understand when to use each channel. You're going to learn a lot about how I think through channels and their unique advantages.

Instagram is a really powerful one, so let's start here.

Like Facebook, Instagram supports profile-based audience targeting. This means saturation will occur at some point. (While people do search Instagram, most of the consumption occurs through passively browsing one's feed.)

Instagram Ad Unit

Here are the characteristics of Instagram Ads:

Marketing channel: Twitter AdsExpand

Now let's walk through the characteristics of Twitter Ads. This is sorta fun, right?

Like Facebook, Twitter is mostly profile-based audience targeting. 

(While it also offers behavior-based targeting, most users don't use Twitter by searching for keywords. Instead, they passively consume their Following feed.)

Twitter Ad Unit

Marketing channel: LinkedIn AdsExpand

LinkedIn supports profile-based audience targeting. 

Let's run through its unique characteristics. I quite like LinkedIn.

LinkedIn Ad Unit

Weak ad channelsExpand

The following channels don't often work. Either their click volume is too low to warrant the time investment or CPA's are too high. Usually both.

Channel: Reddit Ads

A reddit ad for a client

If you execute reddit ads well, clicks can be very relatively low priced (~$0.15) and the ensuing site engagement can be relatively high. But, the total available click volume is extremely limited. This is due to Reddit's unpredictable ad delivery algorithms. 

(Their ads product is rudimentary and, frankly, very amateur. They're improving it now.)

Like the other ad channels, a reddit ad is a native ad — it looks like an organic reddit submission and is differentiated by a subtle "promoted" label.

Targeting works on either an interest- or subreddit-basis. Interest categories include Travel and Technology, for example, and they match your ad with redditors whose upvoting, browsing, and commenting behavior fit into these categories. 

Alternatively, you can target users when they browse specific subreddits. (Subreddits are the subcategories of reddit that are home to broad and niche topics, e.g. movies.)

The difficulty of reddit ads

I started this section by saying, "If you do reddit ads well..." Reddit ads are hit or miss. 

Here's my theory : Their ad unit is always shown in the same spot (the very top of the page), which users can then subconsciously build a reflex to ignore.

In contrast, Twitter, Facebook, and LinkedIn ads are buried among native feed content. This may make all the difference. It's impossible to completely ignore these ads.

The trick then is to ensure your reddit really don't feel like ads. They have to look and sound like organic content that just so happens to be in the top spot. 

Specifically, be especially careful to avoid sales language or hard-to-believe claims. And look at how organic reddit submissions are worded. Mimic that.

But even wording your ad perfectly will only get you so far. The problems with reddit continue: the ads appear without creative — beyond a laughably tiny thumbnail. 

So you also lack the ability to grab attention with dazzling visuals.

Ultimately, reddit is a very weak ad channel. They company seems to know it, however, and they appear to be addressing it now.

All that said, here's the one no-brainer use case for reddit ads: If you're targeting an extremely niche audience, you can probably find a subreddit or two for it. Even for topics so niche you couldn't find a way to target them on every other ad channel.

But don't get to excited because reddit disallows targeting ads to subreddits that are below a certain subscriber count. 

So you'll be looking for a middle ground. Good luck!

If you get reddit ads to work spectacularly, please let me know on Twitter!

Channel: Pinterest

A Pinterest ad for a client

On Pinterest, you intercept people's content searching experience with your ads. So, like Adwords, it's behavior-based targeting. Yay! We growth marketers need more ad channels with behavioral targeting to help keep saturation at bay.

Unfortunately, Pinterest Ads perform extremely poorly. Haha.

While their cost-per-clicks are low (if your ads look enough like organic content), Pinterest users generally don't convert once they arrive at your site. They bounce. 

And I don't have a good theory as to why. Do you? Ping me on Twitter.

Ultimately, I've found Pinterest only worthwhile when:

  • Your product lends itself to visually enticing imagery that organically fits among native ecommerce content, such as apparel and furniture. (Like Instagram ads, there is very little room for text on Pinterest ads. It's all about the visuals.)
  • Your audience skews toward 24 to 35-year-old females. Pinterest disproportionately reps this demographic. (It's 80% female and 60% 24-35 year-olds.)

Channel: Display ads ("banner ads")

Display ads generally convert terribly. 

People don't click banners anymore. They haven't for years.

What display ads are still good for, however, is cheap brand marketing. As in, ads that aren't first and foremost designed to get clicked.

Consider how Verizon and T-Mobile spam every possible site on the web with banner ads all the time. They aren't getting a lot of clicks relative to their ad spend. But they do it because they want to stay top of mind when your phone contract next expires and you're considering a different mobile provider.

Or consider how movie studios aggressively spam sites with banner ads before the release of a film. The same idea is at work here: They're not expecting you to click their ads then convert on the spot. They're hoping you keep the film in mind when you're bored next weekend and stumble into the theater.

If you're a Fortune 1000 brand marketer, give display ads a shot. If not, there's typically only one reason to consider them: retargeting.


Retargeting is when you follow a visitor around the web with your ads after they've visited your site. 

Retargeted impressions tend to convert significantly better than prospecting (first) impressions. So, using display ads for this purpose may be positive ROI. 

I cover retargeting when I dive into how to run a Facebook Ads campaign (in two pages from now). So I'll end this discussion here.

All other ad channels

Curious about the remaining channels I haven't touched upon? Go experiment!

Remember, it’s the job of a growth hacker to rigorously test new channels routinely. 

Put yourself in a position to discover pleasant surprises.

Running ad campaigns

In two pages from now, I walk through how to run a professional ad campaign on Facebook. In learning Facebook, you'll learn how to run ads on every other channel too. It's our proxy for everything.

On the next page, I'll finally talk about how to write and design great ads.

Next page: Copywriting

The next page shows you how to write ads that compel people to click.

Next →
Julian Shapiro headshot