Growth

Nov 30

How to test paid media channels for B2C

Trevor Sookraj

Of the numerous channels B2C startup founders can run experiments on, paid advertising is one of the most popular—and perhaps, also the most intriguing one. Founders can see instant results on paid advertising, which is a stark contrast to the timeline needed to test other channels like organic social, SEO, or email marketing. Like every channel, paid ads come with their own set of challenges… And if a paid ad test is not structured properly, there’s a risk that the founder will end up burning cash too quickly, without getting any significant results in return.

So how can you start testing on paid ad channels effectively without compromising on your budget or the quality of your results? We combined insights from one of our Heads of Growth who specializes in this area to help you out. Read on to learn more.

What is a paid ads test?

A paid ads test is running a paid advertising campaign on a selected channel with clear expectations of results. Based on the results, the operator can make a data-driven decision on:

  1. If the selected channel is a feasible channel for marketing the business
  2. Where to double-down on testing if it is identified as a feasible channel

You can’t make an informed decision on a test that isn’t structure properly, so before starting, make sure you have clearly identified what you’re testing, what you’re expecting to see, and what success or failure looks like. For more on this, read our article How to Run a Growth Experiment.

How do you prioritize paid channels?

The goal of running a paid test is to reach your target audience, but how do you know which channel to prioritize if you haven’t talked to them? Start by asking your customers questions that will help you understand their content consumption patterns: how they discover information, what channels they spend time on, and what format they like to consume content in.

This will help inform your assumptions about what channels would be effective, while also minimizing the risk of running experiments on channels where your target audience is not engaged. For example, targeting 65+ people on Snapchat Ads likely wouldn’t make sense. But this is just your assumption—let your customers confirm that for you.

Next, consider the targeting capabilities of the channel (i.e. Facebook Ads). Targeting by interest varies by platform - Facebook may look at pages they like, Google Display at their browsing history, and Twitter by the accounts they follow and interact with. Depending on how your customers view content, certain platforms may be easier for targeting than others.

Third, review how people prefer to engage with the channel. Some platforms are less ad-friendly than others, i.e. Reddit Ads, whereas others look for user-generated content, i.e. TikTok Ads. Depending on the goal of your test, you may deprioritize certain channels or shift the content that you’ll be promoting on them.

Lastly, remember the intent on the platform—programmatic ads (Google Display Ads, StackAdapt) are typically placed on 3rd party sites where users don’t click on ads or have lower intent, so expectations when running tests on those platforms should be lower. You may have a much more affordable Cost Per Click (CPC) but a lower Conversion Rate (CR).

How do you start testing paid media channels?

Once you have prioritized your channels, use the following steps to figure out how your test will play out:

1. Build your tech stack and check your tracking

Improper tracking is the bane of every test; people often skip this step and end up with results that can’t be relied upon. Google Tag Manager is a must-have in your tech stack to easily deploy and edit pixels, ensure they are firing correctly, and that conversions are being tracked. This is especially important for consumer products where the conversion occurs off-site (i.e. web or mobile apps), and bottom-of-funnel performance (activation, converting to a paid user, etc.) is important. Some ad platforms can be configured to show you this information, but you will often need a 3rd party tool to analyze your results. More on that below!

If you are looking to test quickly, look to landing page builders like Unbounce to leverage their templates. You can build on other stacks (i.e. Webflow) for better reliability but they often require a web designer and take longer to deploy.

2. Build and test audiences

After you have finished curating your tech stack and setting up your tracking, it’s time to focus on identifying the audiences that you want to test. Broad targeting is typically the preference for B2C marketers, which gives the ad algorithm more room to learn. For example, for a consumer health product, you might create audiences of people who are interested in:

  • Health Wearables — Likes WHOOP, Fitbit, Levels Health, etc.
  • Supplements — Likes Thorne, Ritual, Care/of, BioSteel, Athletic Greens
  • Fitness Tracking — Likes MyFitnessPal, Strava, Noom, etc.

Remember that smaller audiences are harder to target, and therefore more expensive, so initial tests should aim to target audiences of 10k+ people.

Your goal is to learn as much about your audience as possible — whether certain demographics (men, specific ages, etc.) convert better, whether device traction (desktop VS mobile) has any impact purchase, and whether logistics like times of day and day of the week have any implications on the results. Depending how mature your company is, your goal might be to understand the engagement / interest of an audience (i.e. click-through rate) opposed to immediate conversions. This information will help you get better with your testing and reach more favourable outcomes.

Audiences have the greatest impact on performance metrics like Cost Per Acquisition (CPA). Focus on testing these audiences with the same landing pages, creative, and copy before moving to other areas.

3) Test landing page Call to Actions (CTAs),  messaging, and creatives

When you have established a good audience to target, you can move to testing landing pages, Call-to-Actions (CTA), and creatives. With this stage, you want to start personalizing your ads to a specific audience, to see which pain points and value props resonate with which segments.  You can coordinate your creative tests with landing pages; i.e. for a telehealth company targeting seniors, your test groups may be:

  • Misalignment between doctors and your medical condition(s)
  • Frustration with pre-defined plans that don’t consider your lifestyle
  • Lack of availability from the specialists you need

This will connect to the CTAs you test as well. You might keep “book a consultation” consistent across all creative / landing page groups, but test “talk to an expert” as another option. Words with essentially the same meaning perform can differently, and it’s important to focus on messaging that resonates with your audience, to get more favorable results.

4) Test taglines and ad copy

Once you have determined the right pain points / messaging for a specific audience, using different creative and landing pages, you are 80% of the way there. The last layer of testing is around optimization — how can you get better CTRs on the ads that you’re showing and convert them at a higher rate? This layer involves testing copy and taglines on ads, including your primary text, image copy, and CTA on the actual ad (i.e. “book now”).

You can use tools like Jasper AI to help with copy ideas and Hotjar to see what areas of your landing pages the potential user/customer is most interested in. You can also A/B test different ad copy variations to see which one results in more engagement, clicks, and a lower cost-per-lead (CPL) to identify the most effective ad copy version.

What budget should you spend on paid ads tests?

The #1 question that a marketer gets on paid media is “how much should I spend on Facebook Ads to see results?”. The short, and the realistic answer is, it depends on the channel you’re testing and your product.

Certain channels are more expensive than others (i.e. Facebook vs Google Display) but have different intent levels. You can’t expect to have the same Cost-Per-Click (CPC) on a programmatic platform like Google Display that you have on a higher intent like Facebook Ads, because your conversion rate (once they hit the landing page) will likely be higher on Facebook Ads. Likewise, smaller audiences will also have higher CPCs than larger audiences, but may have more intent towards your product and convert at a higher rate.

It’s helpful to start with a realistic CPC and CPA number that reflects your budget—any experiment that goes over this number should be paused, re-evaluated, and optimized to make sure you’re not burning cash.

To get adequate data, aim for $3k to $5k per month on a given channel to test. You can break this down as:

  1. Medium: Facebook Ads
  2. CPC: $5 per click
  3. Goal: 100 clicks per audience
  4. Audience number: 3 audiences to test
  5. Total cost: $5 x 100 clicks x 3 audiences = $1,500

The conversion / performance metrics you set will vary based on your product and the platform. WordStream estimates an average CTR of 0.90% and an average conversion rate of 9.21% on Facebook Ads, broken down by industry. However, you should set expectations yourself by working backwards from the value of a conversion. I.e.:

  • Netflix basic is $9.99 per month
  • Lifetime Value (LTV) might be $59.94, assuming a user retains for 6 months
  • Target Cost Per Acquisition (CPA) might be $30, to get paid back in 3 months

Using WordStream’s metrics above, Netflix might optimize for:

  • $30 CPA
  • $2.76 per click (9.21% conversion rate)
  • $9 cost per 1,000 impressions (CPM) (0.90% CTR)

For more information on CAC and payback ratios, check out this article by OpenView Partners. Initial experiments typically have limited data to rely on, so it’s a good idea to ear-mark a higher budget to make it easier to test and hit your goals.

What tools should you use on paid ads tests?

There are LOTS of testing tools available for paid media, so it’s easy to over-invest in your tech stack. To avoid falling into that trap, we recommend breaking down tools into categories of ad deployment and ad analytics. These tools sit on top of platforms like Google Ads and Facebook Ads to give you better visibility. Ultimately, which tools you need will depend on the level of sophistication and details you require for your testing.

For example, a tool like Omneky can simplify the way you view your analytics and deploy your ads, including identifying which creative performs the best and generating new ones for you. They even manage your ad deployment for a small fee, so you can focus on the testing strategy and not the execution.

If you prefer doing everything yourself, there’s a variety of analytics tools you can look at: some for better dashboards (e.g. Swydo), some for optimizing budget at scale (e.g. OutPoint), and some built for specific products like e-commerce (e.g. Triple Whale).

You may also need to consider tools based on your vertical—consumer apps may be more interested in cohort analysis by channel, so AppsFlyer can be helpful, whereas testing your app experience based on audience or channel can be made easier with Apptimize.

Rounding it up

Most tests will not succeed, but it’s important that you learn from them and where to double-down. Create a strong hypothesis for channel tests that is rooted in customer research, then ensure you structure it properly and have good tracking, so that you can rely on the results. Like testing on all other channels, paid ad tests should be run with the intention of getting to know your audience and their purchase habits better, so you can improve your performance metrics and grow more efficiently.

At Divisional, we work with B2C startup founders to understand the goals of their business and run growth experiments that align with those goals. Get in touch to work with a B2C Marketer at Divisional.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

  • List item 1
  • List item 2
  • List item 3
  • List Item 4

  1. List item 1
  2. List item 2
  3. List item 3
  4. List Item 4

Heading 2 - 56/40

Heading 3 - 32/24

Heading 4 - 24/18

Heading 16/14
HEADING 14/14