Efficient Marketing Research for Startups Using Facebook Advertising

Who will buy your product? Which variation do they like best? How does your cofounder’s suggested slogan compare to the one your mom thought was clever? What’s next for your mobile app company? …How do you do this all at once quickly and relatively cheaply?

The answer, of course, is Adwords. But if you want to dive deeper, increase testing efficiency, and decrease costs, I and countless other marketers have consistently found Facebook Advertising to be a marketing research gem. I’m surprised so few startups use it effectively, particularly since potential customers may not be actively googling for the new tech these companies offer. I’ve used Facebook Advertising in general and its Power Editor specifically while consulting for a variety of startups to quickly and cheaply answer core business questions, and it’s a tool I’ve come to rely on due to its flexibility, constant improvements, and wide range of creative uses.

Additionally, Facebook Ads can do things Adwords can’t:

  • Powerful audience targeting including job titles & industry
  • Advertise based on user interest and easy in-depth targeting on users’ behavior off-Facebook through Partner Categories (quick summary here)
  • Immediate social proof: Targets can see which of their friends “Likes” your brand while they view the ad
  • Consistency: Unless a target has actually visited your site and you have retargeting set up, you can’t continually advertise to a high value lead with Adwords.
  • Clear and useful customer analysis through Audience Insights

Major brands like Gilt use Facebook Ads to easily target their customer base (which includes me) and test new ideas. The three ads below were served to me not too long ago. You might notice a trend with the language used. You can bet lots of research and testing went into these campaigns.

This slideshow requires JavaScript.

You can achieve the same efficient research and testing with just a bit of practice. If you’re not using Facebook Ads already, read through this 5-Minute Marketing guide to get ideas on how you can quickly and inexpensively find or examine your target audience, increase your ads’ effectiveness, and test various campaigns and user interest in product features. This guide was specifically written for those of you who have limited resources for marketing research, but the concepts found here are also used by brands with bigger budgets.

I’m also assuming in some of these case studies that you don’t necessarily have a customer list yet or it’s still somewhat small. If you do have a sizable list already, stop reading and go analyze your customers right now using Facebook’s Audience Insights. Introduced earlier this year, Audience Insights is a “tool designed to help marketers learn more about their target audiences, including aggregate information about geography, demographics, purchase behavior and more.” This is different than Page insights “because it looks at trends about your current or potential customers across Facebook, whereas Page Insights looks at the interactions with your Page (i.e., likes, comments and shares)” (see Facebook’s post for full details). Seriously, go check it out now if you haven’t already. You’ll be surprised how much you can learn about your customer base using Facebook.

Down below, I’ve written three brief case studies that are of particular use to startups in the consumer hardware, SaaS, and mobile app space. 

I’m focusing on three of the most common marketing research queries startups have:

  • How to Find Your First Target Market
  • Perfecting Your Sales Pitch
  • Testing New Product Concepts on Your Existing User Base

Quick Disclaimer: This is a simplified guide on how to add a great tool to your marketing research efforts, not an advertising optimization guide.

And Quick Note: You can absolutely use much of the information presented below in Facebook’s regular Ads Manager tool. But I’m suggesting you get acquainted with its Power Editor tool because it makes iterating and reusing extremely easy, and some things simply can’t be done without Power Editor. If you’re going to be doing many more ads in the future, or you’re a marketer interested in improving your digital ad buying skills, I suggest getting comfortable with Power Editor sooner rather than later.

Since I’m recommending checking this out via the Power Editor, let’s start with…

What is Facebook Power Editor and How Do I Use It?

Power Editor is the bulk ad creation and management tool for Facebook Ads. You can access it by going to Facebook’s Ads Manager.

To get started, check out Social Media Examiner’s detailed guide as well as KISSmetrics, but in short…

1. Go into Ads Manager, and find the link to Power Editor on the lefthand sidebar. Make sure you’re using Chrome!

Facebook Ads Sidebar Menu

2. Use the “Download to Power Editor” button at top to, yes, download your existing ad account into Power Editor. When you finish with any changes to your ad campaigns, sets or actual ads in Power Editor, they’ll only go live after you click “Upload Changes”. You must click that button when you’re done or none of your work in Power Editor will matter.

Download to Power Editor and Upload Button

Once your account is ready to go, you can create a Campaign, Ad Sets, and individual Ads for your tests. This is the hierarchy:

Facebook Power Editor Hierarchy

I won’t go much further into the nitty-gritty from here on out. This marketing guide exists to help you get your head around why Facebook Ads and its Power Editor can be valuable for you, to inspire you with marketing research ideas, and to give you real world examples of how other brands use it. Need more in-depth instruction? Read the guides I posted above for step-by-step details.

Let’s jump into…

Case Studies & Ideas 

I’ve used Facebook Ads to perform marketing research for a variety of products – physical consumer goods, B2B software, and mobile apps. Here are three short case studies on how to do it yourself quickly and inexpensively. Remember, this doesn’t replace all the other valuable marketing research that’s part of building a solid business; it’s just a good helper to add to your toolbox.

Selling New Hardware: Finding Your Target Audience

You have a fantastic new IoT product that massages feet while updating a blog with health metrics in real time. But you have no idea who’s most likely to buy it! Should your brand first go after moms? Dads & grads? Yuppie males? Teen girls in Park Slope who like Lindsey Stirling?

Let’s focus our investigation on gender, age, and income for now. Which combination of these basic factors results in the most likely customer?

Let’s go to Power Editor to build our test audiences so we can find out!

  1. In Power Editor, go to the drop down menu on the righthand side to go to the Audiences page.
  2. Once there, look for the green button on the lefthand side to Create Audience.
  3. In the Create Audience menu, select “Saved Target Group” – this option allows you to create an audience profile you can save and reuse.

Here’s the breakdown of the variations I’m testing. These can be changed around depending on how you want to break up your test audiences, of course. For this test, I’m only going to target the U.S., and users who speak English. Change these factors according to what makes sense for your brand.

Age Variations: 18-24, 25-34, 35-45 (These are brackets I made myself based on age brackets I’ve done in the past for other non-Facebook related research projects. I’ve excluded <18 intentionally, but you could include it.)

Gender Variations: Male or Female

Income Variations: $50,000 – $75,000, $75,000 – $100,000, $100,000 – $125,000, over $125,000 (these are exact brackets in Facebook)

Additionally, because you’re selling this cool new product on your website and not in physical retail stores yet, we only want to target people who actually buy products online and have active credit cards. Thanks to Facebook’s targeting, we can do that.

Behavior Targeting Box for Finance and Spending

The behavior targeting on Facebook is fantastic. But details on that are for another post.

Using these factors, I came up with 18 variations to test. I’ve named them as follows: “Audience Group 1: Females, 18-24, $125k+,” “Audience Group 2: Males, 14-24, $125k+” etc. I could create and save each audience easily, go back and edit, and reuse for future campaigns.

Once I have the audiences created and saved, I’ll create an Ad Set (see hierarchy chart above) for each audience variation. For consistency and statistical significance, I’ll use the same exact ads in each Ad Set, and thus for each audience.

To do this…

  1. Make sure you’ve already created an Ad Campaign for all these tests. Call it something like “Consumer Market Tests for IoT Foot Massager”.
  2. Create a new Ad Set. Make sure you place it under the correct campaign name. Name the Ad Set after the first audience you’re testing. Examples: “Audience Group 1 Females, 18-24, $125k+” or “Audience Group 2 Males, 14-24, $125k+”.
  3. In the Ad Set settings, click Edit Audience and select the companion audience group you created earlier.
  4. Once the Ad Set is ready, go to the Ads tab, and create a new ad under the appropriate Campaign and Ad Set. Create a few ads to test out copy, images, etc.
  5. Now that you’ve created a few ads, you can simply duplicate the Ad Set, change the duplicate’s name to match the next audience group you have, and of course edit the audience to select the corresponding Saved Audience.
  6. Rinse and repeat until you’ve completed all 18 (or however many) variations.

You now have all your audience profile variations with identical ads, budgets, and timelines ready to go. Easy peasy. Run the ads and see which audience profile responds to your ads the most. Get your answer in just a few days time and around $1,000. (Length of time and budget depend on what you need to achieve statistical significance! See section below.)

Perfecting Your SaaS Pitch: Testing Marketing Messages on Target Audiences

A month ago I helped a fledging SaaS startup figure out their first pitch to customers. They knew they needed to target CTOs and Directors of [redacted] in [redacted] industries. But because they were doing something that doesn’t quite exist yet, they were struggling with which words to use to describe their innovative new tech. They had a few good options, but weren’t sure which one resonated the most.

We spent about $80 to figure this out. And important to note: We could not have done this on Adwords. You simply can’t target this way on Adwords.

Because they knew the titles and industries, creating our target audience was simple.

Screen Shot 2014-12-11 at 5.32.27 PM

We created the audience profile with all the relevant titles and industries (another demographic option). We built our Campaign, Ad Set, and then built out several ads with variations on the different phrases and pitches. We also tried out different images with each variation of text.

A couple days later, we had good data that helped shape the future of their sales operations. Was this the only research they did? Of course not. But it was a quick and easy way for them to validate some assumptions so they could move on and try new things.

Target Your Current Customers: Facebook Custom Audiences & Lookalike Audiences

Imagine you’re a mobile app developer. You’ve built one app that’s been somewhat successful, but now you’re at a crossroads. You either want to 1) change up  your app to make it more appealing or 2) begin work on a new app.

You have a user base already. Why not test some ideas on them and similar audiences?

We’re going to use Power Editor’s Create Audience feature again in order to target your actual mobile app users who can be matched to Facebook accounts. Check out Facebook’s instructions on how to create a Custom Audience of your users by uploading your mobile user list. You can also target users based on particular actions they take in your app. You’ll need at least 1,000 users to ensure meaningful reach.

With this list now ready, you can experiment by showing them ads about upcoming feature ideas or mock ups of different app ideas you may have. (The link on the latter ad could simply go to a LaunchRock page for the potential app.) Success is of course defined by which concepts get the most clicks, Likes, comments, etc.

By why not see how similar app users react to your ads? Go to “Create Audience” and build a Lookalike Audience based on your Custom Audience of users. This Lookalike Audience is exactly what you think it is – Facebook users who have similar demographics and behaviors to the Custom Audience you previously uploaded or selected.

Bonus Tip: Got an Email List? Target It for Research!

Not a mobile app developer? You can use this same exact concept as above with your own customer data such as email addresses and phone numbers. If Facebook finds a match with this user data, they can be targeted with ads and/or you can analyze your list efficiently. Again, check out this post to read all the great information you can squeeze out of Facebook’s Audience Insights.

Be Confident: Don’t Forget Statistical Significance

Imagine you’re running an ad campaign targeting your current customers using Facebook’s Custom Audience tool. You want to test which of two images gets more clicks. When you review the data, you found that an ad with puppies had a 10% CTR and the same ad with kittens only had a 4% CTR. Job done? Nope. We need to see if the data is reliable or just a fluke.

Let’s assume the puppies ad was viewed by 40 people and 4 clicked on it (10% CTR), and the kittens ad was viewed by 80 people and 3 clicked on it (~4% CTR). When we calculate for statistical significance, we find that the puppies ad was 167% more effective than kittens… but we can only be 89% sure (confidence rating) that the puppies are the reason for the higher CTR. The data is not statistically significant, and we’ll need many people to view both ads in order to reach statistical significance with a 95% confidence rating.

How do you do the math for all this? The internet is full of statistical significance calculators and Excel templates to help you out. Here are two such sites:

Significance Calculator #1

Significance Calculator #2

Confidence does requires more cash, but once you achieve confidence, you should take some time to consider what to do with the raw data you’ve received. Facebook has proven to be great for research, but depending on what you want to accomplish it may not be where you want to keep spending your limited marketing budget.

Et voilà! Marketing Research is Now Easier 

You now know enough to think up more creative ways of using this tool to research your target market, test marketing copy or images, or learn more about your existing customers. Got a better case study or idea? Comment below – I appreciate the help!

Quick closing thought: As I said above, I love Facebook Ads as a tool for basic research and validation, and spreading these concepts ultimately makes my job easier: keeping costs low where you can leaves more marketing budget for other important (and fun) endeavors. All this said, at some point in your company’s life, hiring a real researcher (most likely a contractor) to explore key questions may very well be worth the spend. I’ve personally used the services of professional researchers in the past to get quality, reliable data and in-depth analysis that I simply couldn’t do on my own.

Good luck!

Photo Credit

3 Responses to “Efficient Marketing Research for Startups Using Facebook Advertising”
  1. Mark Gavagan says:

    Thank you for a very helpful article. I’m surprised you don’t have sharing via LinkedIn enabled, since your content speaks to a business owner audience.

  2. Sandy Cormack says:

    Nice article. Your method seems to focus on gauging clickthrough rates rather than gathering survey data. Have you ever used Facebook ads to perform market surveys using a third party tool like Polldaddy or SurveyMonkey?

    • Kaitlin Pike says:

      Yup! I didn’t mention surveys here because I wanted to focus on what the target audience actually does vs. what the audience says they want. As you know, the two don’t always match. Surveys and, indeed, more in-depth user research is essential.

Leave a Reply to Mark Gavagan Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: