Use pixel

For more experienced marketers it is obvious, but still entering the pages of various businesses is short of this short script that allows remarketing. Of course, not every website on the internet needs a pixel and Facebook advertising, but if we are thinking about this form of advertising or are just starting to operate on this platform, pixel is one of the first things we should be interested in.

Pixel is a tracking code that allows you to create remarketing groups. The script connects the person who enters the page with the Facebook account. Thanks to this, by creating an ad, we can reach people who, e.g. have visited your website within 45 days. These groups include people who know your brand through the website. It is therefore the hottest type of recipients, and not using them and letting them go into oblivion is a flaw for the marketer.

Delving into the technicalities and specifics of this script, you will soon discover that you can add conversion events and parameters to it. What does it mean? This means that you can optimize your campaign for a specific conversion, e.g. purchase or filling out a form. Thanks to event installation, you will also learn how many conversions and values were made after clicking on the advertisement. By combining pixels with a catalog (product feed), you can start displaying dynamic remarketing, in which you will show a personalized advertisement with a specific product to the person who has previously viewed the site.

Use Lookalike groups

The previous point concerned the remarketing group from the site. These are extremely valuable recipients. What about the moment when we want to reach new potential customers with the advertisement?

Of course, we can use the classic method, i.e. directing advertising to groups built on the basis of demographic data, i.e. locations, age, gender, interests or behavior. However, we can let Facebook do all the magic for us. Then he will choose those who will most often use our services. Sounds beautiful? Indeed it is. The only thing we have to sacrifice to the platform are our remarketing lists. Of these groups, e.g. 5000 people who visited the site (data from the pixel) within 30 days, Facebook will prepare a list of 200,000 people in the country similar to these people. In the source, each of the 5000 people is hundreds of data that the platform analyzes and searches for profiles with similar characteristics within a given country. You can also create Lookalike based on the list of people who interacted with your Facebook page. Then even the remarketing list obtained from the website will not be needed.

As you can see, at first Facebook Ads are mainly groups of so-called standard, i.e. demographic. Over time, however, as you gain experience, you’ll start to use a combination of remarketing lists and lookalike groups.

Advertising creation

This is the only thing the recipient sees. Not your settings, not the purpose of the campaign, but graphics or video and text are the only point of real contact with the recipient – if this point fails, your campaign will not have a chance of success.

Settle in as the recipient. You are viewing the feed on Instagram or Stories on Facebook. In what form would you like to see your ad? Stock image without an ad slogan and unclear text, unmatched by mobiles? Or maybe eye-catching graphics with a slogan in the form of Call To Action, the right button and a short, understandable advertising text? It is obvious. Unfortunately, advertisers often focus on complicated audience settings and bids, and neglect the only thing that the viewer really sees.

To create the perfect creation, I suggest you follow these guidelines:

Ad text – Probably your ad will be displayed on mobiles, so try to include a maximally simple message in the first three lines that you will see in the preview of the ad in the mobiles feed version. The text can be long and the recipient willingly clicks “view more” to read it, but you need to catch attention through the first, one or two sentences, e.g. to solve the problem of the recipient with the first few words. In the following you can write what you want.

Graphics / Video – does it also stand out in feed / stories? Gets attention? Is it clear and you know what the advertiser meant? Do you know what you, as the recipient, should do after the ad? Is the format adapted to the locations? Is the graphics on Stories vertical (sample format 1080 × 1920) and in the feed square (1080 × 1080) which significantly improves the reception? Is the advertisement based on stocks that everyone already knows or maybe on their own interesting graphics / photos?

Ad headline – Short and succinctly, this should be the headline. It can take the form of a call to action.

Button – If you offer the sale of t-shirts, you will not use the “fill in application” button, eg “buy now”. The “more information” button has already taken care of everything, change it!

Test different solutions – audiences and ads

When launching a Facebook Ads advertising campaign, we set the target, audience in the ad set, and the ad itself. Let’s stop for a moment while developing the middle part, i.e. the group of recipients. We usually set up one and publish the campaign. But why not use two or more and give the algorithm some room to work? Almost every group can be divided into two, at least according to age, locations or gender. Thanks to this breakdown and the simultaneous use of CBO, i.e. optimization of the budget campaign at the level of setting the advertising goal, we will allow the algorithm to experiment and select a group that converts better, i.e. results, e.g., at a lower cost.

We wouldn’t know about it by broadcasting a campaign with one audience.

The same applies to advertising in a set of ads, i.e. graphic or video creation. It’s worth testing different headlines, other texts or even entire graphics. After a while we will notice which performs better. Then we can stop the emission of the weaker creation and implement another one for testing, which may prove to be better than the one that won in the previous test.

Analyze on a representative sample of data

It happens that after one day we receive information from the customer that, e.g. one campaign or advertisement works better than the other, and that the worse one should be turned off. With a relatively low budget, absolutely one day of issue is not a representative sample of data on the basis of which you can make radical decisions such as suspending an element of the campaign (e.g. audience). Let the ads take a few days, 3-7 days is the right range to evalsuate the effect of our actions.