Advertising on Twitter can be a great opportunity to drive online sales. But to reap the benefits you must understand how to test and optimize the channel.

One of the targeting options for Twitter advertising campaigns is device targeting. Most direct marketing campaigns include a call to action with the goal of a form completion on the landing page.

A recent desktop vs. mobile device experiment revealed that targeting mobile devices results in a 60 percent lower conversion rate and a 160 percent increase in cost per acquisition.

Additional key findings from the experiment include:

  • Targeting desktop devices doubles the cost per engagement but only results in half the engagement rate.
  • Users on desktop devices have a higher click-to-visit ratio than mobile device users. If a user on a desktop device clicks on the tweet, they are more likely to visit the landing page than a user on a mobile device.
  • Users on a desktop device has more than double the conversion rate of a visit to a lead than mobile device users.

Targeting desktop devices is more efficient than targeting mobile devices when striving for the lowest cost per acquisition. Users on a mobile device are less likely to fill out a form because it’s more of a hassle using the small keyboard and a bigger disruption in their browsing.

The process to conduct this experiment was to set up two promoted tweet campaigns with identical set-up options except for the device targeting.

When maximizing the performance of your Twitter advertising, be sure to take a scientific approach for testing your campaigns to determine what works and what doesn’t for driving down spend.

Unlike most paid channels that have long planning, set-up and implementation cycles, Twitter advertising offers more flexibility to test and refine new campaigns almost instantaneously. This gives marketers a unique opportunity to be responsive and actionable as the campaign are launched and generating results.

Let’s examine the layout, data, and results from the device targeting experiment for you to replicate or modify.

Set-up & Process

For this experiment I created two Promoted Tweets in Timeline campaigns that ran for the same time period with all the same targeting options except device targeting. Each campaign had the same maximum bid, daily campaign budget, target locations, and target interests.

The Tweet promoted in each campaign had the same ad copy but with a different URL tracking parameters to segment users from the mobile versus desktop targeting campaign.


These are the devices targeting options for each campaign.




Notice that the mobile device targeting already has a significantly different estimate reach compared to targeting desktop devices. The mobile device targeting campaign has an estimated reach of 531,000 users while the desktop targeting campaign has an estimated reach of 856,000 users.

Data & Results


After running the campaigns for 12 hours, these metrics were collected for each campaign using Twitter’s native campaign analytics and Optify Reports.


This Traffic Report shows the visits, leads, conversion rate, and pageviews performance of each individual campaign as well as combined totals.

My initial hypothesis was validated when calculating the conversion rate and cost per lead of each campaign. Targeting mobile devices is a waste of advertising money since it drove fewer leads and at a higher cost.

Each business is unique and has different goals. I know that I won’t be targeting mobile devices for any of my Twitter advertising campaigns because I can get a better cost per lead by targeting desktop devices only. Do your own testing to validate your own assumptions and behavior of your target audience.

Additional Testing Opportunities

These are some additional assumptions and targeting opportunities to test:

  • Interests: Do you have a Twitter user persona developed? What are their interests in Twitter? Which influencers are they following?
  • Location: Where do you do business? Are the locations that you are targeting aligning with the Tweets and content you are prompting? Are you using that country’s natural language in your Twitter ad copy?
  • Gender: Is your product or service gender biased or neutral? Female and males may be more inclined to convert when using a specific type of call to action language than the other.
  • Dates: Does your campaign perform better in the mornings when people are browsing on their way to work versus in the middle of the day when they are surfing Twitter during lunch?

Twitter Advertising can be a great opportunity to drive online sales if you know how to test and optimize the channel. I’d love to hear about results you’ve had with testing different targeting elements in your Twitter Ads campaigns!