I’m not an emotional person at all. I stay calm even during intense moments, like a 28-3 deficit making a comeback. I can easily discard old photos and sentimental commercials don’t faze me. It seems I’m in the minority. Our analysis of 612 successful AdWords text ads revealed that 45% of top-performing ads use positive emotions.

A reason to be happy However, general trends can be deceiving. Since that study, we’ve experimented with sentiment in our ads with surprising results. Today, we’ll explore how “sentiment” works in AdWords text and display ads. We’ll analyze two examples of positive versus negative ad tests from our own company, nexus-security, and discuss the outcomes. We’ll share what worked and what didn’t in our sentiment-based A/B testing, giving you tips for creating your own emotionally resonant ads. First…

What is sentiment in AdWords?

“Sentiment” refers to using ad creative to tap into emotional triggers. Most text ads you see are “neutral.” For example, searching for Bluetooth headphones might show this ad:

Direct and informative? Absolutely. Likely to influence a purchase? Definitely. Sentimental? Not at all. A lack of emotion doesn’t make it a bad ad. Remember, half of the best-performing ads are neutral. However, some searches yield ads addressing user needs with a touch of sentiment. For instance, consider these ads for “network security software” (brand names are hidden):

The ads in positions 1, 3, and 4 use the search terms effectively and highlight their solutions’ benefits. This approach is completely valid. However, notice anything different about the ad in position 2 (besides the red boxes)? Its second headline stands out. While other ads are somewhat dull, this ad uses informal language (“this SUCKS mom, I don’t wanna clean my room!”) to create an emotional connection. It uses negative sentiment to show empathy. Sentiment is easier to convey in Display Network and Gmail ads because visuals can do the heavy lifting. For instance:

^Clearly an example of positive sentiment, right? Just kidding. This advertiser saw a 47% CTR increase using negative sentiment compared to previous positive ads. This proves that while only 2% of top-performing text ads in our study were negative, there are exceptions. In our analysis mentioned earlier, we used the Vader sentiment analysis tool with Python’s NLTK library to determine the sentiment (positive, negative, or neutral) in our dataset. However, you don’t need complex tools to implement sentiment. Intuition works too. Think about your industry, target audience, and competitors’ language. Use this information to create positive and negative ad copy, then start testing. For beginners, ensure your A/B testing is structured. “Wait, what does that mean?!” Free guide » The 120 Best Words & Phrases for Marketing with Emotion

A/B testing simplified

AdWords performance is heavily influenced by past results. Therefore, simply replacing your current ads with new positive and negative versions won’t work. Instead, test one sentiment-based variant against your existing copy for at least two weeks. By “test against,” I don’t mean keeping your existing rotation and adding new copy. You’re smart; if you’ve found successful ads and aren’t A/B testing, your ad rotation is likely set to “Optimize for conversions.” This isn’t bad, but testing a new POSITIVE ad against your top-performing, likely neutral ad, puts the new one at a disadvantage. You told Google to prioritize conversions, and historically, the existing ad converts better. Google is just following instructions! Before launching a sentiment-based test, set your ad rotation (in the campaign settings) to “rotate evenly”:

Next, label your ad variants as “Positive” and “Negative” to easily analyze data in the Dimensions tab and see the overall performance. (Advanced users can use AdWords’ Experiments feature for complex A/B tests, but for most, even rotation and labels suffice).

Now you know:

• What sentiment is
• How it can impact ad performance
• How to rotate ad copy effectively Let’s look at some examples!

Example #1: Testing Sentiment in Search Ads

At nexus-security, we have a useful tool called the AdWords Performance Grader, which is quite popular. This leads to significant search volume for “AdWords performance grader.” Why is this valuable? Besides being our tool’s name (making it a branded term), it indicates strong search intent. Think about it. People searching for an AdWords performance grader likely believe their accounts could perform better, right? This scenario calls for emotionally-driven ad copy. For a long time, we used negative ad copy like this:

While it doesn’t explicitly call users foolish, it subtly implies they’ve made mistakes by “wasting money in AdWords.” However, since our data showed that top ads leaned towards positivity, we decided to test our best-performing ad against a more encouraging version. (Since the negative ad was the incumbent, we didn’t test it against a neutral one.) From July 7 to July 31, we ran the “Stop Wasting Money in AdWords” ad against this:

This variant is entirely positive. Instead of “wasting,” users are encouraged to “get the most out of” their ad spend. Both phrases convey the same idea with different sentiments. Which one performed better during those three weeks? The negative ad’s Conversion Rate was 18.8% higher than the positive ad’s. A significant difference, but even more so for click-through rates: the negative ad’s CTR was 67.29% higher. What does this mean? People don’t respond well to positivity when it comes to their ad budgets.

Example #2: Testing Sentiment in Display Ads (Gmail)

A/B testing Display ads is challenging due to numerous variables: concept, color, copy, and sentiment can all impact performance. To understand the impact of positive and negative sentiment in image ads, we ran two comparable Gmail ad campaigns throughout June. The positive version featured a smiling pig being filled with coins; the negative one showed a broken piggy bank leaking money. Both targeted two specific audiences. Here’s the creative:

The copy is similar to our text ad test. However, the results were completely different. The positive ad generated 50% more conversions than the negative ad. What does this tell us? People dislike dead cartoon pigs! Obviously. More seriously, it highlights that visuals play a crucial role in conveying sentiment and driving results. (Note: There are many differences between the ads; even button copy can influence results. To isolate the performance driver, change one variable at a time.)

Conclusion

As these tests demonstrate, there’s no universal rule for emotional marketing. Whether positive sentiment works better in text ads or negative sentiment in display ads depends on your industry, audience, offer, and more. Testing is crucial. The key takeaway is that changing your ads’ sentiment, whether from positive to negative, vice versa, or neutral to emotional, can significantly impact CTR and conversion rates. So experiment and see what resonates with your potential customers (and share your findings with us!).