3 Tips for Writing Paid Search Ad Copy that Performs

SOAR-2016-web-size-blog-header
The Future of Performance Content
April 25, 2016
CMO-Briefing-2
Generating Social Content for SEO Value
April 27, 2016

3 Tips for Writing Paid Search Ad Copy that Performs

SEARCH2

Post by Heather Kollme, Senior Media Manager and Jay Schmiedeskamp, Media Director

Conveying persuasive messaging in 70 characters or less can be a difficult task for any advertiser. Keep these 3 tips in mind to write paid search ad copy that drives leads and sales.

(1) Tailor ad copy by keyword type, device and landing page content

  • Ad Copy and Search Engine: Ad copy tends to perform differently by (1) brand versus non-brand and (2) engine (Google, Bing & Yahoo). To illustrate, when testing non-brand headlines for upper-funnel generic non-brand health insurance terms, Performics saw a 23% higher click-through rate (CTR) versus branded headlines. Additionally, across three engines, we tested using the language “Official Site” with a trademark symbol in branded headlines for three different client lines of businesses; we found that “Official Site” won across all three for Bing only.
  • Display URLs: Theme-focused display URLs should be utilized whenever applicable. Performics tested a standard domain display (i.e. apple.com) versus a more focused URL including a top performing keyword by category theme (i.e. www.apple.com/iPad) and saw higher CTRs anywhere from 5-13% (performance varied by keyword theme) for the themed URL.
  • Modify Ad Copy for Mobile: Optimizing ad copy specifically for mobile devices will yield stronger results. If generating calls through mobile ad copy is an important KPI, consider including a “Call” call-to-action in the headline. When testing this feature, we found a 20% higher click-to-call rate on branded keywords versus headline copy without the call-to-action.

(2) Use clear and concise language, strong calls-to-action and influential punctuation

By using language that’s straight to the point, and including a strong call-to-action with influential punctuation, brands are likely to see higher CTRs and conversion rates. Within the ad copy, it’s important to highlight what makes your brand unique in comparison to your competitors (offers, products and/or services).

  1. Poor Example: Order Jimmy John’s Delivery Today. Order Online.
  2. Good Example: Order Jimmy John’s Delivery Today. Place Your Order Online Now!

(3) Never stop testing

To maintain high CTRs, advertisers must continually test ad messaging to ensure they’re delivering the most engaging and relevant ad copy to their consumers. There are various ways to measure a copy test, including:

  • Setting a benchmark (control ad copy), to test against to see if performance can be improved vs. the control’s benchmarked performance
  • Letting the ad copy tests run until the test has become statistically significant. What does this mean? Statistical significance is not a term just for statisticians, but also for marketers. Results are considered statistically significant if it’s unlikely that they occurred by chance. There are different levels of confidence in statistical significance: 90%, 95%, 99%—each represents the reliability of the test results. In most cases 95% is an accepted confidence level for A/B testing.
  • Ensure you have set your campaign ad rotation settings to rotate evenly or indefinitely. If you are running a test that will take more than 90 days to reach significance then set your ads to rotate indefinitely.

In summary:

  • Testing each component of ad copy (headline, description line & display URL) by keyword type, engine and device is an effective tool to increase CTRs
  • Remember to always have a control iteration of ad copy when testing to compare or benchmark performance against. This will allow identification of a clear winner whether it’s through A/B testing or through a factorial test. Optimizing and continually testing ad copy will lead to improved performance and impactful learnings for future tests.

Comments are closed.

Performics Newsletter

[raw]



[/raw]