How Google improves AdWords
Google is committed to providing you with the best version of AdWords, so we’re constantly developing new features, asking for your feedback, and researching and testing improvements.
This article will help you understand how Google uses research and experiments to improve AdWords.
How experiments work
Most experiments are aimed at improving our technical systems and infrastructure, and typically aren’t noticeable to advertisers. Other experiments may be noticeable, but typically only by a subset of advertisers for a relatively short period of time. So the chances of any particular experiment meaningfully affecting your campaigns at any given time are low.
With hundreds of millions of consumers running billions of queries on Google each day, and with highly competitive advertisers constantly updating their bids, keywords, and campaign settings to keep up, it’s much more likely that unexpected changes in performance are from changing market conditions than from a temporary AdWords experiment.
In the unlikely case that your campaigns are opted in to an experiment, you might notice some differences in ad formats, ad placements, or even in the AdWords interface. Because anyone could be opted in to an experiment, you could also see changes in competitors’ ads. Regardless of experiments, you are in control. Google always respects your ad settings, even during experiments, and you will never pay more for a click than your maximum CPC.
What to do
Use AdWords like you always do, and monitor your campaign performance to ensure your settings are appropriate for your goals. Our systems constantly incorporate your settings changes, even during experiments, so you’ll remain in control. For example, if you don’t like a particular formatting or ranking change, you can always update your bids or other campaign settings (and as always, you never pay more than your maximum CPC bid). Remember, although some experimental changes may be temporary, others may become permanent if the testing shows positive results. And often the performance changes you’re seeing aren’t related to experiments or system changes at all. So your best bet is to keep close tabs on your campaign performance and update your settings accordingly.