Coping With 8,000 Title Tag Rewrites: A Case Study

This isn’t really a single word change (more like a full swap), and I don’t know why we ended up with two different words here, but what about the actual title – which is very similar to the post title – that prompted the need for a rewrite?
One quick note – remember that Featured Captions are live effects, too, so rewriting will also affect your featured captions. Here’s a similar post/rewrite of another question, appearing as a Featured Snippet:

Also, there is nothing really wrong or wrong about rewriting, except for the lack of clarity as to why it happened. In the context of Featured Snippets, however, rewriting has a greater chance of influencing the intent of the original author(s).
Did Google do you wrong?
It’s the moment you’ve been waiting for – examples of where Google has done bad things. I want to be clear that these, at least in our data set, are few and far between. It’s easy to pick the worst of the worst, but the three examples I’ve chosen here have a common theme, and I think they represent a broader problem.
(7) Last things first
Here’s an example of a breakup, where Google seems to have chosen a parenthesis over the main part of the article:

Many bad examples (or good examples of bad) seem to be where Google splits the topic based on the classifiers and then reconstructs what’s left in a way that doesn’t make sense. It looks especially strange in the case of a statement with parentheses, which should be at the side and less important than what precedes it.
(8) Part of conversation
In some cases, Google uses delimiters as breakpoints, indicating what comes before or after it. Here’s a situation where the “back” method doesn’t work very well:

This is user-generated content and, admittedly, it’s a long topic, but the breaking effect doesn’t make sense out of context. Standard reduction (…) would have been a better route here.
(9) And another thing…
Here is a similar example, but where the cutoff occurs at the hyphen (-). The style of the title is unusual (especially beginning with the subheading “And”), but the hyphen moves it from unusual to ridiculous:

Again, a simple reduction would have been a better bet here.
I get what Google is trying to do — it’s trying to use delimiters (including pipes, hyphens, colons, parentheses, and parentheses) to find natural language breaks, and separate articles from those breaks. Unfortunately, examples show how dangerous this approach can be. Even the classic “Subject:Subtitle” format is often reversed by writers, the (unimportant) part is sometimes used first.
Three courses (and three wins)
Eventually, some rewrites will be fine and most of these rewrites are not worth the time and effort to fix. More than half of Moz
What about negative rewriting, though? I decided to pick three case studies and see if I could get Google to take my suggestions. The process was relatively simple:
-
Review the
tag, trying to keep it under the height limit -
Submit the page to be rewritten in the Google search console
-
If the rewrite didn’t take, update this file
Here are the results of three screen studies (with before and after screenshots):
(1) A shadowy character
This was really our mistake and it was an easy decision to fix. Long story short, the data migration led to a special character attack, which resulted in:

I don’t blame Google for this, but the result was a weird kind of clipping that made “Google Won’t” look like “Google Won”, and made it seem like this was the end of the story. I have corrected and shortened the i

It is interesting that Google chose to use the


