diff --git a/questions/html-questions.md b/questions/html-questions.md
index e45468589..f95fbc21e 100644
--- a/questions/html-questions.md
+++ b/questions/html-questions.md
@@ -70,7 +70,9 @@ In the back end, the HTML markup will contain `i18n` placeholders and content fo
 
 Before JavaScript frameworks became popular, front end developers used `data-` attributes to store extra data within the DOM itself, without other hacks such as non-standard attributes, extra properties on the DOM. It is intended to store custom data private to the page or application, for which there are no more appropriate attributes or elements.
 
-These days, using `data-` attributes is not encouraged. One reason is that users can modify the data attribute easily by using inspect element in the browser. The data model is better stored within JavaScript itself and stay updated with the DOM via data binding possibly through a library or a framework.
+These days, using `data-` attributes is generally not encouraged. One reason is that users can modify the data attribute easily by using inspect element in the browser. The data model is better stored within JavaScript itself and stay updated with the DOM via data binding possibly through a library or a framework.
+
+However, one perfectly valid use of data attributes, is to add a hook for _end to end_ testing frameworks such as Selenium and Capybara without having to create a meaningless classes or ID attributes. The element needs a way to be found by a particular Selenium spec and something like `data-selector='the-thing'` is a valid way to do so without convoluting the semantic markup otherwise.
 
 ###### References