HTML resume parsing - says 100% accuracy but some Company fields empty?

HTML resume parsing - says 100% accuracy but some Company fields empty?

Does anyone have any tips on how to structure an HTML resume so that "import from Document" will *reliably* fill in the Company fields in a candidate record?

With my HTML format, the parser does a pretty good job overall. Out of 10 employments, it gets all of the dates, all of the job titles, all of the summaries, and 7 company names. But for some reason it ignores 3, despite the structure of the repeating employment blocks being the same in each case.

Is this just because it doesn't recognise the company name? Is there anything I can put in the HTML source as a hint to the parser? 

Structure is per below

  1. ### name, summary etc precedes this

    <h2>Work History</h2>

    <p>startDate1 - endDate1</p>

    <h3>JobTitle1</h3>

    <h4>Company1</h4>

    <p>Description1</p>

    <hr>

    <p>startDate2 - endDate2</p>

    <h3>JobTitle2</h3>

    <h4>Company2</h4>

    <p>Description2</p>

    ### more employments and then education, skills, etc, follow


Many thanks,

John