Tables and text boxes, and what a parser does with them
The short answer
Content inside tables is an 8 point finding in our published rule set. Content inside text boxes is another 8. They are separate checks because they fail in different ways, and they sit next to each other because the underlying mistake is the same one: putting something that matters into a container the document body does not own.
Both are worth understanding properly, because both are extremely common in resume templates and neither announces itself when it goes wrong.
Tables: the reading order problem
A table is a layout device. People reach for one because it lines things up: dates on the right, employer on the left, or a skills grid three across. It is the most reliable way to get columns to behave in a word processor, which is exactly why it is everywhere.
The trouble starts when a parser flattens it. A table stores its content cell by cell, row by row. To read it, software has to decide whether the meaning runs across each row or down each column, and it cannot see the visual arrangement that makes the answer obvious to you.
Guess wrong and the results are specific and bad. A job title attaches to the wrong set of dates. An employer name lands in the middle of the previous role's bullets. A skills grid that reads across in three neat columns comes out as a single run of unrelated words. Nothing is missing. Everything is in the wrong place, which for a document whose entire job is to associate the right dates with the right employers is close to as damaging as missing.
The most damaging version is the invisible table, the one with its borders turned off. It looks like ordinary formatting, so nobody thinks to check it, and it behaves exactly like a table underneath.
Text boxes: the not being read at all problem
A text box fails differently. It does not get read in the wrong order. It may not get read.
Text boxes float above the document rather than sitting inside it. They are anchored to a position rather than to the flow of the text, which is the property that makes them useful for design and the property that makes them risky here. Some parsers walk the document body and never look at floating objects, so whatever is in the box is simply absent from the extracted text.
This is the same failure mode as a header or a footer, which is a separate 6 point check and which we cover in where your email silently vanishes. The pattern is identical: a region that is visually prominent and structurally outside the body.
What tends to end up in text boxes makes it worse. Because they are a design device, people use them for the parts of the page they want to stand out: a summary, a highlights panel, a skills sidebar, a call out with the best number on the whole resume in it. The content most likely to be in a text box is the content you most wanted read.
How to find out whether your resume uses either
Neither is obvious from looking at the page, so check the file rather than the layout.
In a word processor, click into the area you are suspicious about and look at what happens. If clicking selects a whole rectangular object with handles around its edge, that is a text box. If your cursor lands inside a cell and the ribbon offers table tools, that is a table, even if you cannot see any borders.
There is also a general test that works on the exported file and takes a minute. Select all the text in the PDF, copy it, and paste it into a blank plain text document. That paste is close to what a parser gets.
Read it through. If a section is missing entirely, look for a text box. If the words are all present but interleaved or attached to the wrong dates, look for a table. This single test catches both problems and several others, and it is worth running on any resume before you send it anywhere.
The fix, and what it costs you
For tables, the fix is usually to replace the table with ordinary paragraphs and tab stops, or with a real column layout where you genuinely need columns.
For a common case, the employer on the left and the dates on the right, a right aligned tab stop does the same visual job with none of the structural risk. The line stays one line in the stored order, so it stays one line when it is read.
For text boxes, move the content into the body and style it there. A summary does not need to be in a box to be a summary. If you want visual separation, a rule above and below, or a shaded paragraph, achieves it without floating the content out of the document.
In both cases the cost is usually smaller than people expect. You are not giving up the design. You are getting the same arrangement with a different structure underneath. That distinction is the whole subject, and it is also what separates a two column resume that parses cleanly from one that does not, which is covered in two column resumes.
Where these sit against everything else
Eight points each is meaningful and it is not the top of the list.
Parse is worth 35 of the 100 points, and it holds nine checks. The two heaviest, password protection and no extractable text, are worth 35 each on their own, because both mean the parser gets nothing whatsoever. If you have not ruled those out, do that before you touch a table, because there is no point improving the structure of text that will never be read. That is the two resume files that score zero.
Below tables and text boxes sit the smaller parse checks. Unreadable characters is 8, contact details in the header or footer is 6, type below 9 points is 4, and decorative bullet glyphs is 3. The last two share a cause worth knowing, and we cover it in junk characters and decorative bullets.
The rule that covers all of it
If you remember one thing from this, make it this: anything load bearing goes in the body.
Your employers, your dates, your titles, your bullets, your contact details, your summary. All of it belongs in the ordinary flow of the document, in paragraphs, in the order you want it read. Decoration can live wherever it looks best, because nothing depends on a parser finding a rule or a shaded panel.
That single rule handles tables, text boxes, headers, footers and most of the layout questions people ask about resumes, without needing to remember which of them is worth 8 points and which is worth 6.
For the whole ranked list with every weight attached, see what actually breaks a resume in an ATS. If you want the findings for your own file instead of the general case, the free ATS checker runs the published rule set in your browser, uploads nothing, and shows each finding against the line it refers to.
Both findings are reported separately by the free ATS checker that iapplyai.app publishes, because the fixes are different: a table needs restructuring and a text box needs emptying into the body. Seeing which one you actually have saves you doing both.
