How to Make a PDF Searchable in Adobe Acrobat

To make a PDF searchable in Adobe Acrobat, open the scanned file, select All tools, choose Scan & OCR, select In this file, set the page range and language, and select Recognize Text. Acrobat adds a text layer behind the page image. Save the result under a new name so the original scan stays untouched.

Those are the steps on Adobe's own help page. The parts that cause trouble in a law office come before and after: whether your licence includes the tool, what to do with two hundred files, how to prove the text is really there, and how to keep the scan you started with.

Which Adobe plan do you need to make a PDF searchable?

You need a paid Acrobat plan for the desktop tool. Adobe's Acrobat plans page lists "Turn scanned documents into editable, searchable PDFs" among the features Acrobat Pro adds to Acrobat Standard. A free route exists for single PDFs: Adobe's online OCR tool, which runs in a browser after you sign in.

Here is what Adobe's pages say about each way in:

The plans page says a desktop subscription also includes Acrobat online tools, the Acrobat Reader mobile app and the Adobe Scan app. The post on what Adobe OCR is and where it lives explains how those pieces differ.

The quickest test of your own licence takes ten seconds. Open any scanned PDF in Acrobat, select All tools, and look for Scan & OCR. If it is there and opens without asking you to upgrade, you have what you need.

One caution on the free route. Adobe's online OCR page describes uploading the file to the tool, so a client document leaves your computer. Check your firm's policy on browser tools before you use it for anything confidential.

Making one scanned PDF searchable in Acrobat

For a single file, Adobe's procedure has five steps: open the PDF, select All tools then Scan & OCR, select In this file, choose the page range and language, and select Recognize Text. Adobe's help page says Acrobat then creates a searchable text layer in your PDF.

The steps below follow Adobe's page on recognizing text in scanned documents, with the preparation a careful office adds around them.

Before you start

1. Copy the scan. Adobe's page carries a tip to save a backup copy of the original scanned PDF before editing, so you can restore it if needed.
2. Check for restrictions. The same page says to remove any security restrictions before working on a scanned PDF. If the file came from someone else with restrictions set, ask the sender for an unrestricted copy.
3. Look at the pages. Note the language, and note whether some pages are already typed text.

The steps

1. Open the scanned PDF in Acrobat.
2. Select All tools, then Scan & OCR.
3. Select In this file.
4. In the dialog box, choose the page range and the language for text recognition.
5. Optionally, select Settings to set additional options.
6. Select Recognize Text.
7. Save the file under a new name.

Choosing the page range and language

Pick all pages unless you have a reason not to. A reason might be a 400-page file where only an appendix was scanned.

Set the language to the language on the page. A Spanish-language contract recognized as English can come out with damaged accents and odd word breaks. If one file holds documents in two languages, run the two page ranges separately, each with its own language.

The Settings button

Adobe's scanned PDF settings page describes two text recognition options: Document Language and Output, which selects the PDF type. The page does not say what each type does to the picture of the page, and for a legal document that is the question that matters. Compare one page before and after recognition on a copy before you process anything you will file or produce. The post on Adobe OCR lists the settings Adobe lets you change before recognition, including the ones that apply while scanning.

If your version of Acrobat looks different, the command may be elsewhere. Adobe's own pages do not all give the same path, and the same post sets out where the Recognize Text command sits in Acrobat.

Making a folder of PDFs searchable in one Acrobat run

For many files, use In multiple files under Scan & OCR. Adobe's steps are: add the PDFs, set a target folder and file names in the Output Options dialog, set the recognition options in the General Settings dialog, and select OK. Acrobat writes searchable versions to the output folder.

In order, from Adobe's help page:

1. Select All tools, then Scan & OCR.
2. Select In multiple files.
3. In the Recognize Text dialog box, select Add files and choose the PDFs.
4. In the Output Options dialog box, specify a target folder for the processed files and set the file name preferences. Select OK.
5. In the Recognize Text - General Settings dialog box, specify the options and select OK.

Adobe's page says Acrobat processes the selected documents and creates searchable versions in the specified output folder. That design helps you: point the output at a new, empty folder and the source files are left alone.

A worked example

Paralegal Marcus Oyelaran receives 140 scanned PDFs from the other side in Brennan Freight v. Castellan Supply. They sit in a folder named Castellan_Prod_001_as_received.

• He makes that folder read-only and creates a second folder, Castellan_Prod_001_searchable.
• He runs the multiple-file command on three files first, with the output going to the second folder, and opens each result to test it.
• Satisfied, he adds the remaining 137 files and runs the batch over lunch.
• Afterwards he compares file counts in the two folders. Both show 140.
• He spot-checks ten files, choosing the largest, the smallest and eight at random.

The three-file trial is the step people skip. It costs five minutes and catches a wrong language setting or a wrong output folder before it is repeated 140 times.

On each finished PDF, search for the exhibit number, a word from the complaint caption, and a date on the deposition cover. In a mixed discovery folder, pick one motion title you can see and search that too. Read the highlight on the page image, using the first page, a middle page, and the last page of a long set.

Settings apply to the whole batch

The General Settings dialog is filled in once for the run. Sort the files first when they differ: French-language documents in one batch, English in another.

Action Wizard

Adobe's plans page mentions a second way to automate this in Acrobat Pro: Action Wizard, which it describes as a way to create and run custom sequences of tasks, such as OCR, redaction, watermarking and file conversion, on multiple PDFs at once. It is worth a look if recognition is one step of several you repeat. For productions that run to thousands of pages, the post on batch OCR for discovery documents covers planning and quality control.

How do you tell that Acrobat added searchable text?

Search for a word you can see on the page. Press Ctrl+F, or Command+F on a Mac, type the word, and Acrobat should highlight it on the scan. Then drag across a sentence: individual words should select. If both work, the text layer is there.

Acrobat gives you two further signals.

The prompt. Adobe's page on fixing text issues in scanned PDFs notes that if OCR has not been applied and you try to select text or use Read Out Loud, Acrobat prompts you to run OCR. Seeing that prompt on a file you thought you had processed tells you it has no text.

The suspects. The same page explains that when a match is uncertain, Acrobat marks the word as a suspect, showing the original image with searchable text behind it. The Correct recognized text command under Scan & OCR steps through them. The post on Adobe OCR describes that review and what it does not catch.

A test that takes two minutes per document:

• Search for one word from the first page, one from the middle and one from the last page.
• Search for a party's surname and compare the number of hits with what you expect.
• Copy one paragraph into a blank document and read it against the scan.
• Check one date and one dollar figure by eye.

If the copy does not match the page, Adobe's instruction is to correct errors by hand or run OCR again with adjusted settings. The post on why OCR output is full of errors explains how to tell a scan problem from a software problem when the copy comes out wrong.

When Acrobat refuses to recognize text on a page

Adobe's help pages name two conditions for recognition in Acrobat: remove any security restrictions first, and supply an image of at least 72 dpi. They list no error messages and do not say how Acrobat treats a page that already holds text, so test such pages on a copy.

The file is restricted or locked

Adobe's recognize-text page tells you to remove any security restrictions before working on a scanned PDF. Its tips on editing scanned documents repeat the point: if you cannot edit the PDF, check for security restrictions. If you did not set the restriction, do not work around it. Ask for a clean copy. A restriction on a produced document may reflect a protective order.

The page already has text

A page saved from a word processor holds real characters, so it does not need recognition. Mixed files are common: a typed cover letter followed by scanned exhibits. Use the page range to run recognition on the scanned pages only. The section on what makes a scanned PDF searchable explains the difference between a page of characters and a picture of a page.

A harder case is a scan that arrived with a poor text layer from the copier that made it. You can select words, but they paste as nonsense. Whether and how Acrobat replaces an existing layer is not covered on the Adobe pages cited here, so test on a copy.

The image is too poor

Adobe's settings page says all of the output options require an input of at least 72 dpi. That is a floor, and far below what reads well. If the pages are faint, skewed or speckled, Adobe's page on enhancing scanned documents describes the Enhance scanned file command, whose settings include filters such as deskew and background removal. Rescanning the paper, when you have it, usually beats any filter. The post on OCR for a low-quality PDF covers what else you can do.

A quick diagnosis

If a file defeats Acrobat after two attempts, stop and change approach. The post on when to look for an Adobe Acrobat OCR alternative describes when a second tool is worth having. A browser converter such as Schema Law takes a PDF of up to 20 MB as an upload and exports a searchable PDF or a Word file; the how it works section of the homepage shows the steps.

Keeping an untouched copy of the scan before Acrobat changes it

Save the original before you run anything, and never write the result over it. Adobe's own help page gives the same instruction as a tip: keep a backup of the original scanned PDF so you can restore it. For a legal file the original is also your record of what you received.

Recognition in one file changes that file. Acrobat adds a text layer, and Adobe's settings page does not promise that every output type leaves the page image as it was. Once you save, the file on disk is no longer the file that arrived. If anyone later asks what the other side actually sent, you want to be able to hand over exactly that.

A routine that works for one file or a thousand:

1. Keep an "as received" folder. Put every incoming scan there first. Make the folder read-only if your system allows it.
2. Work on copies in a second folder. Run recognition there.
3. Use Save As for single files. Give the result a suffix, for example Lease_2019_scan.pdf and Lease_2019_scan_searchable.pdf.
4. Use a separate output folder for batches. The multiple-file command asks for a target folder. Never point it at the source folder.
5. Record what you did. A line in the matter notes is enough: the date, the tool, the language and the output setting.
6. Compare page counts. The searchable file should have exactly as many pages as the original.

Whether the searchable copy can itself be filed or produced depends on whether its page images match the original. Adobe's settings page does not say which output types keep the page image, so compare pages yourself. The section on keeping the page image identical to the original explains what to look for. Court rules on searchable filings differ, so check the rules of the court you are filing in.