Getting Started with Petition Checker

What is Petition Checker?

Petition Checker helps campaigns verify that petition signatures are valid registered voters. When a petition is collected, each signer's name and address need to be checked against the official voter registration database. Doing this by hand is slow and error-prone — Petition Checker automates most of it.

Here's the basic flow of using this tool

1. 1 You upload a scanned PDF of a petition.
2. 2 The system reads the handwriting and extracts each row of signature data automatically.
3. 3 A user confirms that the system has correctly read the pdf and makes manual correction as needed.
4. 4 The tool searches the voter file to find the best matching registered voters for each signature.
5. 5 You review the suggested matches and confirm, reject, or flag each one.
6. 5 The user handles duplicates and keeps track of everything that was struck such as no matches or out of district voters.

User roles & what you can do

Every account has a role for each campaign. Your role determines which actions you can take. You can see your role on the campaign card on the dashboard.

After signing in, you land on the dashboard. You'll see a card for each campaign you have access to. Each card shows:

• — The campaign name, office, and election date
• — The status of recent uploads (e.g., "OCR Processing", "OCR Complete")
• — Quick links: Upload, OCR Review, Match Review, and Cleanup

Click Upload on a campaign card to submit a new batch of petition pages.

Before you upload

Make sure your petition PDF is a clean scan — the clearer the scan, the more accurately the system can read the handwriting. Multi-page PDFs are fine; each page will be processed separately.

Defining columns

Before uploading, you tell the system what columns are on your petition form (e.g., "Printed Name", "Street Address", "City", "Zip Code"). For each column, you specify:

• — Column name — what's printed on the petition header
• — Field type — what kind of data it is (name, address, city, zip, etc.)
• — Required — whether this field must be filled in to be counted

If your petition already has a value printed in a column (for example, the county is pre-printed as "Cook"), you can enter it as a hardcoded value and the system will fill it in automatically for every row.

After you upload

The system works in the background to split your PDF and read the handwriting. This usually takes a few minutes depending on the number of pages. You can watch the status update on the dashboard — when it says OCR Complete, the pages are ready to review.

"OCR" stands for Optical Character Recognition — it's the technology that reads handwriting and turns it into text. It's very good, but not perfect. This step lets you catch and fix any mistakes before the system tries to find voter matches.

How it works

Click OCR Review from the dashboard. You'll see a queue of pages waiting for review. Click into a page and you'll get a split-screen view:

• Left — The scanned image of the petition page. You can zoom in and pan around.
• Right — A table of extracted rows, one row per signature line.

Compare the extracted text on the right with the handwriting on the left. If something was misread, click the cell and type the correction.

Confirming a page

When you're satisfied the extracted data looks correct, click Confirm Page. This locks in the corrections and kicks off voter matching for that page. You can then move on to the next page in the queue.

This is the main working step. The system has already tried to find a matching registered voter for each signature. Your job is to confirm whether the suggestion is correct.

Click Match Review from the dashboard. You'll see a queue of pages with unreviewed rows. Click into a page to start reviewing.

The review screen

You'll see three panels:

• Left The scanned page image so you can refer to the original.
• Center The list of signatures on this page, with the extracted name and address. The currently selected row is highlighted.
• Right Candidate voter matches for the selected row, ranked from best to worst. Each candidate shows the voter's registered name, address, and a match score.

Match scores

Each candidate has a score that reflects how closely they match the extracted name and address. The score goes from 0 to 1:

Your three choices for each row

Completing a page

Once every row on a page has been reviewed, click Complete. The system will automatically take you to the next page in the queue. Keep going until the queue is empty.

After Match Review is done, the Cleanup page helps you tidy up two remaining categories: struck rows and duplicate signatures.

Struck Lines tab

This tab lists every row that was marked as No Match, Illegible, Out of District, or identified as a duplicate. For each row you can:

• — Toggle Struck on Paper — mark that the signer physically crossed out their own line on the petition. This is useful for record-keeping. Until this box is checked the row will be highlighted.
• — Click Restore — if you think a row was incorrectly struck, restore it to "pending" so you can re-review it in Match Review.

Duplicates tab

Sometimes the same registered voter appears on multiple petition sheets. This tab groups those rows together so you can decide what to do:

• — Strike Duplicate — Remove extra occurrences and keep only the first valid signature.
• — Keep Duplicate — Mark the duplicate you intend to keep. This does not prevent the potential dupe from appearing again if a later upload also contains the same voter.

Tips & common questions