Most Excel trainers tell you to drag a PNG into Excel and call it 'converted'. That’s like calling a photocopy of a contract a legally binding signature. PNGs aren’t data — they’re pixels. And Excel doesn’t read pixels. It reads values, formulas, and structured ranges. If you’re pasting a screenshot into Sheet1 and hoping for columns, you’re building on sand.
The Setup
You get a WhatsApp message from your procurement lead: a PNG screenshot of a supplier quote table taken on a phone. No CSV. No email attachment. Just this image — and the finance team needs the numbers in Excel by noon.
Here’s what that PNG actually shows (reconstructed from real OCR output):
| Supplier | Item Code | Qty | Unit Price | Total |
|---|---|---|---|---|
| Acme Corp | A782-BX | 12 | $42.50 | $510.00 |
| BrightLine Ltd | BL-904T | 5 | $189.99 | $949.95 |
| DynaTech Solutions | DT-SM33 | 20 | $12.75 | $255.00 |
| GreenField Co | GF-22X | 8 | $64.30 | $514.40 |
| Horizon Labs | HL-77Z | 1 | $1,245.00 | $1,245.00 |
| Jade Systems | JS-88R | 15 | $27.40 | $411.00 |
| Nexus Dynamics | ND-55P | 3 | $219.95 | $659.85 |
| Stellar Group | SG-11M | 7 | $89.00 | $623.00 |
The Challenge
This isn’t about ‘importing’ a PNG — Excel has no native PNG-to-table engine. You’re really doing optical character recognition (OCR), then cleaning, then structuring. The trap? Assuming any OCR tool gives clean, aligned, typed-text output. Real-world screenshots introduce misaligned columns, decimal confusion ($12.75 → $1275), merged cells disguised as spaces, and inconsistent spacing between digits.
What makes it tricky isn’t the tech — it’s the assumptions. People assume the PNG is ‘clean enough’. It rarely is. Your job isn’t to convert — it’s to reconstruct.
Walking Through It
We’ll use Excel’s built-in Insert > Picture > Insert Screenshot, then leverage Alt+N+P to paste as image, followed by Alt+T+O (Data > From Picture) — but only if you're on Microsoft 365 (v2206 or later). That’s the first surprise: this feature *only works on Windows with an active M365 subscription*. Mac users? Skip straight to step 3.
Step 1: Paste & Trigger OCR
Paste the PNG into cell A1. Right-click → ‘Extract Data from Picture’. Excel runs OCR and drops raw text into a new sheet. It looks messy — one long column, no headers, line breaks everywhere.
| Raw OCR Output (Sheet2!A1:A24) |
|---|
| Supplier Item Code Qty Unit Price Total |
| Acme Corp A782-BX 12 $42.50 $510.00 |
| BrightLine Ltd BL-904T 5 $189.99 $949.95 |
| DynaTech Solutions DT-SM33 20 $12.75 $255.00 |
| GreenField Co GF-22X 8 $64.30 $514.40 |
Step 2: Split & Clean
Select A1:A24. Go to Data > Text to Columns > Delimited > Next. Uncheck Tab, check Space, and crucially — check ‘Treat consecutive delimiters as one’. Why? Because some rows have 2–3 spaces between columns. Without that box, Excel creates empty columns. Then click Finish.
You’ll get 5–7 columns. Delete blanks. Use =SUBSTITUTE(A2,"$","") in E2 to strip dollar signs. Drag down. Format E2:E10 as Currency.
Step 3: Fix Misreads (The Counterintuitive Tip)
Look at row 5: “Horizon Labs HL-77Z 1 $1,245.00 $1,245.00”. OCR sees the comma and treats it as a delimiter — splitting $1,245.00 into two cells: $1 and 245.00. The fix? Don’t remove commas first. Instead, use =TEXTJOIN("",TRUE,IF(ISNUMBER(--MID(A5,ROW(INDIRECT("1:"&LEN(A5))),1)),MID(A5,ROW(INDIRECT("1:"&LEN(A5))),1),"")) — but only for numeric columns. Simpler: wrap all currency fields in =VALUE(SUBSTITUTE(SUBSTITUTE(E2,",",""),"$","")). Yes — remove both comma AND dollar sign *before* converting to number. That’s the surprise: Excel’s VALUE() fails on $1,245.00, but succeeds on 1245.00.
The Result
After filtering, sorting, and applying TEXT formatting to Item Code (to preserve leading zeros if any), here’s your final, validated table — ready for pivot tables, SUMIFS, or export:
| Supplier | Item Code | Qty | Unit Price | Total |
|---|---|---|---|---|
| Acme Corp | A782-BX | 12 | $42.50 | $510.00 |
| BrightLine Ltd | BL-904T | 5 | $189.99 | $949.95 |
| DynaTech Solutions | DT-SM33 | 20 | $12.75 | $255.00 |
| GreenField Co | GF-22X | 8 | $64.30 | $514.40 |
| Horizon Labs | HL-77Z | 1 | $1,245.00 | $1,245.00 |
| Jade Systems | JS-88R | 15 | $27.40 | $411.00 |
| Nexus Dynamics | ND-55P | 3 | $219.95 | $659.85 |
| Stellar Group | SG-11M | 7 | $89.00 | $623.00 |
What Could Go Wrong
Three mistakes I see daily — each causing hours of rework:
- Mistake #1: Skipping the ‘consecutive delimiters’ checkbox. Results in 12 columns instead of 5 — with random blanks between every word. You spend 20 minutes deleting empties instead of fixing structure.
- Mistake #2: Using ‘Fixed Width’ instead of ‘Delimited’ for text-to-columns. Excel guesses split points based on spacing — and misplaces them where company names have internal spaces (e.g., “BrightLine Ltd” splits after “Bright”).
- Mistake #3: Applying
VALUE()before removing commas. Returns#VALUE!— then people manually edit each cell. The real fix takes one formula dragged across.
Here’s what to do next — right now:
| Action | Shortcut / Command | When to Use It |
|---|---|---|
| Paste PNG as image | Alt+N+P | Before triggering OCR — keeps original resolution |
| Launch Text to Columns | Alt+A+E | On raw OCR output — never skip ‘consecutive delimiters’ |
| Clean currency strings | =VALUE(SUBSTITUTE(SUBSTITUTE(A2,",",""),"$","")) | For any column with $ + comma — works on 10K rows instantly |
| Preserve leading zeros (e.g., item codes) | Format column as Text before pasting | Prevents Excel from auto-converting “00782” → “782” |
