Color shifting between print runs? Why, and how to lock it
The pouch you reordered looks a shade off next to last quarter's on the shelf. Nine times out of ten that drift isn't bad luck — it's the absence of a number everyone is printing to.
THE SHORT ANSWER
The single most likely cause of color shifting between runs is that there was never a fixed, measured target color — so each batch gets matched by eye to a moving reference, and every uncontrolled variable underneath (substrate lot, ink batch, press condition, even the vendor) is free to push the result around. Lock color by turning "our brand green" into a number, then holding it:
- Set a measured target — a named spot color or a shared press characterization, not a printed sample.
- Agree a tolerance — a Delta E your printer must hold each run to, in writing.
- Lock the variables — same substrate spec, same process, ideally the same line, on every reorder.
- Measure, don't eyeball — verify each run with a spectrophotometer against the target, not against memory.
What color drift between runs looks like
Color drift shows up when you put two batches of the same SKU side by side and the brand color no longer matches: a coffee bag that reads warmer than the last one, a carton whose blue has gone slightly green, a label whose "signature" red has crept toward orange. On a single unit it can look fine — the problem only becomes obvious in comparison, which is why it slips through a run and gets caught by a customer or a buyer at retail.
The important distinction up front: this is a consistency problem, not necessarily a match problem. Either batch on its own might be an acceptable color; what's failing is that the two don't agree. That points you away from "the printer got it wrong once" and toward "nobody defined what right is." If instead your very first run already looked different from your proof, that's a related but distinct issue — start with print doesn't match the proof for that one. This guide is part of our packaging problems diagnostic library, and color consistency is one of its most common and most fixable entries.
Why color shifts between print runs
Drift almost always has a single dominant cause with a stack of contributing variables underneath. Work the list top to bottom — the higher causes are both more common and easier to fix at the spec stage.
| Likely cause | Why it happens | How to tell |
|---|---|---|
| No fixed numeric target | Color was approved from a printed sample or a screen, so there's no measurable standard to reprint against — every run is matched by eye to a reference that itself fades and varies. | Ask what the target is. If the answer is "the last run" or "the approved sample" rather than a spot number or a characterization dataset, this is your cause. |
| Substrate lot variation | The same ink on a whiter, bluer, or more absorbent material lot reads as a different color. Paperboard shade, film clarity, and metallized bases all move between lots. | The drift tracks with a new material delivery, or a job on a different stock. Unprinted areas of the two runs look slightly different too. |
| Ink batch & viscosity | A new ink batch, a re-mixed spot, or a change in viscosity, tack, or dryer settings changes how much colorant lands and how it looks. | Color moves after an ink change or partway through a long run as viscosity drifts. Density readings differ run to run. |
| Press condition drift | Anodes, plates, aniline/anilox wear, impression pressure, and temperature all shift ink transfer over time and between makereadies. | Same file, same material, same ink, still different — and the operator can't hit the number without a lot of tweaking. |
| Different vendor or process | A reorder placed with a new printer, or moved from flexo to digital (or offset to flexo), reproduces color through a different gamut and different limitations. | You changed suppliers or were told the job moved presses. Fine detail and color both look different, not just one hue. |
| Judging by eye | Approving and QC-ing color visually under inconsistent lighting means "close enough" moves with the viewer, the light, and the day. | There's no instrument in the loop. Sign-off happens at a desk or under whatever light is handy rather than in a controlled booth to a number. |
Notice the pattern: the top cause is a process gap (no target), the middle causes are materials and machine, and the bottom two are who and how you judge. You fix drift by closing all three — but a target has to exist before any of the rest can be held to it.
How to hold color across print runs
Here's the procedure, in the order you should actually do it. Each step removes one of the variables above.
- Set a measured target for every critical color. Turn each brand color into a real specification — a named spot color (a Pantone or a custom mix with a measured Lab value) or, for images and process builds, a shared press characterization such as a GRACoL or G7 dataset. A printed sample is a reference, not a target; make the target a number. Our packaging color management guide walks through building that standard.
- Agree a Delta E tolerance in writing. Decide how close each run has to be and put it in the spec and the PO — for example, a stated Delta E ceiling for each critical color, measured a defined way. Without a written tolerance, "match" is an opinion; with one, it's a pass/fail.
- Lock the substrate. Specify the same material — the same board grade or film structure, ideally the same supplier and, for tight programs, a consistent lot family — because the same ink on a different white is a different color. If the substrate has to change, expect to re-approve the target on the new stock.
- Use spot colors for the colors you can't afford to lose. A critical brand color reproduced as a spot ink is far more stable run to run than the same color built from process (CMYK), which depends on registering and balancing several inks at once. See spot color vs CMYK for when to spend a spot on it.
- Measure every run against the target, not against memory. Verify with a spectrophotometer to the target Lab value and the agreed tolerance, in a controlled viewing condition — not by holding the new run next to a drawer copy of the old one. Instrument sign-off is what makes "consistent" repeatable.
- Keep the same process and vendor on reorders. Re-running the same file, on the same process, at the same printer, to the same target is the single biggest lever for consistency. When you must switch, treat it like a new qualification: send the target and tolerance, get a measured draw-down or proof, and approve to the number before the full run.
What a Delta E tolerance actually buys you
A tolerance is what converts color from a debate into a measurement. Delta E collapses the difference between your target color and what actually printed into one number, so a run either passes or it doesn't — no arguing about whether the green "looks off." As a rough, widely-cited rule of thumb, a Delta E of around 2 sits near the point where an average observer starts to notice a difference at all, so brand-critical colors are often held to roughly that, while less sensitive work may accept something looser like 4 to 5. Treat those as typical, approximate figures — the exact number, the Delta E formula, and how it's measured should be agreed with your printer and written into the spec, because a tolerance nobody committed to isn't a tolerance.
Two practical notes. Tighter is not automatically better: an unrealistically tight tolerance on a process that can't hold it just guarantees arguments and rejects, so set it to what the process can actually repeat. And a tolerance only means something paired with a fixed target and a defined measurement condition — the same instrument geometry, backing, and light — which is why steps 1 and 2 come as a pair.
Is it your fault or your supplier's?
Color consistency is a shared responsibility, and honestly apportioning it saves a lot of friction. Roughly:
- It's on you (the buyer/brand side) when there was no measured target, no written tolerance, the color was approved from a screen or a one-off printed sample, the substrate was left open, or a reorder was moved to a new printer or process without re-qualifying. A printer can't hold a number you never gave them.
- It's on the supplier when a clear target and tolerance were provided and a run still lands outside it — that's press control, ink management, or QC not doing its job. Delivering an out-of-tolerance run without flagging it is a supplier miss, full stop.
- It's genuinely shared when the substrate lot moved, a spot was rebuilt as process to save a station, or the tolerance was set tighter than the chosen process can repeat. These trace back to spec decisions made together, and they're fixed together.
The fastest way to find which bucket you're in is to ask two questions: what was the target? and what did this run measure? If either answer doesn't exist, the gap is upstream of the press. Setting the target and tolerance up front — ideally at proof approval — is also the cheapest insurance; a shared proof-and-target agreement turns "it looks different" into a measurable conversation.
When to escalate
Most drift is solved by adding a target and a tolerance and moving on. Escalate — to a formal color program, a press qualification, or a supplier corrective-action request — when:
- The same printer misses an agreed tolerance across multiple runs even after re-mixing the spot and re-profiling — that's a process-capability problem, not a one-off.
- A brand-critical color appears across many SKUs, materials, or vendors and has to match as a family — that warrants a documented standard and a shared characterization everyone qualifies to.
- Retail or a licensor holds you to a color spec you can't currently prove you meet — you need instrument records, not assurances.
- You're moving a line to a new converter or process and can't afford a visible shift — qualify to the number before you commit the volume.
How PackOS keeps color consistent
Consistency is a spec-time decision far more than a press-time rescue, and that's where PackOS works. When it reads your artwork it captures the color intent — which colors are spots and which are process, and what the critical brand colors are — so the standard travels with the job instead of living in someone's memory. That spec is what a reorder reproduces against, letting a target and tolerance be attached before the first run rather than reconstructed after a complaint. See how the platform handles color and print-readiness on the technology page, dig into the discipline in our color management guide, or lock the spec for your next run with Quick Quote.
Frequently asked questions
Why does my packaging color look different on every reorder?
Usually because there was never a fixed, measured target color to reprint against. When color is approved from a printed sample or a screen instead of a number, each run gets matched by eye to a reference that itself drifts, so substrate lot, ink batch, and press condition are free to push the result around. Set a measured target and a written tolerance and the reorders stop wandering.
What is an acceptable Delta E for packaging color?
It depends on the process and how critical the color is, so agree the number with your printer rather than assuming one. As a rough rule of thumb, a Delta E of around 2 is near the point an average observer starts to notice a difference, so brand-critical colors are often held roughly there, while less sensitive work may accept something looser like 4 to 5. Treat these as typical, approximate figures and write the exact tolerance and measurement method into the spec.
Do spot colors really stay more consistent than CMYK?
Yes, for a critical brand color a spot ink is generally more stable run to run than the same color built from process. A spot is one pre-mixed ink measured to a target, while a CMYK build depends on registering and balancing several inks at once, so it has more ways to drift. For the colors you cannot afford to lose, spending a spot station on them is usually worth it.
Can I get identical color from two different printers?
You can get close, but only if both print to the same measured target and the same tolerance rather than to each other's samples. Two printers reproduce color through different presses, inks, and sometimes different processes, so without a shared numeric standard they will drift apart. Give each the target and tolerance, qualify a measured draw-down before the full run, and expect the match to be to the number, not perfect.
Whose fault is color drift, mine or the printer's?
It is the buyer's side when there was no measured target, no written tolerance, or the substrate and process were left open on a reorder, because a printer cannot hold a number it was never given. It is the supplier's when a clear target and tolerance were provided and a run still lands outside it without being flagged. It is shared when a substrate lot changed or the tolerance was set tighter than the process can repeat.