Right first time (RFT) is the share of work completed correctly on the first attempt, with no rework, correction, or deviation from the process. RFT = work completed with no rework or deviation ÷ total work completed. It is broader than a defect count: in a batch process it judges the whole batch, parts, process, and paperwork, as one pass-or-fail.
That breadth is what makes RFT a culture signal rather than just another yield number. A defect metric asks whether the parts were good. RFT asks a bigger question: did the whole system, the operators, the procedure, the settings, the records, do the job right the first time, without anyone having to go back and fix, adjust, or explain? This guide covers what RFT is, how it differs from first pass yield in a batch, why it reads as a quality-culture signal, and how to measure and improve it.
What is right first time?
Right first time, also called first time right, is the proportion of units, orders, or batches completed correctly the first time, without rework, correction, retest, or deviation from the intended process. The formula: RFT = number completed with no rework or deviation ÷ total number completed, as a percentage. A high RFT means work flows through clean; a low RFT means a lot of it is looping back for fixes, corrections, and sign-offs that were never supposed to be needed.
The key word is deviation. RFT does not only fail on a bad part; it fails on a departure from the process even when the part is fine, a missed step, an out-of-sequence operation, a blank field on a batch record, a correction made after the fact. That is deliberate. RFT is designed to catch the times the system did not run as intended, because those are the times a defect is most likely to escape next.
This makes RFT a demanding number, and that is the point. A plant can hit its shipment, pass its final inspection, and still have an ugly RFT because the work only got there through a string of corrections and workarounds. Each of those corrections is a small tax, a retest, a re-sign, an investigation, a supervisor pulled off the floor to approve a deviation, and RFT is the one metric that adds the tax up. It is less forgiving than yield precisely because forgiveness is what hides the cost.
How is RFT different from first pass yield in a batch process?
They share the same philosophy, get it right the first time, count rework as failure, but they differ in scope, and the difference is sharpest in a batch process. First pass yield is usually a unit-level, step-level metric: it counts how many individual units passed a given step clean. RFT typically works at the level of the whole batch, order, or run, and it includes process compliance and documentation, not just the physical units.
In a batch process that distinction bites. Imagine a batch of 1,000 units where 995 units are physically good, but the batch record has a documentation deviation, a missed signature, an out-of-limit reading that was not addressed. At the unit level, first pass yield is 99.5%. At the batch level, RFT is 0% for that batch: the batch did not go right the first time, because it needs a correction, an investigation, or a deviation write-up before it can be released. One is a parts number; the other is a whole-batch verdict on parts, process, and paperwork together.
Why is RFT a quality-culture signal?
Because it measures whether the whole system did its job the first time, not just whether the parts survived inspection. RFT catches the near-misses, the workarounds, and the quiet corrections that a defect count ignores, and those are the truest signals of how a quality culture actually behaves under pressure. A plant that reworks quietly and signs off later can post fine parts numbers and a terrible RFT, and the RFT is telling the truer story.
That is why RFT is often used as a leadership scorecard rather than a station metric. It answers questions a defect rate cannot: are the procedures right, are people trained to follow them, are the records complete the first time, does the process hold when no one is watching? A rising RFT means the system is getting more capable of doing things right without a safety net of rework. A falling RFT is an early warning that the safety net is doing more and more of the work, which is expensive, and eventually fails.
There is a behavioral edge to watch here. Because RFT reflects on the whole team, there is a strong pull to keep it high by quietly fixing deviations and not recording them, the same trap that catches every honest quality metric. The discipline that protects RFT is the same one that makes it valuable: a deviation logged is a lesson available, and a deviation hidden is a lesson thrown away. Leaders who react to a failed batch by hunting the cause rather than the person who logged it are the ones whose RFT actually climbs, because their people keep telling them the truth.
The cost is real. The American Society for Quality reports that the cost of poor quality, dominated by internal failure like rework and correction, commonly runs 15–20% of sales revenue, reaching as high as 40% at poor performers. Every point of RFT you lose is a point of that hidden work you are paying for. RFT turns an abstract culture question into a number leaders can trend and act on, and it pairs naturally with the broader cost of quality.
How do you measure and improve RFT?
Measuring RFT well starts with a strict, written definition of what "right the first time" means for a batch, then counting honestly against it. A dependable sequence:
- Define pass, in writing, for the whole batch. A batch is right first time only if it needed no rework, no correction, no deviation, and no incomplete record. Write down every condition, because a soft definition turns RFT back into a simple yield number.
- Count at the batch or order level, not just the unit. RFT is a whole-run verdict. Track it per batch so a documentation or process deviation fails the batch even when the units are physically fine.
- Log every deviation with a reason code. An RFT number without causes is a scoreboard with no game plan. A shared code list feeding defect tracking turns failed batches into a ranked list of what to fix.
- Separate parts failures from process and documentation failures. They have different fixes, a parts defect points at the process, a documentation deviation points at the record and the training. Lumping them hides which one to work.
- Attack the top cause and mistake-proof it. Work the most common deviation with root-cause analysis and statistical process control where a variable drives it, and design the failure out where you can.
- Trend RFT as a leadership metric and re-measure. Review it at the batch-release or daily meeting, not in an operator's file, and watch the trend climb as causes are removed. RFT only improves when the system changes, not when the logging gets quieter.
How does RFT connect to OEE and the rest of the plant?
RFT sits alongside first pass yield and rework rate as the quality family of a plant's manufacturing KPIs: FPY at the unit and step, rework rate on the recovery effort, RFT on the whole batch and the process around it. They should be built from the same captured records, and if they disagree it usually means a rework or a deviation is being logged in one place and not another, which is worth an afternoon to reconcile. RFT is also where quality touches OEE: a batch that fails RFT and loops back for correction burns capacity and shows up as a quality loss.
The thing RFT depends on completely is honest, complete capture at the point of work, and in a batch process that means the batch record, the readings, and the deviations, not just the parts count. Paper batch records are exactly where deviations get corrected quietly and RFT gets inflated. Plants that capture batch activity digitally as it happens, the way Harmony logs production and quality records live and feeds automatic production reporting off one source of truth (see the platform), get an RFT they can trust and act on. The shift from paper to real-time capture is worked through in the CLS case study.