Blog / OEE
OEE · Guide

How to Start an OEE Program: The First 30 Days

OL By OEE Lab Editorial |Updated 2 November 2026 |6 min read
Short answer

Spend the first 30 days on one line: week 1 to agree the rules and name an owner, week 2 to log every shift, week 3 to read the first Pareto and pick one loss, and week 4 to fix it and report. On day 30 you should have an honest baseline, one tested fix and a decision on what comes next.

What this month is for

You have been asked to "get OEE going", and Monday is close. This page is the calendar for that first month, with the outputs each week must produce.

For the full method and the long-term rollout, read how to implement OEE. If your boss has not bought in yet, start with how to explain OEE.

The goal of month one is modest: one line, a number people accept, and one loss visibly smaller. Do not aim for a plant-wide dashboard yet.

The calendar

DaysFocusWhat must exist at the end
1-7Agree the rulesA one-page rule sheet, and a named owner
8-14Log every shiftA week of logs, checked against counts
15-21Baseline and ParetoOEE split into A, P and Q, and one chosen loss
22-30Test one fix and reportBefore and after figures, and a one-page report
If a week's output is not ready, take the extra days. Moving on without it costs more later.

Days 1-7: agree the rules

Day 1: pick the line and the owner

Choose a line whose output matters, with a supervisor who wants this. The owner is one named person, with time set aside.

Day 2: tell the crews, before anyone sees a clipboard

Say what it is for and what it is not for. Operator buy-in has the words to use, and the first promise to make.

Days 3-5: write the rule sheet

It fits on one page and answers five questions:

  • What is planned production time, and do breaks count?
  • What is the ideal cycle time for each product?
  • How long must a stop last to get logged? See stop thresholds.
  • What counts as a good unit?
  • What are the reasons? Keep to about eight, using reason codes as a guide.

Days 6-7: teach it and test it

Run a short session with each shift, using OEE training for operators, and log one test shift. Expect to change the reason list.

Days 8-14: log every shift

Start the log on the downtime log template or the OEE tracking template. Paper is fine this month, since you are learning the loss model, not proving a fix.

The owner visits the line on every shift, including nights, at least once this week. Most of what is wrong with the log shows up on the first visit.

Every day, check the log against the count. Ideal time for the count plus logged stops should come near the planned time, and a big gap means stops went unlogged.

Run time = 480 - 62 logged stops = 418 min
Ideal time = 7,500 units x 3 s = 375 min
Unexplained = 418 - 375 = 43 min
Hypothetical shift. The 43 minutes are slow running or stops nobody wrote down, so go and watch the line.

Do not punish gaps. Ask what made logging hard that day and fix that, and see operator downtime logging for more.

Days 15-21: first number, first Pareto

Add up the logged week, and put it through the OEE calculator. Use the standard split from Overall Equipment Effectiveness (R. C. Hansen, 2001): availability x performance x quality.

Example: the week comes to availability 78%, performance 81% and quality 97%, so OEE = 61.3%. Do not judge that figure yet, because a first number mostly tells you how good your measuring is.

Then build a downtime Pareto and look at both minutes and the number of stops. Pick one loss with these three filters:

  • It is in the top three.
  • The crew can do something about it this month.
  • You can measure it next week with the same log.

Example: the top three are changeover overruns, a jam at the labeller and waiting for pallets. A labeller redesign will take longer than a month, so the team picks changeovers.

Book a 10-minute slot each day from now on, and use the daily production meeting format to look at yesterday's top loss.

Days 22-30: one fix, one report

Keep the fix small and cheap. Example: stage tools and the first pallet of material before each changeover, so nobody walks to fetch them.

Keep logging exactly as before, so the figures compare. Changing the rules now would make the result meaningless.

OEE, baseline week61.3%
OEE, week after the fix63.6%
Availability, baseline78%
Availability, after the fix81%
Hypothetical first month, on a shared 0-100% scale. Availability +3 points is 14.4 minutes in a 480-minute shift.

One week is not proof, because product mix alone can move OEE by that much. Say so on the report, and keep tracking the loss you fixed, not only the headline.

The one-page report for day 30

  • The line, the rules and how the data was taken, including what it misses.
  • Baseline OEE, split into availability, performance and quality.
  • The top five losses, in minutes.
  • The fix, and what moved. Use the point value calculator to put it in money.
  • The next loss, and what you need to go after it.

Who does what

  • Sponsor: a plant or production manager. Protects the owner's time and asks about the top loss each week, not about the score.
  • Owner: keeps the rule sheet, checks the log, runs the Pareto and writes the report.
  • Operators: log stops, say when a reason does not fit, and suggest fixes.
  • Technician: helps with the fix and checks that counts are right.

Traps in the first month

  • Setting a target on day 1. You do not yet know what the number means. Read how to set OEE targets after the baseline.
  • Shopping for software first. A month of logging will make your requirements list far better.
  • Comparing shifts straight away. Early gaps are mostly differences in logging, so publishing them starts a fight.
  • Skipping the fix. Without a visible result, the log feels like extra work and quality falls.

Day 31

If the crew accepts the number, keep this line going and start one more. If short stops turned out to be a big unknown, it may be time to capture stops from the machine, and OEE data collection methods compares the options.

See it live in Fabrico

A paper log is right for the first month, and later it may leave too many short stops unexplained. When you get there, Fabrico takes stops, speed and counts live from the machine, and operators tag the cause at the line, with OEE per line and shift. See how Fabrico tracks stops.

What people ask next

What is the first step in starting an OEE program?

Pick one line and name one owner. Then write a one-page rule sheet before anyone logs a stop. Skipping it leads to arguments over the data later.

Should I start an OEE program on the bottleneck?

Yes, if you can. A gained minute on a bottleneck adds output for the whole line, so the first result is easier to value. If its team is unwilling, choose a nearby line and return later.

Do I need software to start an OEE program?

No. A paper log and a spreadsheet are enough for the first 30 days, as OEE without software shows. You will write better requirements once you know your own losses.

How often should OEE be reviewed in a new program?

Daily for a few minutes at the line, and weekly with the sponsor. Monthly reviews come too late for anyone to remember what happened. Short interval control shows how to go further.

What OEE should I expect in the first month?

On a hand log, expect availability to look high and performance low, because missed short stops show up as slow running. Compare the result with your own later data, not with other plants. Is 85% OEE realistic? covers the outside figure traced to Nakajima.

See OEE per line and shift on your own machines

Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.

Book a Fabrico demo

Related guides: OEE for One Machine: A Simple Way to Start · MES Implementation Steps: The Rollout in Order

← More OEE Lab guides · How to implement OEE →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES