Coffee Quality Control: A Weekly Tasting Routine for Cafés

Ask most café owners how they check their coffee and the honest answer is that they taste a shot now and then during service. Which means the only time anybody assesses the product is standing up, between orders, with a queue forming and a milk jug steaming — the exact condition under which nobody can taste carefully. It is also why quality problems in cafés are almost never caught by the café. They are caught by a regular who stops coming and never says why.

The fix is not a certificate or a lab. It is half an hour a week, sitting down, with the coffee in front of you and someone writing things down.

Why tasting during service does not work

Three reasons, and all of them are about conditions rather than skill. Your palate is already saturated — by the time you pull that check shot you have been in a room full of coffee aroma for four hours. You are tasting one drink in isolation, so you have nothing to compare it against except your memory of a drink you had yesterday, which is not reliable. And you are rushing, so you are really only checking that nothing is badly wrong, not whether anything has moved.

Quality in a café rarely fails suddenly. It drifts. A grinder wears, a new bag behaves slightly differently, a barista adjusts a little to the left and then a little further, the water filter approaches the end of its life. Each step is small enough to pass a mid-service check. Six weeks of them is a noticeably different cup, and nobody in the building noticed it happen, because everyone recalibrated their expectations along with it. A structured comparison is the only thing that catches a slow movement, because it puts two reference points side by side instead of relying on anyone's memory.

What to put side by side

The value is entirely in the comparison, so never taste a single coffee alone. Pick a small number of pairings and keep them consistent week to week — the same comparisons, repeated, are what turn individual impressions into a trend you can see.

• The new bag against the one you have been serving. The single most useful comparison you can run, and the one that catches a roaster's batch variation before your customers do.
• Your espresso against your filter, from the same coffee. If the filter tastes sweet and clean while the espresso tastes harsh, the problem is your extraction or your machine, not the beans.
• Your drink against the café down the road. Buy one, bring it back, taste them together. Uncomfortable, and more informative than any internal check.
• The same coffee at different ages. A bag just rested against one near the end of its life tells you whether your ordering rhythm matches how fast you actually get through it.
• Milk as it is actually served. Most of what you sell is a milk drink, so a tasting that only covers black coffee is missing the majority of your sales.

Brew the comparison the same way every week, with the same water, the same vessels and the same recipe you use on the bar. The absolute method matters far less than keeping it identical, because anything you change becomes a variable that hides the one you are looking for. Formal cupping protocols exist and are worth learning if you have the appetite, but you do not need the full apparatus to run something useful on a Tuesday morning.

Who should be in the room

Not only the owner, and not only the head barista. The two or three people who actually make most of your drinks should be there, because the tasting is half quality control and half training — it is where a team builds a shared vocabulary for what your coffee is supposed to taste like. A barista who has sat down and tasted an over-extracted shot next to a good one, with everyone naming what they are noticing, will catch that shape on the bar at four in the afternoon. One who has only been told to watch the timer will not.

Taste before anyone speaks. The moment the most senior person in the room says the word bitter, everyone tastes bitter, and the exercise becomes an agreement rather than an assessment. Everyone writes a note first, then you compare. This also surfaces the genuinely useful disagreements — if three people describe the same cup three different ways, something in it is unstable. Run it on a fixed day at a quiet hour and protect it, because a tasting that happens when there is time happens about once a quarter. It sits naturally alongside the rest of your barista training programme rather than as a separate thing to find time for.

Writing it down so a drift becomes visible

Here is the part that makes this worth doing rather than merely enjoyable: an impression you did not record is gone by Thursday. The point of the log is not the individual entry, it is the column you can read down in three months to see that something has been moving steadily in one direction.

Keep it short enough that it actually gets filled in. A phone note or a shared sheet is fine; an elaborate form is a thing that gets abandoned by week four.

Note the equipment context too. Grinder burrs wear gradually and the resulting change is almost impossible to attribute without a record — the same goes for a water filter approaching the end of its service life. Keep your tasting log and your maintenance dates near each other and the pattern usually explains itself. For the daily side of this, a dialling-in routine handles the morning calibration; the weekly tasting is what tells you whether the thing you are dialling in has itself moved. And if the coffee has been sitting too long before it reaches the hopper, the cause may be upstream — how you store beans shows up in the cup faster than most owners expect.

What to escalate to the roaster, and how

A tasting routine is also the only thing that gives you standing with your supplier. Walking in and saying the coffee tastes off is a conversation that goes nowhere, because they cannot act on it and you cannot support it. Walking in with four weeks of notes, the roast dates and a specific description — this batch is noticeably flatter than the last two, here is when it changed — is a conversation that gets a sample, a replacement, or at minimum a real explanation.

Escalate when a change persists across more than one session and you have ruled out your own equipment, when a batch is clearly different from the same coffee you had before, or when roast dates start arriving older than you agreed. Keep it factual rather than adversarial. A good roaster wants this information, because you are the only person who can tell them how their coffee is performing in a working bar. If delivery age, batch consistency or who absorbs a bad batch is not written down anywhere, that belongs in your supplier agreement rather than in a difficult phone call.

One more honest limit: a tasting panel tells you about the coffee, not about the cup your customer received. Those are different questions. The drink that left your bar at the busiest hour, made by whoever was fastest, is the one your reputation rests on — which is what a mystery shopper check is for. Run both and you cover the product and the delivery of it.

Where MidaOne fits

The tasting itself needs paper and attention, not software. What your system can tell you is the context around it: when the bag arrived, how fast you are getting through it, and whether what you tasted matches what customers did. MidaOne tracks stock as you sell, so you can see whether your coffee is turning over fast enough that freshness is not the variable. Sales by item and by day sit alongside it, so a week where the tasting flagged something and espresso sales dipped is a pattern you can see rather than a suspicion. Honest about the limit: the judgement is yours and the log is yours. The system just keeps the surrounding facts straight.

Frequently asked questions

Nobody opens a café intending to serve worse coffee than they did last year, and yet it is the most common thing that happens, quietly, to businesses that are otherwise doing fine. Half an hour a week with the cups lined up is a small price for being the first to know.