Why 99.2% and 97.8% on the same lot can both be correct

Why 99.2% and 97.8% on the same lot can both be correct I have a specific reason for asking rather than idle curiosity, and the context is below.

I would like to understand what this number means before I repeat it anywhere.

A Janoshik report on a retatrutide lot gives 98% purity. The supplier certificate for the same lot states 98.8%. Both documents name a reversed-phase method; neither states the same gradient.

My question is not "who is right". It is: given that those two figures were produced by different methods, what is the largest difference I should expect from method alone, and at what point does a gap stop being explainable that way?

Taking the opening post at face value and following it one step further.

System suitability is not paperwork. If the replicate injections failed, the run did not happen — the numbers from it are uninterpretable rather than approximate.

That is where I would start, not where I would stop.

Two things can be true about 99.2% and 97.8% at once: the mechanism is plausible and the evidence for the size of the effect is thin. Most of the argument here is people defending the first against attacks on the second.

Agreed on all of that, and I have nothing to add to it.

Adding the boring version of 99.2% and 97.8%, because the interesting version keeps getting posted and the boring one is usually right.

Check the ordinary explanations, in order, and stop when one of them accounts for what you are seeing. Most of the time the second one does.

This follows post #5 rather than contradicting it.

On integration: where the baseline is drawn matters more than most people realise. On a clean chromatogram with well-resolved peaks the choice is inconsequential. On a chromatogram with a trailing shoulder or a rising baseline it matters. Differences of one to two percentage points between defensible integrations are ordinary.

A qualification I should have led with rather than closed on.

The confident answers on 99.2% and 97.8% and the well-sourced answers are not the same answers, which is the most useful thing I have learned reading this category.

Change the wavelength and the proportions change even though the sample has not. That is the reason the wavelength has to be on the document for the number to mean anything.

I changed my mind about 99.2% and 97.8% after someone here asked me for the source and I could not produce one. That is worth saying out loud because it is the ordinary way it happens.

99.2% and 97.8% is well covered in the tag pages, and the older discussions are better than the recent ones because they were argued out properly. Worth twenty minutes before adding to this one.

Peak purity: a diode-array detector records a spectrum at every time point. If a peak contains two co-eluting species with different spectra, the spectrum changes across the peak. A passing peak-purity result says the spectrum is constant; it is weak evidence of homogeneity if the impurities have similar spectra.

I would rather say I do not know than round it up to an answer.

Adding the measurement that post #10 says would settle it.

Taking 99.2% and 97.8% seriously for a moment rather than deflecting: the honest position is that the community has observations and no controlled comparison, and those two things support very different sentences.

Typical suitability criteria are a replicate area relative standard deviation below about two per cent, a tailing factor inside a defined window, a resolution minimum against a specified peak, and a plate-count floor.