There is a disagreement in my family about chai that has been running since before I was born. It concerns the order of operations and everybody involved is certain.
Last year I made all four versions in one afternoon and tasted them next to each other. The results settled nothing, because everybody decided their version had won, and I did learn something.
The four methods
My father boils water with tea and spices first, then adds milk and brings it back up.
My aunt boils milk and water together from the start with everything in it.
My grandmother boiled water and spices, added tea, then milk, and brought it to a boil three separate times with a lowering of heat in between.
My cousin, who is the youngest and the most dismissed, adds everything cold to the pot and brings it up once, slowly.
What actually differed
More than I expected, and along dimensions I had not anticipated.
Strength of tea was highest in my father's version. Tea extracts more effectively in water than in milk, because the fat and protein in milk bind some of the compounds. Boiling the tea in water first gets more out of it.
Body was highest in my aunt's, where the milk cooked longest. Prolonged heating changes milk proteins and produces a thicker mouthfeel, closer to the very reduced chai you get from a good stall.
Spice intensity was highest in my grandmother's, presumably because the repeated boiling and settling gave more extraction time.
My cousin's was the mildest of all four and also the smoothest. Nobody expected that and nobody admitted it.
The bitterness question
The main argument in the family is about bitterness, and here there was a clear result.
Prolonged boiling of tea leaves in water produces bitterness, because tannins keep extracting. My father's version was the most bitter, which is exactly what he likes and what my aunt objects to.
Adding milk earlier limits that, because the milk proteins bind tannins. So the milk-early methods are genuinely less bitter, and that is a chemical effect rather than a preference.
Which means the argument is not about who is right. It is about whether you want the bitterness, and both sides have been arguing past each other for forty years.
The things everybody agreed on
Interestingly, there were some.
Ginger crushed rather than sliced or grated. Everybody does this and everybody says it releases more.
Cardamom crushed at the moment of use, not in advance.
Loose tea rather than bags, and a strong small-leaf tea rather than anything delicate. Delicate tea is wasted here, since it will be boiled with milk and spice.
And sugar in the pot rather than in the cup, which everybody insisted made a difference and nobody could explain. I could not detect a difference in the test and I have kept doing it anyway.
What the stalls do
Worth noting since stall chai is what most people are actually comparing against.
Most vendors keep a large pot at a rolling boil more or less continuously, and it reduces over the course of hours. That reduction is a large part of why stall chai has a concentration that home versions do not.
They also generally use more tea and more sugar than home cooks do, by a margin that surprised me when I asked.
And the milk is frequently boiled and reboiled for a long time, which produces exactly the thick, slightly caramelised character that is hard to reproduce in ten minutes.
So the home version cannot match it without either time or a different technique, which is a fair thing to know before being disappointed.
Where I landed
My method now is a hybrid that would please nobody in my family.
Water, crushed ginger and cardamom brought to a boil and held for a couple of minutes to extract the spices. Tea added and given about ninety seconds. Milk added, brought back to a boil, then taken off almost immediately.
This gets the spice extraction of the long-boil methods, most of the tea strength, and stops before the bitterness becomes dominant.
My father says it is weak. My aunt says it is bitter. I take that as a reasonable indication that it is somewhere in the middle.
The water is not neutral
One variable none of us had considered until I moved cities and my chai changed without any change in method.
Hard water affects tea extraction noticeably. In a hard water area the same tea produces a duller, flatter cup, and the difference is large enough that professional tasters control for it.
There is not much to be done about it domestically beyond using filtered water, which does help and which I now do without being able to say precisely which mineral is responsible.
It is worth knowing mostly as an explanation. If your chai suddenly tastes different and nothing else has changed, the pipes are a more likely culprit than the tea.
Reheating
Reheated chai is worse and everybody knows it. The reason is that the tea continues to extract while it sits and the milk proteins have already been through one heat cycle.
Making smaller quantities more often is the only real answer, which is inconvenient and is what every stall effectively does by keeping the pot moving.