Nearly a third of coded diagnostics rolling into workshops carry multiple fault codes simultaneously. When a car rolls into a bay, the textbook idea of one code leading to one fix rarely survives contact with the OBD reader.
Looking at our own platform data between February and October 2026, we pulled 3,495 total diagnostic sessions. Out of the 1,928 that included at least one readable fault code, 598 of them carried multiple codes simultaneously. That means 31 percent of coded jobs are multi-code jobs. The average across all those coded sessions was 1.72 readable codes per diagnostic.
The single-code illusion
When a technician gets just one code, which happened in 38 percent of our total sample, they treat it like a straight lookup. They only bother to write a symptom description 61 percent of the time. The expectation is that the bare code will point directly to the broken part.
Things change the moment a second code appears. Two codes together appeared in 8 percent of all diagnostics, and this is typically where you see a root cause and its direct consequence. When technicians face a pair of codes, they write a symptom description 76 percent of the time.
Cause and consequence
You see this in the common exact code sets. We saw P0562 and U0416 together 11 times, most often on an Audi. We saw P0102 and P0193 together 9 times, usually on a Land Rover. Reading the pair correctly beats looking up each code in isolation. The combination tells a story that a single code cannot. We also see sets like P0401, P042E, P049C, P1402, and P1412 together, which showed up 5 times, frequently on a Volvo.
The extremes of the diagnostic spectrum
The most common scenario in our data does not involve codes at all. Exactly 1,567 diagnostics, or 45 percent of the total, started with no codes. Either the fault throws no code, or the technician is asking about symptoms before plugging in the scanner. In these cases, mechanics wrote a description 93 percent of the time.
At the other end of the spectrum, 2 percent of jobs present six or more codes. This looks less like a targeted diagnostic and more like a full memory dump. For example, a massive string of P0171, P1141, P1171, P1542, P1564, P1565, and P1568 appeared 12 times, usually on a Volkswagen. Faced with a dump like this, mechanics write a description 89 percent of the time. These massive code sets also trigger a follow-up chat 10 percent of the time, suggesting that a huge list of codes needs more manual sorting to make sense of the real fault.
What this means for the bay
Treating every code as an isolated event ignores how cars actually break. A diagnostic approach needs to handle the reality that a third of coded jobs are going to present a web of related faults. Are you seeing this same split between single-code lookups and complex multi-code jobs in your own workshop?
