Do you need a tracker for search arbitrage?
In practice, yes. Costs live in the traffic source and revenue lives in the feed, and neither sees the other. A tracker joins them: it records each visit, pulls in feed revenue, calculates profit by campaign, creative and keyword, and sends revenue back to the traffic source so automated bidding can optimise for money earned.
A tracker gives every visit a Click ID and stores where it came from using UTM parameters and a Sub ID. When revenue arrives from the feed, through an API or a postback, it attaches the money to the visit.
Without this you cannot tell which of fifty creatives is paying for itself, and the traffic platform optimises only for cheap clicks. Some hosted feed providers include basic tracking, which is enough for a first test. As spend grows, a dedicated tool also runs automation rules, for example pausing an ad when trailing ROI falls below a threshold. Remember that most feed revenue arrives late and is estimated, so tracker profit is provisional.