Large language models generate inconsistent brand and product recommendations, making precise ranking measurements unreliable, according to a study by market research firm SparkToro and…