How so? It was a routine occurrence for steppe confederations to launch raids on their agricultural neighbors, but that doesn't really make the Mongol conquests any less brutal.
So first of all, the relationship between the steppes nomads and more traditional civilizations wasn't one sided. Steppes Nomads did raid, but just as often were sponsored by Empires such as those of the Iranian, Chinese and Roman civilizations to war against each other and their enemies. Likewise, they were often tributaries and vassals of those civilizations. Genocides done on the steppes nomads and by the steppes nomads at the behest of Agricultural states was very normal. The mongols were not exceptional in this regard (at first anyway, after Temujin's death things got a bit chaotic).
But indeed under him, the nomadic conquering was a hell of a lot milder than normal in its brutality as I will get to in a second, but first I shall quickly address the issues with determining how many people did actually die due to Genghis Khan.
So Genghis Khan is known in his death toll for his conquest of northern China and Persia.
China: with his conquest of the Jurchen, its important to consider that China was in the middle of a horrific drought. A large death toll was naturally expected, along with the usual (or as I will get to, not so usual) fatalities of war.
Genghis Khan however had an interesting policy regarding peasants. Wanting them for his empire, he did some rather brilliant psychological warfare by scaring them off, burning their villages and essentially herding them towards the cities with the hopes of demoralising the enemy. Why this is so unusual is that it was the usual steppe routine to execute the males and take the women and children as slaves. This was a continuation of Genghis Khan's unusual policy of absorbing tribes into the mongol identity by killing the leaders but otherwise accepting them in (yet another way he was far less bloodthirsty than usual steppe fare).
So now you have a drought, cities taking on larger populations and more importantly literal resource drains when needed most. Millions died.
When we quantify those deaths, how do we accurately consider which deaths were due to Genghis Khan, and which deaths were a part of the existing wars going on at the time/drought. Undoubtedly Genghis Khan added to the situation, but the end number of deaths given in most text books usually puts all related drought deaths on him. In short, we dont actually have an accurate account of how many died in northern China due to him, and unfortunately can't get an accurate number either.
Persia: as I'm running out of time, he is most well known for the massacre at Urgench. The primary source we have for this however is a persian historian who Genghis Khan hired to make him sound more fearsome and generally recorded more deaths in the conquest of a given city than there were actually people living in it (as we would see with Hulagu and Baghdad). When we go outside of the mongol propaganda machine, we see historians such as Ibn Batuta (apologies if the name is spelt incorrectly) in the early 14th century commenting about how populous the city was. All the more remarkable this is when you factor in the eventual Timurid genocide of the city with which it did not recover.
In short, our source for how many died in Persia is a guy who literally had the job to exacerbate numbers. Like in China, the death toll is hard to quantity.
Finally there is just bad history that sells well. Despite the academic concencus being now about how difficult it is to calculate how many died, it really sells well on the history channel and in history books to make the mongols as vicious as possible by going with the higher estimates. Shockingly, this practice is made worse by often including deaths attributed to his progeny after his death onto the kill count.