What you would probably need for this to occur is the Japanese becoming very enamored of airpower in general. For a divergence point, have the Japanese attempt to hold on to their gains in Eastern Siberia during and after the RCW. They manage to keep northern Sakhalin, but are eventually driven off of the mainland by weight of Soviet numbers. As a result of this experience, the Japanese prioritize force multipliers as a way of allowing them to fight on more or less equal terms with their potential enemies, all of whom vastly outweigh the Japanese in population or industry.
A high-quality air arm is seen as a particularly valuable method of achieving this, allowing the Japanese to attrit enemy forces that are still well behind the front lines. If you want, have the Japanese score one or two spectacular successes against the Soviets due to aircraft while still losing the overall war, giving aircraft a boost in the opinions of the military establishment.
The other element would probably be a different attitude by the Japanese going into the Washington Naval Conference, perhaps accompanied by the Japanese actually managing for once to use a diplomatic code that the US hadn't thoroughly penetrated. Have the Japanese agree, as a result of their experiences with aircraft and as a result of having observed things like Billy Mitchell's test of the effectiveness of aircraft against capital ships, to give up on most of the battleship tonnage they historically received in exchange for aircraft carrier tonnage. Britain and especially the US, both of whom were still very much in favor of the large battleline at this time, happily agree to what they see as the Japanese wasting naval resources.
This would probably cause somewhat of a change in what Japanese battleships would be built to do. With carriers having taken over the job of destroying the enemy fleet, battleships would probably be designed for shore bombardment in support of amphibious operations, as command ships, and as a more resilient defense platform for the carriers against air attack or any enemy capital ships that somehow did reach the carrier formation. This probably leads the Japanese to prioritize speed for the latter two purposes in order for their battleships to keep up with the carriers.
Other major changes are that the Japanese air arm will be much more effective than in OTL due to having more resources dedicated to improving it. Also figure that keeping all of Sakhalin and therefore getting the oil fields reduces Japanese worries about the availability of fuel; one big consequence is that they can build their aircraft engines for power rather than efficiency, allowing them to put better armor and weaponry on their aircraft without having to sacrifice too much range or maneuverability. And with the increased attention to aircraft, the Japanese probably realize the potential of radar much sooner than in OTL. Having effective radar by the mid to late 1930s would vastly change the Japanese naval situation; early raid warning could probably have prevented them from being caught flat-footed at Midway, for example.
It also might be a more remote possibility, but Japanese foreign policy might be a little more cautious, particularly in regard to China, as a result of being mauled by the Soviets in the 1920s. The experience of being thoroughly beaten due to trying to grab too much too fast from a stronger opponent should encourage the Japanese military to consolidate its gains before moving further, and to not try to take more than they think they can control. For example, while the Japanese would probably still grab Manchuria in this scenario, I doubt that the Second Sino-Japanese War would occur as in OTL. More likely is the Japanese going for some specific strategic targets, such as the Shandong Peninsula, Hainan, and possibly Shanghai, while attempting to support the KMT's rivals within China in the hopes of throwing it back into civil war.
What other effects might a Japanese shift towards attempts to offset its numerical disadvantage via quality and technology have?