With ten funds on a menu, the naive approach puts 10% in each. It sounds unsophisticated, and it ignores everything an optimiser would consider, yet it is surprisingly hard to beat out of sample because it makes no estimation errors.
The catch is that 1/N is only as good as the menu. If seven of the ten choices are equity funds, the equally weighted portfolio is 70% equities whether or not that suits the investor. Equal weighting across a badly chosen list produces a badly chosen portfolio.
Use it as a sanity benchmark. If a carefully optimised allocation cannot beat 1/N across a realistic range of assumptions, the optimisation is probably fitting noise. See mean-variance-optimization.
Related: diworsification, mean-variance-optimization, asset-allocation, diversification