Mobile Benchmarks 2026

Mob il e r e te n tio n
b en ch m ark s 2 0 26
Key h ig h lig h ts f r o m t h e
G am eA naly tic s r e p ort
Julia Iljuk

D1, D 7, a n d D 30 r e te n tio n d eclin ed
aga in i n 2 0 25 .
R ete n tio n k e e p s g e ttin g
hard er
Th e g a p b etw een a v e ra ge g a m es a n d
to p p erfo rm ers k e ep s w id en in g.

~2 2 %
Med ia n D 1
re te n tio n
~ 4 0 %
D1 r e te n tio n i n
t o p 1 0 % g a m es
R ete n tio n D 1

~4 %
Med ia n D 7
re te n tio n
~ 1 2 %
D7 r e te n tio n i n
t o p 1 0 % g a m es
R ete n tio n D 7

If v alu e i s n ‘t c le ar i n t h e f ir s t 5 –1 5
m in ute s , p la y e rs c h urn .
E arly e n ga ge m en t i s
e ve ry th in g
Desig n o n b oard in g a s a
d em on stra tio n o f p le asu re , n ot j u st
a t u to ria l.
© O xana F omina

~0 .7 %
Med ia n D 30
re te n tio n
~ 4 %
D30 r e te n tio n i n
t o p 1 0 % g a m es
R ete n tio n D 30

Con te xt m atte rs
GameAnaly tics is a great
accessible t ool widely used b y
indies and early-stag e teams ,
while man y top -perf orming
studios st ore da ta in-house and
never appear in public
benchmarks .
In addition, g enre dif fer ences
need t o be tak en into accoun t.
Long -term r enention in h ybrid
casual and midcor e games will
not be the same .
Th ese m ed ia n n um bers a re r a th er
sh ockin g, b ut t h ey n eed c o nte xt.

So, i s m ob il e r e te n tio n
“b ro ke n ”?
Is the situa tion so bad? I w ouldn’t sa y so .
Does the classic 40 /20/1 0 (D1 /D7/D30)
ret en tion rule need upda ting? Probably .
A more realistic benchmark t oday can look lik e:
35 /1 5 /5
Wha t do y ou think?
Share your perspectiv e in the commen ts
below.