14
#measure engagement FOCUS: DAU / MAU – THE WHAT, WHY, AND HOW MEASURECAMPNL 2016-04-09, JULES STUIFBERGEN

#Measure Engagement - DAU / MAU and AAD

Embed Size (px)

Citation preview

Page 1: #Measure Engagement - DAU / MAU and AAD

#measure engagementFOCUS: DAU / MAU – THE WHAT, WHY, AND HOW

MEASURECAMPNL 2016-04-09, JULES STUIFBERGEN

Page 2: #Measure Engagement - DAU / MAU and AAD

Some metricsPage metrics

time on page

exit rate

scroll-to-bottom

Session metrics

visit duration

pageviews per visit

User metrics

sessions per user

# multi-session users

More compound metrics

Return visits within x days

DAU/MAU and Average Active Days

Page 3: #Measure Engagement - DAU / MAU and AAD

What is DAU/MAU ?Definition:

DAU (Daily Active Users)

MAU (Monthly Active Users)

Hence:

DAU / MAU = percentage of monthly active users that we see active this day

Page 4: #Measure Engagement - DAU / MAU and AAD

What does it say?https://mixpanel.com/help/platform/tutorials/dau-mau

DAU/MAU is a common metric used to measure stickiness (i.e. how engaged your users are). It's derived by dividing DAU (Daily Active Users) by MAU (Monthly Active Users).

http://www.startupdefinition.com/dau-over-mau

The ratio of Daily Active Users to Monthly Active Users. This metric is used to measure the depth of engagement of a company's users.

Page 5: #Measure Engagement - DAU / MAU and AAD

High = good (mostly)Same users visit often

Loyal visitors

Sticky Site

For games, content-sites, community sites, auction sites.. when you want habit or loyalty, you want high DAU/MAU rates.

Low DAU/MAU rates?

Customer service pages, .. ? ..

Page 6: #Measure Engagement - DAU / MAU and AAD

Average DAU/MAU and ”AAD”http://techcrunch.com/2009/10/29/how-to-measure-the-true-stickiness-and-success-of-a-facebook-app/

[An Average] [..] DAU/MAU ratio of 50% would mean that the average user of your app is using it 15 out of 30 days that month. [ correction mine ]

The sum of all DAU/MAU ratio’s in a period: Let’s call that “Average Active Days”: AAD

AAD / [number of days] = Average DAU/MAU

sum(DAU day 1, .., DAU day n) / MAU = 15

sum(DAU day 1 / MAU, … DAU day n / MAU) = 15

sum(DAU day 1 / MAU, … DAU day n / MAU) / n = 15/30 = 50%

Page 7: #Measure Engagement - DAU / MAU and AAD

So, how to calculate in Google Analytics?Static (once for a period)

period = 28 days (4 whole weeks)

Query 1: users per day (28 rows)

Query 2: users per period (1 row)

DAU / MAU for each day = Q1 day 1 / Q2 … Q1 day n / Q2

AAD = sum(Q1) / Q2

Page 8: #Measure Engagement - DAU / MAU and AAD

So, how to calculate in Google Analytics?Rolling (useful for segmented traffic (e.g. New Users), or to correct for period before peaks:

period = 28 days (4 whole weeks)

Query 1 28x (for every day in period): users per day (28 rows per query)use only 1 row per query

Query 2 28x (for every day in period): users per period (1 row per query)

Page 9: #Measure Engagement - DAU / MAU and AAD

Automate it (in R)library(rga)

i <- "12345" # Google Analytics View ID

mau <- ga$getData(i, start.date = s, end.date = e,

metrics = "ga:users", dimensions = "",

samplingLevel = "HIGHER_PRECISION")

dau <- ga$getData(i, start.date = s, end.date = e,

metrics = "ga:users", dimensions = ”ga:date",

samplingLevel = "HIGHER_PRECISION")

Page 10: #Measure Engagement - DAU / MAU and AAD

Automation is useful for rolling..dau.rolling <- data.frame()

for (day in seq(from = period.end – 27, to = period.end, by=1) {

q <- ga$getData(i, start.date = day, end.date = day + 27,

metrics = "ga:users", dimensions = "ga:date"

segment = "gaid:-101") #multi-session users (built-in)

# add the row we want

dau.rolling <- rbind(dau.rolling, q[ which(q$date == day),])

}

Page 11: #Measure Engagement - DAU / MAU and AAD

Plot it in ggplot

Multi-Session users compared to ALL traffic.

Logically, this differs.

Also plotted: the Average DAU/MAU (which is AAD)

Source: not-so-sticky website

Missing: absolute numbers

Page 12: #Measure Engagement - DAU / MAU and AAD

All sessions vs Customers: Big difference (Same site, all users vs. UID view)

Page 13: #Measure Engagement - DAU / MAU and AAD

Static vs Rolling: see the pre-peak difference(AAD higher)

Page 14: #Measure Engagement - DAU / MAU and AAD

Comments / Discussion

Would YOU use DAU/MAU? Is it useful?

How would you communicate what it means?

Useful segments to compare?

BTW: script available for R-lovers, ask me @zjuul